remove all tx generation in runner

The 0.35 release cycle renamed the 'fastsync' functionality to 'blocksync'. This change brings the configuration parameters in line with that change. Namely, it updates the configuration file `[fastsync]` field to be `[blocksync]` and changes the command line flag and config file parameters `--fast-sync` and `fast-sync` to `--enable-block-sync` and `enable-block-sync` respectively.

Error messages were added to help users encountering these changes be able to quickly make the needed update to their files/scripts.

When using the old command line argument for fast-sync, the following is printed

```
./build/tendermint start --proxy-app=kvstore --consensus.create-empty-blocks=false --fast-sync=false
ERROR: invalid argument "false" for "--fast-sync" flag: --fast-sync has been deprecated, please use --enable-block-sync
```

When using one of the old config file parameters, the following is printed:
```
./build/tendermint start --proxy-app=kvstore --consensus.create-empty-blocks=false
ERROR: error in config file: a configuration parameter named 'fast-sync' was found in the configuration file. The 'fast-sync' parameter has been renamed to 'enable-block-sync', please update the 'fast-sync' field in your configuration file to 'enable-block-sync'
```

.github/codecov.yml

@@ -1,27 +0,0 @@
-coverage:
-  precision: 2
-  round: down
-  range: "70...100"
-  status:
-    project:
-      default:
-        threshold: 1%
-    patch: on
-    changes: off
-
-github_checks:
-  annotations: false
-
-comment:
-  layout: "diff, files"
-  behavior: default
-  require_changes: no
-  require_base: no
-  require_head: yes
-
-ignore:
-  - "docs"
-  - "DOCKER"
-  - "scripts"
-  - "**/*.pb.go"
-  - "libs/pubsub/query/query.peg.go"

.github/workflows/coverage.yml

@@ -121,7 +121,3 @@ jobs:
       - run: |
           cat ./*profile.out | grep -v "mode: atomic" >> coverage.txt
         if: env.GIT_DIFF
-      - uses: codecov/codecov-action@v2.0.2
-        with:
-          file: ./coverage.txt
-        if: env.GIT_DIFF

.github/workflows/docker.yml

@@ -40,7 +40,7 @@ jobs:
           platforms: all
 
       - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v1.5.0
+        uses: docker/setup-buildx-action@v1.6.0
 
       - name: Login to DockerHub
         if: ${{ github.event_name != 'pull_request' }}

.github/workflows/proto-docker.yml

@@ -34,7 +34,7 @@ jobs:
           echo ::set-output name=tags::${TAGS}
 
       - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v1.5.0
+        uses: docker/setup-buildx-action@v1.6.0
 
       - name: Login to DockerHub
         uses: docker/login-action@v1.10.0

.github/workflows/release.yml

@@ -2,7 +2,7 @@ name: "Release"
 
 on:
   push:
-    branches: 
+    branches:
       - "RC[0-9]/**"
     tags:
       - "v[0-9]+.[0-9]+.[0-9]+" # Push events to matching v*, i.e. v1.0, v20.15.10
@@ -20,9 +20,6 @@ jobs:
         with:
           go-version: '1.16'
 
-      - run: echo https://github.com/tendermint/tendermint/blob/${GITHUB_REF#refs/tags/}/CHANGELOG.md#${GITHUB_REF#refs/tags/} > ../release_notes.md 
-        if: startsWith(github.ref, 'refs/tags/')
-
       - name: Build
         uses: goreleaser/goreleaser-action@v2
         if: ${{ github.event_name == 'pull_request' }}
@@ -35,6 +32,6 @@ jobs:
         if: startsWith(github.ref, 'refs/tags/')
         with:
           version: latest
-          args: release --rm-dist --release-notes=../release_notes.md
+          args: release --rm-dist
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

.golangci.yml

@@ -1,14 +1,17 @@
 linters:
   enable:
+    - asciicheck
     - bodyclose
     - deadcode
     - depguard
     - dogsled
     - dupl
     - errcheck
+    - exportloopref
     # - funlen
     # - gochecknoglobals
     # - gochecknoinits
+    # - gocognit
     - goconst
     - gocritic
     # - gocyclo
@@ -22,11 +25,11 @@ linters:
     - ineffassign
     # - interfacer
     - lll
-    - misspell
     # - maligned
+    - misspell
     - nakedret
+    - nolintlint
     - prealloc
-    - scopelint
     - staticcheck
     - structcheck
     - stylecheck
@@ -37,9 +40,6 @@ linters:
     - varcheck
     # - whitespace
     # - wsl
-    # - gocognit
-    - nolintlint
-    - asciicheck
 
 issues:
   exclude-rules:

CHANGELOG.md

@@ -1,6 +1,160 @@
 # Changelog
 
-Friendly reminder, we have a [bug bounty program](https://hackerone.com/tendermint).
+Friendly reminder: We have a [bug bounty program](https://hackerone.com/tendermint).
+
+## v0.35
+
+Special thanks to external contributors on this release: @JayT106, @bipulprasad, @alessio, @Yawning, @silasdavis,
+@cuonglm, @tanyabouman, @JoeKash, @githubsands, @jeebster, @crypto-facs, @liamsi, and @gotjoshua
+
+### BREAKING CHANGES
+
+- CLI/RPC/Config
+  - [pubsub/events] \#6634 The `ResultEvent.Events` field is now of type `[]abci.Event` preserving event order instead of `map[string][]string`. (@alexanderbez)
+  - [config] \#5598 The `test_fuzz` and `test_fuzz_config` P2P settings have been removed. (@erikgrinaker)
+  - [config] \#5728 `fastsync.version = "v1"` is no longer supported (@melekes)
+  - [cli] \#5772 `gen_node_key` prints JSON-encoded `NodeKey` rather than ID and does not save it to `node_key.json` (@melekes)
+  - [cli] \#5777 use hyphen-case instead of snake_case for all cli commands and config parameters (@cmwaters)
+  - [rpc] \#6019 standardise RPC errors and return the correct status code (@bipulprasad & @cmwaters)
+  - [rpc] \#6168 Change default sorting to desc for `/tx_search` results (@melekes)
+  - [cli] \#6282 User must specify the node mode when using `tendermint init` (@cmwaters)
+  - [state/indexer] \#6382 reconstruct indexer, move txindex into the indexer package (@JayT106)
+  - [cli] \#6372 Introduce `BootstrapPeers` as part of the new p2p stack. Peers to be connected on  startup (@cmwaters)
+  - [config] \#6462 Move `PrivValidator` configuration out of `BaseConfig` into its own section. (@tychoish)
+  - [rpc] \#6610 Add MaxPeerBlockHeight into /status rpc call (@JayT106)
+  - [blocksync/rpc] \#6620 Add TotalSyncedTime & RemainingTime to SyncInfo in /status RPC  (@JayT106)
+  - [rpc/grpc] \#6725 Mark gRPC in the RPC layer as deprecated.
+  - [blocksync/v2] \#6730 Fast Sync v2 is deprecated, please use v0
+  - [rpc] Add genesis_chunked method to support paginated and parallel fetching of large genesis documents.
+  - [rpc/jsonrpc/server] \#6785 `Listen` function updated to take an `int` argument, `maxOpenConnections`, instead of an entire config object. (@williambanfield)
+  - [rpc] \#6820 Update RPC methods to reflect changes in the p2p layer, disabling support for `UnsafeDialPeers` and `UnsafeDialPeers` when used with the new p2p layer, and changing the response format of the peer list in `NetInfo` for all users.
+  - [cli] \#6854 Remove deprecated snake case commands. (@tychoish)
+
+- Apps
+  - [ABCI] \#6408 Change the `key` and `value` fields from `[]byte` to `string` in the `EventAttribute` type. (@alexanderbez)
+  - [ABCI] \#5447 Remove `SetOption` method from `ABCI.Client` interface
+  - [ABCI] \#5447 Reset `Oneof` indexes for  `Request` and `Response`.
+  - [ABCI] \#5818 Use protoio for msg length delimitation. Migrates from int64 to uint64 length delimiters.
+  - [ABCI] \#3546 Add `mempool_error` field to `ResponseCheckTx`. This field will contain an error string if Tendermint encountered an error while adding a transaction to the mempool. (@williambanfield)
+  - [Version] \#6494 `TMCoreSemVer` has been renamed to `TMVersion`.
+	- It is not required any longer to set ldflags to set version strings
+  - [abci/counter] \#6684 Delete counter example app
+
+- Go API
+  - [pubsub] \#6634 The `Query#Matches` method along with other pubsub methods, now accepts a `[]abci.Event` instead of `map[string][]string`. (@alexanderbez)
+  - [p2p] \#6618 \#6583 Move `p2p.NodeInfo`, `p2p.NodeID` and `p2p.NetAddress` into `types` to support use in external packages. (@tychoish)
+  - [node] \#6540 Reduce surface area of the `node` package by making most of the implementation details private. (@tychoish)
+  - [p2p] \#6547 Move the entire `p2p` package and all reactor implementations into `internal`.  (@tychoish)
+  - [libs/log] \#6534 Remove the existing custom Tendermint logger backed by go-kit. The logging interface, `Logger`, remains. Tendermint still provides a default logger backed by the performant zerolog logger. (@alexanderbez)
+  - [libs/time] \#6495 Move types/time to libs/time to improve consistency. (@tychoish)
+  - [mempool] \#6529 The `Context` field has been removed from the `TxInfo` type. `CheckTx` now requires a `Context` argument. (@alexanderbez)
+  - [abci/client, proxy] \#5673 `Async` funcs return an error, `Sync` and `Async` funcs accept `context.Context` (@melekes)
+  - [p2p] Remove unused function `MakePoWTarget`. (@erikgrinaker)
+  - [libs/bits] \#5720 Validate `BitArray` in `FromProto`, which now returns an error (@melekes)
+  - [proto/p2p] Rename `DefaultNodeInfo` and `DefaultNodeInfoOther` to `NodeInfo` and `NodeInfoOther` (@erikgrinaker)
+  - [proto/p2p] Rename `NodeInfo.default_node_id` to `node_id` (@erikgrinaker)
+  - [libs/os] Kill() and {Must,}{Read,Write}File() functions have been removed. (@alessio)
+  - [store] \#5848 Remove block store state in favor of using the db iterators directly (@cmwaters)
+  - [state] \#5864 Use an iterator when pruning state (@cmwaters)
+  - [types] \#6023 Remove `tm2pb.Header`, `tm2pb.BlockID`, `tm2pb.PartSetHeader` and `tm2pb.NewValidatorUpdate`.
+	- Each of the above types has a `ToProto` and `FromProto` method or function which replaced this logic.
+  - [light] \#6054 Move `MaxRetryAttempt` option from client to provider.
+	- `NewWithOptions` now sets the max retry attempts and timeouts (@cmwaters)
+  - [all] \#6077 Change spelling from British English to American (@cmwaters)
+	- Rename "Subscription.Cancelled()" to "Subscription.Canceled()" in libs/pubsub
+	- Rename "behaviour" pkg to "behavior" and internalized it in blocksync v2
+  - [rpc/client/http] \#6176 Remove `endpoint` arg from `New`, `NewWithTimeout` and `NewWithClient` (@melekes)
+  - [rpc/client/http] \#6176 Unexpose `WSEvents` (@melekes)
+  - [rpc/jsonrpc/client/ws_client] \#6176 `NewWS` no longer accepts options (use `NewWSWithOptions` and `OnReconnect` funcs to configure the client) (@melekes)
+  - [internal/libs] \#6366 Move `autofile`, `clist`,`fail`,`flowrate`, `protoio`, `sync`, `tempfile`, `test` and `timer` lib packages to an internal folder
+  - [libs/rand] \#6364 Remove most of libs/rand in favour of standard lib's `math/rand` (@liamsi)
+  - [mempool] \#6466 The original mempool reactor has been versioned as `v0` and moved to a sub-package under the root `mempool` package.
+	Some core types have been kept in the `mempool` package such as `TxCache` and it's implementations, the `Mempool` interface itself
+	and `TxInfo`. (@alexanderbez)
+  - [crypto/sr25519] \#6526 Do not re-execute the Ed25519-style key derivation step when doing signing and verification.  The derivation is now done once and only once.  This breaks `sr25519.GenPrivKeyFromSecret` output compatibility. (@Yawning)
+  - [types] \#6627 Move `NodeKey` to types to make the type public.
+  - [config] \#6627 Extend `config` to contain methods `LoadNodeKeyID` and `LoadorGenNodeKeyID`
+  - [blocksync] \#6755 Rename `FastSync` and `Blockchain` package to `BlockSync` (@cmwaters)
+
+- Data Storage
+  - [store/state/evidence/light] \#5771 Use an order-preserving varint key encoding (@cmwaters)
+  - [mempool] \#6396 Remove mempool's write ahead log (WAL), (previously unused by the tendermint code). (@tychoish)
+  - [state] \#6541 Move pruneBlocks from consensus/state to state/execution. (@JayT106)
+
+- Tooling
+  - [tools] \#6498 Set OS home dir to instead of the hardcoded PATH. (@JayT106)
+  - [cli/indexer] \#6676 Reindex events command line tooling. (@JayT106)
+
+### FEATURES
+
+- [config] Add `--mode` flag and config variable. See [ADR-52](https://github.com/tendermint/tendermint/blob/master/docs/architecture/adr-052-tendermint-mode.md) @dongsam
+- [rpc] \#6329 Don't cap page size in unsafe mode (@gotjoshua, @cmwaters)
+- [pex] \#6305 v2 pex reactor with backwards compatability. Introduces two new pex messages to
+  accomodate for the new p2p stack. Removes the notion of seeds and crawling. All peer
+  exchange reactors behave the same. (@cmwaters)
+- [crypto] \#6376 Enable sr25519 as a validator key type
+- [mempool] \#6466 Introduction of a prioritized mempool. (@alexanderbez)
+  - `Priority` and `Sender` have been introduced into the `ResponseCheckTx` type, where the `priority` will determine the prioritization of
+  the transaction when a proposer reaps transactions for a block proposal. The `sender` field acts as an index.
+  - Operators may toggle between the legacy mempool reactor, `v0`, and the new prioritized reactor, `v1`, by setting the
+  `mempool.version` configuration, where `v1` is the default configuration.
+  - Applications that do not specify a priority, i.e. zero, will have transactions reaped by the order in which they are received by the node.
+  - Transactions are gossiped in FIFO order as they are in `v0`.
+- [config/indexer] \#6411 Introduce support for custom event indexing data sources, specifically PostgreSQL. (@JayT106)
+- [blocksync/event] \#6619 Emit blocksync status event when switching consensus/blocksync (@JayT106)
+- [statesync/event] \#6700 Emit statesync status start/end event (@JayT106)
+- [inspect] \#6785 Add a new `inspect` command for introspecting the state and block store of a crashed tendermint node. (@williambanfield)
+
+### IMPROVEMENTS
+
+- [libs/log] Console log formatting changes as a result of \#6534 and \#6589. (@tychoish)
+- [statesync] \#6566 Allow state sync fetchers and request timeout to be configurable. (@alexanderbez)
+- [types] \#6478 Add `block_id` to `newblock` event (@jeebster)
+- [crypto/ed25519] \#5632 Adopt zip215 `ed25519` verification. (@marbar3778)
+- [crypto/ed25519] \#6526 Use [curve25519-voi](https://github.com/oasisprotocol/curve25519-voi) for `ed25519` signing and verification. (@Yawning)
+- [crypto/sr25519] \#6526 Use [curve25519-voi](https://github.com/oasisprotocol/curve25519-voi) for `sr25519` signing and verification. (@Yawning)
+- [privval] \#5603 Add `--key` to `init`, `gen_validator`, `testnet` & `unsafe_reset_priv_validator` for use in generating `secp256k1` keys.
+- [privval] \#5725 Add gRPC support to private validator.
+- [privval] \#5876 `tendermint show-validator` will query the remote signer if gRPC is being used (@marbar3778)
+- [abci/client] \#5673 `Async` requests return an error if queue is full (@melekes)
+- [mempool] \#5673 Cancel `CheckTx` requests if RPC client disconnects or times out (@melekes)
+- [abci] \#5706 Added `AbciVersion` to `RequestInfo` allowing applications to check ABCI version when connecting to Tendermint. (@marbar3778)
+- [blocksync/v1] \#5728 Remove blocksync v1 (@melekes)
+- [blocksync/v0] \#5741 Relax termination conditions and increase sync timeout (@melekes)
+- [cli] \#5772 `gen_node_key` output now contains node ID (`id` field) (@melekes)
+- [blocksync/v2] \#5774 Send status request when new peer joins (@melekes)
+- [store] \#5888 store.SaveBlock saves using batches instead of transactions for now to improve ACID properties. This is a quick fix for underlying issues around tm-db and ACID guarantees. (@githubsands)
+- [consensus] \#5987 and \#5792 Remove the `time_iota_ms` consensus parameter. Merge `tmproto.ConsensusParams` and `abci.ConsensusParams`. (@marbar3778, @valardragon)
+- [types] \#5994 Reduce the use of protobuf types in core logic. (@marbar3778)
+  - `ConsensusParams`, `BlockParams`, `ValidatorParams`, `EvidenceParams`, `VersionParams`, `sm.Version` and `version.Consensus` have become native types. They still utilize protobuf when being sent over the wire or written to disk.
+- [rpc/client/http] \#6163 Do not drop events even if the `out` channel is full (@melekes)
+- [node] \#6059 Validate and complete genesis doc before saving to state store (@silasdavis)
+- [state] \#6067 Batch save state data (@githubsands & @cmwaters)
+- [crypto] \#6120 Implement batch verification interface for ed25519 and sr25519. (@marbar3778)
+- [types] \#6120 use batch verification for verifying commits signatures.
+  - If the key type supports the batch verification API it will try to batch verify. If the verification fails we will single verify each signature.
+- [privval/file] \#6185 Return error on `LoadFilePV`, `LoadFilePVEmptyState`. Allows for better programmatic control of Tendermint.
+- [privval] \#6240 Add `context.Context` to privval interface.
+- [rpc] \#6265 set cache control in http-rpc response header (@JayT106)
+- [statesync] \#6378 Retry requests for snapshots and add a minimum discovery time (5s) for new snapshots.
+- [node/state] \#6370 graceful shutdown in the consensus reactor (@JayT106)
+- [crypto/merkle] \#6443 Improve HashAlternatives performance (@cuonglm)
+- [crypto/merkle] \#6513 Optimize HashAlternatives (@marbar3778)
+- [p2p/pex] \#6509 Improve addrBook.hash performance (@cuonglm)
+- [consensus/metrics] \#6549 Change block_size gauge to a histogram for better observability over time (@marbar3778)
+- [statesync] \#6587 Increase chunk priority and re-request chunks that don't arrive (@cmwaters)
+- [state/privval] \#6578 No GetPubKey retry beyond the proposal/voting window (@JayT106)
+- [rpc] \#6615 Add TotalGasUsed to block_results response (@crypto-facs)
+- [cmd/tendermint/commands] \#6623 replace `$HOME/.some/test/dir` with `t.TempDir` (@tanyabouman)
+- [statesync] \6807 Implement P2P state provider as an alternative to RPC (@cmwaters)
+
+### BUG FIXES
+
+- [privval] \#5638 Increase read/write timeout to 5s and calculate ping interval based on it (@JoeKash)
+- [evidence] \#6375 Fix bug with inconsistent LightClientAttackEvidence hashing (cmwaters)
+- [rpc] \#6507 Ensure RPC client can handle URLs without ports (@JayT106)
+- [statesync] \#6463 Adds Reverse Sync feature to fetch historical light blocks after state sync in order to verify any evidence (@cmwaters)
+- [blocksync] \#6590 Update the metrics during blocksync (@JayT106)
 
 ## v0.34.12
 

CHANGELOG_PENDING.md

@@ -9,153 +9,18 @@ Friendly reminder: We have a [bug bounty program](https://hackerone.com/tendermi
 ### BREAKING CHANGES
 
 - CLI/RPC/Config
-  - [pubsub/events] \#6634 The `ResultEvent.Events` field is now of type `[]abci.Event` preserving event order instead of `map[string][]string`. (@alexanderbez)
-  - [config] \#5598 The `test_fuzz` and `test_fuzz_config` P2P settings have been removed. (@erikgrinaker)
-  - [config] \#5728 `fast_sync = "v1"` is no longer supported (@melekes)
-  - [cli] \#5772 `gen_node_key` prints JSON-encoded `NodeKey` rather than ID and does not save it to `node_key.json` (@melekes)
-  - [cli] \#5777 use hyphen-case instead of snake_case for all cli commands and config parameters (@cmwaters)
-  - [rpc] \#6019 standardise RPC errors and return the correct status code (@bipulprasad & @cmwaters)
-  - [rpc] \#6168 Change default sorting to desc for `/tx_search` results (@melekes)
-  - [cli] \#6282 User must specify the node mode when using `tendermint init` (@cmwaters)
-  - [state/indexer] \#6382 reconstruct indexer, move txindex into the indexer package (@JayT106)
-  - [cli] \#6372 Introduce `BootstrapPeers` as part of the new p2p stack. Peers to be connected on  startup (@cmwaters)
-  - [config] \#6462 Move `PrivValidator` configuration out of `BaseConfig` into its own section. (@tychoish)
-  - [rpc] \#6610 Add MaxPeerBlockHeight into /status rpc call (@JayT106)
-  - [fastsync/rpc] \#6620 Add TotalSyncedTime & RemainingTime to SyncInfo in /status RPC  (@JayT106)
-  - [rpc/grpc] \#6725 Mark gRPC in the RPC layer as deprecated.
-  - [blockchain/v2] \#6730 Fast Sync v2 is deprecated, please use v0
-  - [rpc] \#6820 Update RPC methods to reflect changes in the p2p layer, disabling support for `UnsafeDialPeers` and `UnsafeDialPeers` when used with the new p2p layer, and changing the response format of the peer list in `NetInfo` for all users.
 
 - Apps
-  - [ABCI] \#6408 Change the `key` and `value` fields from `[]byte` to `string` in the `EventAttribute` type. (@alexanderbez)
-  - [ABCI] \#5447 Remove `SetOption` method from `ABCI.Client` interface
-  - [ABCI] \#5447 Reset `Oneof` indexes for  `Request` and `Response`.
-  - [ABCI] \#5818 Use protoio for msg length delimitation. Migrates from int64 to uint64 length delimiters.
-  - [ABCI] \#3546 Add `mempool_error` field to `ResponseCheckTx`. This field will contain an error string if Tendermint encountered an error while adding a transaction to the mempool. (@williambanfield)
-  - [Version] \#6494 `TMCoreSemVer` has been renamed to `TMVersion`.
-	- It is not required any longer to set ldflags to set version strings
-  - [abci/counter] \#6684 Delete counter example app
 
 - P2P Protocol
 
 - Go API
-  - [pubsub] \#6634 The `Query#Matches` method along with other pubsub methods, now accepts a `[]abci.Event` instead of `map[string][]string`. (@alexanderbez)
-  - [p2p] \#6618 Move `p2p.NodeInfo` into `types` to support use of the SDK. (@tychoish)
-  - [p2p] \#6583 Make `p2p.NodeID` and `p2p.NetAddress` exported types to support their use in the RPC layer. (@tychoish)
-  - [node] \#6540 Reduce surface area of the `node` package by making most of the implementation details private. (@tychoish)
-  - [p2p] \#6547 Move the entire `p2p` package and all reactor implementations into `internal`.  (@tychoish)
-  - [libs/log] \#6534 Remove the existing custom Tendermint logger backed by go-kit. The logging interface, `Logger`, remains. Tendermint still provides a default logger backed by the performant zerolog logger. (@alexanderbez)
-  - [libs/time] \#6495 Move types/time to libs/time to improve consistency. (@tychoish)
-  - [mempool] \#6529 The `Context` field has been removed from the `TxInfo` type. `CheckTx` now requires a `Context` argument. (@alexanderbez)
-  - [abci/client, proxy] \#5673 `Async` funcs return an error, `Sync` and `Async` funcs accept `context.Context` (@melekes)
-  - [p2p] Remove unused function `MakePoWTarget`. (@erikgrinaker)
-  - [libs/bits] \#5720 Validate `BitArray` in `FromProto`, which now returns an error (@melekes)
-  - [proto/p2p] Rename `DefaultNodeInfo` and `DefaultNodeInfoOther` to `NodeInfo` and `NodeInfoOther` (@erikgrinaker)
-  - [proto/p2p] Rename `NodeInfo.default_node_id` to `node_id` (@erikgrinaker)
-  - [libs/os] Kill() and {Must,}{Read,Write}File() functions have been removed. (@alessio)
-  - [store] \#5848 Remove block store state in favor of using the db iterators directly (@cmwaters)
-  - [state] \#5864 Use an iterator when pruning state (@cmwaters)
-  - [types] \#6023 Remove `tm2pb.Header`, `tm2pb.BlockID`, `tm2pb.PartSetHeader` and `tm2pb.NewValidatorUpdate`.
-	- Each of the above types has a `ToProto` and `FromProto` method or function which replaced this logic.
-  - [light] \#6054 Move `MaxRetryAttempt` option from client to provider.
-	- `NewWithOptions` now sets the max retry attempts and timeouts (@cmwaters)
-  - [all] \#6077 Change spelling from British English to American (@cmwaters)
-	- Rename "Subscription.Cancelled()" to "Subscription.Canceled()" in libs/pubsub
-	- Rename "behaviour" pkg to "behavior" and internalized it in blockchain v2
-  - [rpc/client/http] \#6176 Remove `endpoint` arg from `New`, `NewWithTimeout` and `NewWithClient` (@melekes)
-  - [rpc/client/http] \#6176 Unexpose `WSEvents` (@melekes)
-  - [rpc/jsonrpc/client/ws_client] \#6176 `NewWS` no longer accepts options (use `NewWSWithOptions` and `OnReconnect` funcs to configure the client) (@melekes)
-  - [internal/libs] \#6366 Move `autofile`, `clist`,`fail`,`flowrate`, `protoio`, `sync`, `tempfile`, `test` and `timer` lib packages to an internal folder
-  - [libs/rand] \#6364 Remove most of libs/rand in favour of standard lib's `math/rand` (@liamsi)
-  - [mempool] \#6466 The original mempool reactor has been versioned as `v0` and moved to a sub-package under the root `mempool` package.
-	Some core types have been kept in the `mempool` package such as `TxCache` and it's implementations, the `Mempool` interface itself
-	and `TxInfo`. (@alexanderbez)
-  - [crypto/sr25519] \#6526 Do not re-execute the Ed25519-style key derivation step when doing signing and verification.  The derivation is now done once and only once.  This breaks `sr25519.GenPrivKeyFromSecret` output compatibility. (@Yawning)
-  - [types] \#6627 Move `NodeKey` to types to make the type public.
-  - [config] \#6627 Extend `config` to contain methods `LoadNodeKeyID` and `LoadorGenNodeKeyID`
-  - [blocksync] \#6755 Rename `FastSync` and `Blockchain` package to `BlockSync`
-	(@cmwaters)
 
 - Blockchain Protocol
 
-- Data Storage
-  - [store/state/evidence/light] \#5771 Use an order-preserving varint key encoding (@cmwaters)
-  - [mempool] \#6396 Remove mempool's write ahead log (WAL), (previously unused by the tendermint code). (@tychoish)
-  - [state] \#6541 Move pruneBlocks from consensus/state to state/execution. (@JayT106)
-
-- Tooling
-  - [tools] \#6498 Set OS home dir to instead of the hardcoded PATH. (@JayT106)
-  - [cli/indexer] \#6676 Reindex events command line tooling. (@JayT106)
-
 ### FEATURES
 
-- [config] Add `--mode` flag and config variable. See [ADR-52](https://github.com/tendermint/tendermint/blob/master/docs/architecture/adr-052-tendermint-mode.md) @dongsam
-- [rpc] \#6329 Don't cap page size in unsafe mode (@gotjoshua, @cmwaters)
-- [pex] \#6305 v2 pex reactor with backwards compatability. Introduces two new pex messages to
-  accomodate for the new p2p stack. Removes the notion of seeds and crawling. All peer
-  exchange reactors behave the same. (@cmwaters)
-- [crypto] \#6376 Enable sr25519 as a validator key
-- [mempool] \#6466 Introduction of a prioritized mempool. (@alexanderbez)
-  - `Priority` and `Sender` have been introduced into the `ResponseCheckTx` type, where the `priority` will determine the prioritization of
-  the transaction when a proposer reaps transactions for a block proposal. The `sender` field acts as an index.
-  - Operators may toggle between the legacy mempool reactor, `v0`, and the new prioritized reactor, `v1`, by setting the
-  `mempool.version` configuration, where `v1` is the default configuration.
-  - Applications that do not specify a priority, i.e. zero, will have transactions reaped by the order in which they are received by the node.
-  - Transactions are gossiped in FIFO order as they are in `v0`.
-- [config/indexer] \#6411 Introduce support for custom event indexing data sources, specifically PostgreSQL. (@JayT106)
-- [fastsync/event] \#6619 Emit fastsync status event when switching consensus/fastsync (@JayT106)
-- [statesync/event] \#6700 Emit statesync status start/end event (@JayT106)
-
 ### IMPROVEMENTS
 
-- [libs/log] Console log formatting changes as a result of \#6534 and \#6589. (@tychoish)
-- [statesync] \#6566 Allow state sync fetchers and request timeout to be configurable. (@alexanderbez)
-- [types] \#6478 Add `block_id` to `newblock` event (@jeebster)
-- [crypto/ed25519] \#5632 Adopt zip215 `ed25519` verification. (@marbar3778)
-- [crypto/ed25519] \#6526 Use [curve25519-voi](https://github.com/oasisprotocol/curve25519-voi) for `ed25519` signing and verification. (@Yawning)
-- [crypto/sr25519] \#6526 Use [curve25519-voi](https://github.com/oasisprotocol/curve25519-voi) for `sr25519` signing and verification. (@Yawning)
-- [privval] \#5603 Add `--key` to `init`, `gen_validator`, `testnet` & `unsafe_reset_priv_validator` for use in generating `secp256k1` keys.
-- [privval] \#5725 Add gRPC support to private validator.
-- [privval] \#5876 `tendermint show-validator` will query the remote signer if gRPC is being used (@marbar3778)
-- [abci/client] \#5673 `Async` requests return an error if queue is full (@melekes)
-- [mempool] \#5673 Cancel `CheckTx` requests if RPC client disconnects or times out (@melekes)
-- [abci] \#5706 Added `AbciVersion` to `RequestInfo` allowing applications to check ABCI version when connecting to Tendermint. (@marbar3778)
-- [blockchain/v1] \#5728 Remove in favor of v2 (@melekes)
-- [blockchain/v0] \#5741 Relax termination conditions and increase sync timeout (@melekes)
-- [cli] \#5772 `gen_node_key` output now contains node ID (`id` field) (@melekes)
-- [blockchain/v2] \#5774 Send status request when new peer joins (@melekes)
-- [consensus] \#5792 Deprecates the `time_iota_ms` consensus parameter, to reduce the bug surface. The parameter is no longer used. (@valardragon)
-- [store] \#5888 store.SaveBlock saves using batches instead of transactions for now to improve ACID properties. This is a quick fix for underlying issues around tm-db and ACID guarantees. (@githubsands)
-- [consensus] \#5987 Remove `time_iota_ms` from consensus params. Merge `tmproto.ConsensusParams` and `abci.ConsensusParams`. (@marbar3778)
-- [types] \#5994 Reduce the use of protobuf types in core logic. (@marbar3778)
-  - `ConsensusParams`, `BlockParams`, `ValidatorParams`, `EvidenceParams`, `VersionParams`, `sm.Version` and `version.Consensus` have become native types. They still utilize protobuf when being sent over the wire or written to disk.
-- [rpc/client/http] \#6163 Do not drop events even if the `out` channel is full (@melekes)
-- [node] \#6059 Validate and complete genesis doc before saving to state store (@silasdavis)
-- [state] \#6067 Batch save state data (@githubsands & @cmwaters)
-- [crypto] \#6120 Implement batch verification interface for ed25519 and sr25519. (@marbar3778)
-- [types] \#6120 use batch verification for verifying commits signatures.
-  - If the key type supports the batch verification API it will try to batch verify. If the verification fails we will single verify each signature.
-- [privval/file] \#6185 Return error on `LoadFilePV`, `LoadFilePVEmptyState`. Allows for better programmatic control of Tendermint.
-- [privval] \#6240 Add `context.Context` to privval interface.
-- [rpc] \#6265 set cache control in http-rpc response header (@JayT106)
-- [statesync] \#6378 Retry requests for snapshots and add a minimum discovery time (5s) for new snapshots.
-- [node/state] \#6370 graceful shutdown in the consensus reactor (@JayT106)
-- [crypto/merkle] \#6443 Improve HashAlternatives performance (@cuonglm)
-- [crypto/merkle] \#6513 Optimize HashAlternatives (@marbar3778)
-- [p2p/pex] \#6509 Improve addrBook.hash performance (@cuonglm)
-- [consensus/metrics] \#6549 Change block_size gauge to a histogram for better observability over time (@marbar3778)
-- [statesync] \#6587 Increase chunk priority and re-request chunks that don't arrive (@cmwaters)
-- [state/privval] \#6578 No GetPubKey retry beyond the proposal/voting window (@JayT106)
-- [rpc] \#6615 Add TotalGasUsed to block_results response (@crypto-facs)
-- [cmd/tendermint/commands] \#6623 replace `$HOME/.some/test/dir` with `t.TempDir` (@tanyabouman)
-
 ### BUG FIXES
 
-- [privval] \#5638 Increase read/write timeout to 5s and calculate ping interval based on it (@JoeKash)
-- [blockchain/v1] [\#5701](https://github.com/tendermint/tendermint/pull/5701) Handle peers without blocks (@melekes)
-- [blockchain/v1] \#5711 Fix deadlock (@melekes)
-- [evidence] \#6375 Fix bug with inconsistent LightClientAttackEvidence hashing (cmwaters)
-- [rpc] \#6507 Ensure RPC client can handle URLs without ports (@JayT106)
-- [statesync] \#6463 Adds Reverse Sync feature to fetch historical light blocks after state sync in order to verify any evidence (@cmwaters)
-- [fastsync] \#6590 Update the metrics during fast-sync (@JayT106)
-- [gitignore] \#6668 Fix gitignore of abci-cli (@tanyabouman)

CONTRIBUTING.md

@@ -328,7 +328,6 @@ If there were no release candidates, begin by creating a backport branch, as des
    - Bump TMVersionDefault version in  `version.go`
    - Bump P2P and block protocol versions in  `version.go`, if necessary
    - Bump ABCI protocol version in `version.go`, if necessary
-   - Add any release notes you would like to be added to the body of the release to `release_notes.md`.
 4. Open a PR with these changes against the backport branch.
 5. Once these changes are on the backport branch, push a tag with prepared release details.
    This will trigger the actual release `v0.35.0`.
@@ -355,7 +354,6 @@ To create a minor release:
    - Bump the ABCI version number, if necessary.
      (Note that ABCI follows semver, and that ABCI versions are the only versions
      which can change during minor releases, and only field additions are valid minor changes.)
-   - Add any release notes you would like to be added to the body of the release to `release_notes.md`.
 4. Open a PR with these changes that will land them back on `v0.35.x`
 5. Once this change has landed on the backport branch, make sure to pull it locally, then push a tag.
    - `git tag -a v0.35.1 -m 'Release v0.35.1'`

README.md

@@ -82,32 +82,12 @@ and familiarize yourself with our
 Tendermint uses [Semantic Versioning](http://semver.org/) to determine when and how the version changes.
 According to SemVer, anything in the public API can change at any time before version 1.0.0
 
-To provide some stability to Tendermint users in these 0.X.X days, the MINOR version is used
-to signal breaking changes across a subset of the total public API. This subset includes all
-interfaces exposed to other processes (cli, rpc, p2p, etc.), but does not
-include the Go APIs.
+To provide some stability to users of 0.X.X versions of Tendermint, the MINOR version is used
+to signal breaking changes across Tendermint's API. This API includes all
+publicly exposed types, functions, and methods in non-internal Go packages as well as
+the types and methods accessible via the Tendermint RPC interface.
 
-That said, breaking changes in the following packages will be documented in the
-CHANGELOG even if they don't lead to MINOR version bumps:
-
-- crypto
-- config
-- libs
-    - bits
-    - bytes
-    - json
-    - log
-    - math
-    - net
-    - os
-    - protoio
-    - rand
-    - sync
-    - strings
-    - service
-- node
-- rpc/client
-- types
+Breaking changes to these public APIs will be documented in the CHANGELOG.
 
 ### Upgrades
 

UPGRADING.md

@@ -2,7 +2,7 @@
 
 This guide provides instructions for upgrading to specific versions of Tendermint Core.
 
-## Unreleased
+## v0.35
 
 ### ABCI Changes
 
@@ -17,7 +17,16 @@ This guide provides instructions for upgrading to specific versions of Tendermin
 
 ### Config Changes
 
-* `fast_sync = "v1"` and `fast_sync = "v2"` are no longer supported. Please use `v0` instead.
+* The configuration file field `[fastsync]` has been renamed to `[blocksync]`.
+
+* The top level configuration file field `fast-sync` has moved under the new `[blocksync]`
+  field as `blocksync.enable`.
+
+* `blocksync.version = "v1"` and `blocksync.version = "v2"` (previously `fastsync`)
+  are no longer supported. Please use `v0` instead. During the v0.35 release cycle, `v0` was
+  determined to suit the existing needs and the cost of maintaining the `v1` and `v2` modules
+  was determined to be greater than necessary.
+
 
 * All config parameters are now hyphen-case (also known as kebab-case) instead of snake_case. Before restarting the node make sure
   you have updated all the variables in your `config.toml` file.
@@ -35,7 +44,7 @@ This guide provides instructions for upgrading to specific versions of Tendermin
 * The fast sync process as well as the blockchain package and service has all
   been renamed to block sync
 
-### Key Format Changes
+### Database Key Format Changes
 
 The format of all tendermint on-disk database keys changes in
 0.35. Upgrading nodes must either re-sync all data or run a migration
@@ -60,6 +69,8 @@ if needed.
 
 * You must now specify the node mode (validator|full|seed) in `tendermint init [mode]`
 
+* The `--fast-sync` command line option has been renamed to `--blocksync.enable`
+
 * If you had previously used `tendermint gen_node_key` to generate a new node
   key, keep in mind that it no longer saves the output to a file. You can use
   `tendermint init validator` or pipe the output of `tendermint gen_node_key` to
@@ -74,8 +85,8 @@ if needed.
 
 ### API Changes
 
-The p2p layer was reimplemented as part of the 0.35 release cycle, and
-all reactors were refactored. As part of that work these
+The p2p layer was reimplemented as part of the 0.35 release cycle and
+all reactors were refactored to accomodate the change. As part of that work these
 implementations moved into the `internal` package and are no longer
 considered part of the public Go API of tendermint. These packages
 are:
@@ -98,13 +109,11 @@ will need to change to accommodate these changes. Most notably:
   longer exported and have been replaced with `node.New` and
   `node.NewDefault` which provide more functional interfaces.
 
-### RPC changes
-
-#### gRPC Support
+### gRPC Support
 
 Mark gRPC in the RPC layer as deprecated and to be removed in 0.36.
 
-#### Peer Management Interface
+### Peer Management Interface
 
 When running with the new P2P Layer, the methods `UnsafeDialSeeds` and
 `UnsafeDialPeers` RPC methods will always return an error. They are
@@ -116,6 +125,58 @@ method changes in this release to accommodate the different way that
 the new stack tracks data about peers. This change affects users of
 both stacks.
 
+### Using the updated p2p library
+
+The P2P library was reimplemented in this release. The new implementation is
+enabled by default in this version of Tendermint. The legacy implementation is still
+included in this version of Tendermint as a backstop to work around unforeseen
+production issues. The new and legacy version are interoperable. If necessary, 
+you can enable the legacy implementation in the server configuration file.
+
+To make use of the legacy P2P implemementation add or update the following field of 
+your server's configuration file under the `[p2p]` section:
+
+```toml
+[p2p]
+...
+use-legacy = true
+...
+```
+
+If you need to do this, please consider filing an issue in the Tendermint repository
+to let us know why. We plan to remove the legacy P2P code in the next (v0.36) release.
+
+#### New p2p queue types
+
+The new p2p implementation enables selection of the queue type to be used for
+passing messages between peers.
+
+The following values may be used when selecting which queue type to use:
+
+* `fifo`: (**default**) An unbuffered and lossless queue that passes messages through
+in the order in which they were received.
+
+* `priority`: A priority queue of messages.
+
+* `wdrr`: A queue implementing the Weighted Deficit Round Robin algorithm. A 
+weighted deficit round robin queue is created per peer. Each queue contains a 
+separate 'flow' for each of the channels of communication that exist between any two
+peers. Tendermint maintains a channel per message type between peers. Each WDRR
+queue maintains a shared buffered with a fixed capacity through which messages on different
+flows are passed.
+For more information on WDRR scheduling, see: https://en.wikipedia.org/wiki/Deficit_round_robin
+
+To select a queue type, add or update the following field under the `[p2p]`
+section of your server's configuration file.
+
+```toml
+[p2p]
+...
+queue-type = wdrr
+...
+```
+
+
 ### Support for Custom Reactor and Mempool Implementations
 
 The changes to p2p layer removed existing support for custom

cmd/tendermint/commands/gen_node_key.go

@@ -12,11 +12,9 @@ import (
 // GenNodeKeyCmd allows the generation of a node key. It prints JSON-encoded
 // NodeKey to the standard output.
 var GenNodeKeyCmd = &cobra.Command{
-	Use:     "gen-node-key",
-	Aliases: []string{"gen_node_key"},
-	Short:   "Generate a new node key",
-	RunE:    genNodeKey,
-	PreRun:  deprecateSnakeCase,
+	Use:   "gen-node-key",
+	Short: "Generate a new node key",
+	RunE:  genNodeKey,
 }
 
 func genNodeKey(cmd *cobra.Command, args []string) error {

cmd/tendermint/commands/gen_validator.go

@@ -13,11 +13,9 @@ import (
 // GenValidatorCmd allows the generation of a keypair for a
 // validator.
 var GenValidatorCmd = &cobra.Command{
-	Use:     "gen-validator",
-	Aliases: []string{"gen_validator"},
-	Short:   "Generate new validator keypair",
-	RunE:    genValidator,
-	PreRun:  deprecateSnakeCase,
+	Use:   "gen-validator",
+	Short: "Generate new validator keypair",
+	RunE:  genValidator,
 }
 
 func init() {

cmd/tendermint/commands/inspect.go

@@ -0,0 +1,87 @@
+package commands
+
+import (
+	"context"
+	"os"
+	"os/signal"
+	"syscall"
+
+	"github.com/spf13/cobra"
+
+	cfg "github.com/tendermint/tendermint/config"
+	"github.com/tendermint/tendermint/inspect"
+	"github.com/tendermint/tendermint/state"
+	"github.com/tendermint/tendermint/state/indexer/sink"
+	"github.com/tendermint/tendermint/store"
+	"github.com/tendermint/tendermint/types"
+)
+
+// InspectCmd is the command for starting an inspect server.
+var InspectCmd = &cobra.Command{
+	Use:   "inspect",
+	Short: "Run an inspect server for investigating Tendermint state",
+	Long: `
+	inspect runs a subset of Tendermint's RPC endpoints that are useful for debugging
+	issues with Tendermint.
+
+	When the Tendermint consensus engine detects inconsistent state, it will crash the
+	tendermint process. Tendermint will not start up while in this inconsistent state. 
+	The inspect command can be used to query the block and state store using Tendermint
+	RPC calls to debug issues of inconsistent state.
+	`,
+
+	RunE: runInspect,
+}
+
+func init() {
+	InspectCmd.Flags().
+		String("rpc.laddr",
+			config.RPC.ListenAddress, "RPC listenener address. Port required")
+	InspectCmd.Flags().
+		String("db-backend",
+			config.DBBackend, "database backend: goleveldb | cleveldb | boltdb | rocksdb | badgerdb")
+	InspectCmd.Flags().
+		String("db-dir", config.DBPath, "database directory")
+}
+
+func runInspect(cmd *cobra.Command, args []string) error {
+	ctx, cancel := context.WithCancel(cmd.Context())
+	defer cancel()
+
+	c := make(chan os.Signal, 1)
+	signal.Notify(c, syscall.SIGTERM, syscall.SIGINT)
+	go func() {
+		<-c
+		cancel()
+	}()
+
+	blockStoreDB, err := cfg.DefaultDBProvider(&cfg.DBContext{ID: "blockstore", Config: config})
+	if err != nil {
+		return err
+	}
+	blockStore := store.NewBlockStore(blockStoreDB)
+	stateDB, err := cfg.DefaultDBProvider(&cfg.DBContext{ID: "state", Config: config})
+	if err != nil {
+		if err := blockStoreDB.Close(); err != nil {
+			logger.Error("error closing block store db", "error", err)
+		}
+		return err
+	}
+	genDoc, err := types.GenesisDocFromFile(config.GenesisFile())
+	if err != nil {
+		return err
+	}
+	sinks, err := sink.EventSinksFromConfig(config, cfg.DefaultDBProvider, genDoc.ChainID)
+	if err != nil {
+		return err
+	}
+	stateStore := state.NewStore(stateDB)
+
+	ins := inspect.New(config.RPC, blockStore, stateStore, sinks, logger)
+
+	logger.Info("starting inspect server")
+	if err := ins.Run(ctx); err != nil {
+		return err
+	}
+	return nil
+}

cmd/tendermint/commands/probe_upnp.go

@@ -11,11 +11,9 @@ import (
 
 // ProbeUpnpCmd adds capabilities to test the UPnP functionality.
 var ProbeUpnpCmd = &cobra.Command{
-	Use:     "probe-upnp",
-	Aliases: []string{"probe_upnp"},
-	Short:   "Test UPnP functionality",
-	RunE:    probeUpnp,
-	PreRun:  deprecateSnakeCase,
+	Use:   "probe-upnp",
+	Short: "Test UPnP functionality",
+	RunE:  probeUpnp,
 }
 
 func probeUpnp(cmd *cobra.Command, args []string) error {

cmd/tendermint/commands/reindex_event.go

@@ -31,7 +31,7 @@ var ReIndexEventCmd = &cobra.Command{
 	Long: `
 	reindex-event is an offline tooling to re-index block and tx events to the eventsinks,
 	you can run this command when the event store backend dropped/disconnected or you want to replace the backend.
-	The default start-height is 0, meaning the tooling will start reindex from the base block height(inclusive); and the 
+	The default start-height is 0, meaning the tooling will start reindex from the base block height(inclusive); and the
 	default end-height is 0, meaning the tooling will reindex until the latest block height(inclusive). User can omits
 	either or both arguments.
 	`,
@@ -106,7 +106,7 @@ func loadEventSinks(cfg *tmcfg.Config) ([]indexer.EventSink, error) {
 			if conn == "" {
 				return nil, errors.New("the psql connection settings cannot be empty")
 			}
-			es, _, err := psql.NewEventSink(conn, chainID)
+			es, err := psql.NewEventSink(conn, chainID)
 			if err != nil {
 				return nil, err
 			}

cmd/tendermint/commands/replay.go

@@ -17,11 +17,9 @@ var ReplayCmd = &cobra.Command{
 // ReplayConsoleCmd allows replaying of messages from the WAL in a
 // console.
 var ReplayConsoleCmd = &cobra.Command{
-	Use:     "replay-console",
-	Aliases: []string{"replay_console"},
-	Short:   "Replay messages from WAL in a console",
+	Use:   "replay-console",
+	Short: "Replay messages from WAL in a console",
 	Run: func(cmd *cobra.Command, args []string) {
 		consensus.RunReplayFile(config.BaseConfig, config.Consensus, true)
 	},
-	PreRun: deprecateSnakeCase,
 }

cmd/tendermint/commands/reset_priv_validator.go

@@ -14,11 +14,9 @@ import (
 // ResetAllCmd removes the database of this Tendermint core
 // instance.
 var ResetAllCmd = &cobra.Command{
-	Use:     "unsafe-reset-all",
-	Aliases: []string{"unsafe_reset_all"},
-	Short:   "(unsafe) Remove all the data and WAL, reset this node's validator to genesis state",
-	RunE:    resetAll,
-	PreRun:  deprecateSnakeCase,
+	Use:   "unsafe-reset-all",
+	Short: "(unsafe) Remove all the data and WAL, reset this node's validator to genesis state",
+	RunE:  resetAll,
 }
 
 var keepAddrBook bool
@@ -31,11 +29,9 @@ func init() {
 
 // ResetPrivValidatorCmd resets the private validator files.
 var ResetPrivValidatorCmd = &cobra.Command{
-	Use:     "unsafe-reset-priv-validator",
-	Aliases: []string{"unsafe_reset_priv_validator"},
-	Short:   "(unsafe) Reset this node's validator to genesis state",
-	RunE:    resetPrivValidator,
-	PreRun:  deprecateSnakeCase,
+	Use:   "unsafe-reset-priv-validator",
+	Short: "(unsafe) Reset this node's validator to genesis state",
+	RunE:  resetPrivValidator,
 }
 
 // XXX: this is totally unsafe.

cmd/tendermint/commands/root.go

@@ -2,7 +2,6 @@ package commands
 
 import (
 	"fmt"
-	"strings"
 	"time"
 
 	"github.com/spf13/cobra"
@@ -65,10 +64,3 @@ var RootCmd = &cobra.Command{
 		return nil
 	},
 }
-
-// deprecateSnakeCase is a util function for 0.34.1. Should be removed in 0.35
-func deprecateSnakeCase(cmd *cobra.Command, args []string) {
-	if strings.Contains(cmd.CalledAs(), "_") {
-		fmt.Println("Deprecated: snake_case commands will be replaced by hyphen-case commands in the next major release")
-	}
-}

cmd/tendermint/commands/run_node.go

@@ -3,6 +3,8 @@ package commands
 import (
 	"bytes"
 	"crypto/sha256"
+	"errors"
+	"flag"
 	"fmt"
 	"io"
 	"os"
@@ -33,7 +35,22 @@ func AddNodeFlags(cmd *cobra.Command) {
 		"socket address to listen on for connections from external priv-validator process")
 
 	// node flags
-	cmd.Flags().Bool("fast-sync", config.FastSyncMode, "fast blockchain syncing")
+	cmd.Flags().Bool("blocksync.enable", config.BlockSync.Enable, "enable fast blockchain syncing")
+
+	// TODO (https://github.com/tendermint/tendermint/issues/6908): remove this check after the v0.35 release cycle
+	// This check was added to give users an upgrade prompt to use the new flag for syncing.
+	//
+	// The pflag package does not have a native way to print a depcrecation warning
+	// and return an error. This logic was added to print a deprecation message to the user
+	// and then crash if the user attempts to use the old --fast-sync flag.
+	fs := flag.NewFlagSet("", flag.ExitOnError)
+	fs.Func("fast-sync", "deprecated",
+		func(string) error {
+			return errors.New("--fast-sync has been deprecated, please use --blocksync.enable")
+		})
+	cmd.Flags().AddGoFlagSet(fs)
+
+	cmd.Flags().MarkHidden("fast-sync") //nolint:errcheck
 	cmd.Flags().BytesHexVar(
 		&genesisHash,
 		"genesis-hash",
@@ -158,7 +175,7 @@ func checkGenesisHash(config *cfg.Config) error {
 	// Compare with the flag.
 	if !bytes.Equal(genesisHash, actualHash) {
 		return fmt.Errorf(
-			"--genesis_hash=%X does not match %s hash: %X",
+			"--genesis-hash=%X does not match %s hash: %X",
 			genesisHash, config.GenesisFile(), actualHash)
 	}
 

cmd/tendermint/commands/show_node_id.go

@@ -8,11 +8,9 @@ import (
 
 // ShowNodeIDCmd dumps node's ID to the standard output.
 var ShowNodeIDCmd = &cobra.Command{
-	Use:     "show-node-id",
-	Aliases: []string{"show_node_id"},
-	Short:   "Show this node's ID",
-	RunE:    showNodeID,
-	PreRun:  deprecateSnakeCase,
+	Use:   "show-node-id",
+	Short: "Show this node's ID",
+	RunE:  showNodeID,
 }
 
 func showNodeID(cmd *cobra.Command, args []string) error {

cmd/tendermint/commands/show_validator.go

@@ -16,11 +16,9 @@ import (
 
 // ShowValidatorCmd adds capabilities for showing the validator info.
 var ShowValidatorCmd = &cobra.Command{
-	Use:     "show-validator",
-	Aliases: []string{"show_validator"},
-	Short:   "Show this node's validator info",
-	RunE:    showValidator,
-	PreRun:  deprecateSnakeCase,
+	Use:   "show-validator",
+	Short: "Show this node's validator info",
+	RunE:  showValidator,
 }
 
 func showValidator(cmd *cobra.Command, args []string) error {

cmd/tendermint/main.go

@@ -28,6 +28,7 @@ func main() {
 		cmd.ShowNodeIDCmd,
 		cmd.GenNodeKeyCmd,
 		cmd.VersionCmd,
+		cmd.InspectCmd,
 		cmd.MakeKeyMigrateCommand(),
 		debug.DebugCmd,
 		cli.NewCompletionCmd(rootCmd, true),

config/config.go

@@ -76,7 +76,7 @@ type Config struct {
 	P2P             *P2PConfig             `mapstructure:"p2p"`
 	Mempool         *MempoolConfig         `mapstructure:"mempool"`
 	StateSync       *StateSyncConfig       `mapstructure:"statesync"`
-	BlockSync       *BlockSyncConfig       `mapstructure:"fastsync"`
+	BlockSync       *BlockSyncConfig       `mapstructure:"blocksync"`
 	Consensus       *ConsensusConfig       `mapstructure:"consensus"`
 	TxIndex         *TxIndexConfig         `mapstructure:"tx-index"`
 	Instrumentation *InstrumentationConfig `mapstructure:"instrumentation"`
@@ -152,7 +152,7 @@ func (cfg *Config) ValidateBasic() error {
 		return fmt.Errorf("error in [statesync] section: %w", err)
 	}
 	if err := cfg.BlockSync.ValidateBasic(); err != nil {
-		return fmt.Errorf("error in [fastsync] section: %w", err)
+		return fmt.Errorf("error in [blocksync] section: %w", err)
 	}
 	if err := cfg.Consensus.ValidateBasic(); err != nil {
 		return fmt.Errorf("error in [consensus] section: %w", err)
@@ -194,12 +194,6 @@ type BaseConfig struct { //nolint: maligned
 	//   - No priv_validator_key.json, priv_validator_state.json
 	Mode string `mapstructure:"mode"`
 
-	// If this node is many blocks behind the tip of the chain, FastSync
-	// allows them to catchup quickly by downloading blocks in parallel
-	// and verifying their commits
-	// TODO: This should be moved to the blocksync config
-	FastSyncMode bool `mapstructure:"fast-sync"`
-
 	// Database backend: goleveldb | cleveldb | boltdb | rocksdb
 	// * goleveldb (github.com/syndtr/goleveldb - most popular implementation)
 	//   - pure go
@@ -242,23 +236,24 @@ type BaseConfig struct { //nolint: maligned
 	// If true, query the ABCI app on connecting to a new peer
 	// so the app can decide if we should keep the connection or not
 	FilterPeers bool `mapstructure:"filter-peers"` // false
+
+	Other map[string]interface{} `mapstructure:",remain"`
 }
 
 // DefaultBaseConfig returns a default base configuration for a Tendermint node
 func DefaultBaseConfig() BaseConfig {
 	return BaseConfig{
-		Genesis:      defaultGenesisJSONPath,
-		NodeKey:      defaultNodeKeyPath,
-		Mode:         defaultMode,
-		Moniker:      defaultMoniker,
-		ProxyApp:     "tcp://127.0.0.1:26658",
-		ABCI:         "socket",
-		LogLevel:     DefaultLogLevel,
-		LogFormat:    log.LogFormatPlain,
-		FastSyncMode: true,
-		FilterPeers:  false,
-		DBBackend:    "goleveldb",
-		DBPath:       "data",
+		Genesis:     defaultGenesisJSONPath,
+		NodeKey:     defaultNodeKeyPath,
+		Mode:        defaultMode,
+		Moniker:     defaultMoniker,
+		ProxyApp:    "tcp://127.0.0.1:26658",
+		ABCI:        "socket",
+		LogLevel:    DefaultLogLevel,
+		LogFormat:   log.LogFormatPlain,
+		FilterPeers: false,
+		DBBackend:   "goleveldb",
+		DBPath:      "data",
 	}
 }
 
@@ -268,7 +263,6 @@ func TestBaseConfig() BaseConfig {
 	cfg.chainID = "tendermint_test"
 	cfg.Mode = ModeValidator
 	cfg.ProxyApp = "kvstore"
-	cfg.FastSyncMode = false
 	cfg.DBBackend = "memdb"
 	return cfg
 }
@@ -345,6 +339,28 @@ func (cfg BaseConfig) ValidateBasic() error {
 		return fmt.Errorf("unknown mode: %v", cfg.Mode)
 	}
 
+	// TODO (https://github.com/tendermint/tendermint/issues/6908) remove this check after the v0.35 release cycle.
+	// This check was added to give users an upgrade prompt to use the new
+	// configuration option in v0.35. In future release cycles they should no longer
+	// be using this configuration parameter so the check can be removed.
+	// The cfg.Other field can likely be removed at the same time if it is not referenced
+	// elsewhere as it was added to service this check.
+	if fs, ok := cfg.Other["fastsync"]; ok {
+		if _, ok := fs.(map[string]interface{}); ok {
+			return fmt.Errorf("a configuration section named 'fastsync' was found in the " +
+				"configuration file. The 'fastsync' section has been renamed to " +
+				"'blocksync', please update the 'fastsync' field in your configuration file to 'blocksync'")
+		}
+	}
+	if fs, ok := cfg.Other["fast-sync"]; ok {
+		if fs != "" {
+			return fmt.Errorf("a parameter named 'fast-sync' was found in the " +
+				"configuration file. The parameter to enable or disable quickly syncing with a blockchain" +
+				"has moved to the [blocksync] section of the configuration file as blocksync.enable. " +
+				"Please move the 'fast-sync' field in your configuration file to 'blocksync.enable'")
+		}
+	}
+
 	return nil
 }
 
@@ -694,13 +710,14 @@ type P2PConfig struct { //nolint: maligned
 	// Force dial to fail
 	TestDialFail bool `mapstructure:"test-dial-fail"`
 
-	// DisableLegacy is used mostly for testing to enable or disable the legacy
-	// P2P stack.
-	DisableLegacy bool `mapstructure:"disable-legacy"`
+	// UseLegacy enables the "legacy" P2P implementation and
+	// disables the newer default implementation. This flag will
+	// be removed in a future release.
+	UseLegacy bool `mapstructure:"use-legacy"`
 
 	// Makes it possible to configure which queue backend the p2p
 	// layer uses. Options are: "fifo", "priority" and "wdrr",
-	// with the default being "fifo".
+	// with the default being "priority".
 	QueueType string `mapstructure:"queue-type"`
 }
 
@@ -732,6 +749,7 @@ func DefaultP2PConfig() *P2PConfig {
 		DialTimeout:             3 * time.Second,
 		TestDialFail:            false,
 		QueueType:               "priority",
+		UseLegacy:               false,
 	}
 }
 
@@ -882,15 +900,46 @@ func (cfg *MempoolConfig) ValidateBasic() error {
 
 // StateSyncConfig defines the configuration for the Tendermint state sync service
 type StateSyncConfig struct {
-	Enable              bool          `mapstructure:"enable"`
-	TempDir             string        `mapstructure:"temp-dir"`
-	RPCServers          []string      `mapstructure:"rpc-servers"`
-	TrustPeriod         time.Duration `mapstructure:"trust-period"`
-	TrustHeight         int64         `mapstructure:"trust-height"`
-	TrustHash           string        `mapstructure:"trust-hash"`
-	DiscoveryTime       time.Duration `mapstructure:"discovery-time"`
+	// State sync rapidly bootstraps a new node by discovering, fetching, and restoring a
+	// state machine snapshot from peers instead of fetching and replaying historical
+	// blocks. Requires some peers in the network to take and serve state machine
+	// snapshots. State sync is not attempted if the node has any local state
+	// (LastBlockHeight > 0). The node will have a truncated block history, starting from
+	// the height of the snapshot.
+	Enable bool `mapstructure:"enable"`
+
+	// State sync uses light client verification to verify state. This can be done either
+	// through the P2P layer or the RPC layer. Set this to true to use the P2P layer. If
+	// false (default), the RPC layer will be used.
+	UseP2P bool `mapstructure:"use-p2p"`
+
+	// If using RPC, at least two addresses need to be provided. They should be compatible
+	// with net.Dial, for example: "host.example.com:2125".
+	RPCServers []string `mapstructure:"rpc-servers"`
+
+	// The hash and height of a trusted block. Must be within the trust-period.
+	TrustHeight int64  `mapstructure:"trust-height"`
+	TrustHash   string `mapstructure:"trust-hash"`
+
+	// The trust period should be set so that Tendermint can detect and gossip
+	// misbehavior before it is considered expired. For chains based on the Cosmos SDK,
+	// one day less than the unbonding period should suffice.
+	TrustPeriod time.Duration `mapstructure:"trust-period"`
+
+	// Time to spend discovering snapshots before initiating a restore.
+	DiscoveryTime time.Duration `mapstructure:"discovery-time"`
+
+	// Temporary directory for state sync snapshot chunks, defaults to os.TempDir().
+	// The synchronizer will create a new, randomly named directory within this directory
+	// and remove it when the sync is complete.
+	TempDir string `mapstructure:"temp-dir"`
+
+	// The timeout duration before re-requesting a chunk, possibly from a different
+	// peer (default: 15 seconds).
 	ChunkRequestTimeout time.Duration `mapstructure:"chunk-request-timeout"`
-	Fetchers            int32         `mapstructure:"fetchers"`
+
+	// The number of concurrent chunk and block fetchers to run (default: 4).
+	Fetchers int32 `mapstructure:"fetchers"`
 }
 
 func (cfg *StateSyncConfig) TrustHashBytes() []byte {
@@ -919,49 +968,51 @@ func TestStateSyncConfig() *StateSyncConfig {
 
 // ValidateBasic performs basic validation.
 func (cfg *StateSyncConfig) ValidateBasic() error {
-	if cfg.Enable {
-		if len(cfg.RPCServers) == 0 {
-			return errors.New("rpc-servers is required")
-		}
+	if !cfg.Enable {
+		return nil
+	}
 
+	// If we're not using the P2P stack then we need to validate the
+	// RPCServers
+	if !cfg.UseP2P {
 		if len(cfg.RPCServers) < 2 {
-			return errors.New("at least two rpc-servers entries is required")
+			return errors.New("at least two rpc-servers must be specified")
 		}
 
 		for _, server := range cfg.RPCServers {
-			if len(server) == 0 {
+			if server == "" {
 				return errors.New("found empty rpc-servers entry")
 			}
 		}
+	}
 
-		if cfg.DiscoveryTime != 0 && cfg.DiscoveryTime < 5*time.Second {
-			return errors.New("discovery time must be 0s or greater than five seconds")
-		}
+	if cfg.DiscoveryTime != 0 && cfg.DiscoveryTime < 5*time.Second {
+		return errors.New("discovery time must be 0s or greater than five seconds")
+	}
 
-		if cfg.TrustPeriod <= 0 {
-			return errors.New("trusted-period is required")
-		}
+	if cfg.TrustPeriod <= 0 {
+		return errors.New("trusted-period is required")
+	}
 
-		if cfg.TrustHeight <= 0 {
-			return errors.New("trusted-height is required")
-		}
+	if cfg.TrustHeight <= 0 {
+		return errors.New("trusted-height is required")
+	}
 
-		if len(cfg.TrustHash) == 0 {
-			return errors.New("trusted-hash is required")
-		}
+	if len(cfg.TrustHash) == 0 {
+		return errors.New("trusted-hash is required")
+	}
 
-		_, err := hex.DecodeString(cfg.TrustHash)
-		if err != nil {
-			return fmt.Errorf("invalid trusted-hash: %w", err)
-		}
+	_, err := hex.DecodeString(cfg.TrustHash)
+	if err != nil {
+		return fmt.Errorf("invalid trusted-hash: %w", err)
+	}
 
-		if cfg.ChunkRequestTimeout < 5*time.Second {
-			return errors.New("chunk-request-timeout must be at least 5 seconds")
-		}
+	if cfg.ChunkRequestTimeout < 5*time.Second {
+		return errors.New("chunk-request-timeout must be at least 5 seconds")
+	}
 
-		if cfg.Fetchers <= 0 {
-			return errors.New("fetchers is required")
-		}
+	if cfg.Fetchers <= 0 {
+		return errors.New("fetchers is required")
 	}
 
 	return nil
@@ -970,13 +1021,18 @@ func (cfg *StateSyncConfig) ValidateBasic() error {
 //-----------------------------------------------------------------------------
 
 // BlockSyncConfig (formerly known as FastSync) defines the configuration for the Tendermint block sync service
+// If this node is many blocks behind the tip of the chain, BlockSync
+// allows them to catchup quickly by downloading blocks in parallel
+// and verifying their commits.
 type BlockSyncConfig struct {
+	Enable  bool   `mapstructure:"enable"`
 	Version string `mapstructure:"version"`
 }
 
 // DefaultBlockSyncConfig returns a default configuration for the block sync service
 func DefaultBlockSyncConfig() *BlockSyncConfig {
 	return &BlockSyncConfig{
+		Enable:  true,
 		Version: BlockSyncV0,
 	}
 }

config/toml.go

@@ -97,11 +97,6 @@ moniker = "{{ .BaseConfig.Moniker }}"
 #   - No priv_validator_key.json, priv_validator_state.json
 mode = "{{ .BaseConfig.Mode }}"
 
-# If this node is many blocks behind the tip of the chain, FastSync
-# allows them to catchup quickly by downloading blocks in parallel
-# and verifying their commits
-fast-sync = {{ .BaseConfig.FastSyncMode }}
-
 # Database backend: goleveldb | cleveldb | boltdb | rocksdb | badgerdb
 # * goleveldb (github.com/syndtr/goleveldb - most popular implementation)
 #   - pure go
@@ -270,8 +265,8 @@ pprof-laddr = "{{ .RPC.PprofListenAddress }}"
 #######################################################
 [p2p]
 
-# Enable the new p2p layer.
-disable-legacy = {{ .P2P.DisableLegacy }}
+# Enable the legacy p2p layer.
+use-legacy = {{ .P2P.UseLegacy }}
 
 # Select the p2p internal queue
 queue-type = "{{ .P2P.QueueType }}"
@@ -305,6 +300,7 @@ persistent-peers = "{{ .P2P.PersistentPeers }}"
 upnp = {{ .P2P.UPNP }}
 
 # Path to address book
+# TODO: Remove once p2p refactor is complete in favor of peer store.
 addr-book-file = "{{ js .P2P.AddrBook }}"
 
 # Set true for strict address routability rules
@@ -330,6 +326,8 @@ max-connections = {{ .P2P.MaxConnections }}
 max-incoming-connection-attempts = {{ .P2P.MaxIncomingConnectionAttempts }}
 
 # List of node IDs, to which a connection will be (re)established ignoring any existing limits
+# TODO: Remove once p2p refactor is complete.
+# ref: https://github.com/tendermint/tendermint/issues/5670
 unconditional-peer-ids = "{{ .P2P.UnconditionalPeerIDs }}"
 
 # Maximum pause when redialing a persistent peer (if zero, exponential backoff is used)
@@ -426,22 +424,30 @@ ttl-num-blocks = {{ .Mempool.TTLNumBlocks }}
 # starting from the height of the snapshot.
 enable = {{ .StateSync.Enable }}
 
-# RPC servers (comma-separated) for light client verification of the synced state machine and
-# retrieval of state data for node bootstrapping. Also needs a trusted height and corresponding
-# header hash obtained from a trusted source, and a period during which validators can be trusted.
-#
-# For Cosmos SDK-based chains, trust-period should usually be about 2/3 of the unbonding time (~2
-# weeks) during which they can be financially punished (slashed) for misbehavior.
+# State sync uses light client verification to verify state. This can be done either through the
+# P2P layer or RPC layer. Set this to true to use the P2P layer. If false (default), RPC layer
+# will be used.
+use-p2p = {{ .StateSync.UseP2P }}
+
+# If using RPC, at least two addresses need to be provided. They should be compatible with net.Dial,
+# for example: "host.example.com:2125"
 rpc-servers = "{{ StringsJoin .StateSync.RPCServers "," }}"
+
+# The hash and height of a trusted block. Must be within the trust-period.
 trust-height = {{ .StateSync.TrustHeight }}
 trust-hash = "{{ .StateSync.TrustHash }}"
+
+# The trust period should be set so that Tendermint can detect and gossip misbehavior before 
+# it is considered expired. For chains based on the Cosmos SDK, one day less than the unbonding 
+# period should suffice.
 trust-period = "{{ .StateSync.TrustPeriod }}"
 
 # Time to spend discovering snapshots before initiating a restore.
 discovery-time = "{{ .StateSync.DiscoveryTime }}"
 
-# Temporary directory for state sync snapshot chunks, defaults to the OS tempdir (typically /tmp).
-# Will create a new, randomly named directory within, and remove it when done.
+# Temporary directory for state sync snapshot chunks, defaults to os.TempDir().
+# The synchronizer will create a new, randomly named directory within this directory
+# and remove it when the sync is complete.
 temp-dir = "{{ .StateSync.TempDir }}"
 
 # The timeout duration before re-requesting a chunk, possibly from a different
@@ -454,10 +460,15 @@ fetchers = "{{ .StateSync.Fetchers }}"
 #######################################################
 ###       Block Sync Configuration Connections       ###
 #######################################################
-[fastsync]
+[blocksync]
+
+# If this node is many blocks behind the tip of the chain, BlockSync
+# allows them to catchup quickly by downloading blocks in parallel
+# and verifying their commits
+enable = {{ .BlockSync.Enable }}
 
 # Block Sync version to use:
-#   1) "v0" (default) - the legacy block sync implementation
+#   1) "v0" (default) - the standard Block Sync implementation
 #   2) "v2" - DEPRECATED, please use v0
 version = "{{ .BlockSync.Version }}"
 

config/toml_test.go

@@ -36,9 +36,7 @@ func TestEnsureRoot(t *testing.T) {
 	data, err := ioutil.ReadFile(filepath.Join(tmpDir, defaultConfigFilePath))
 	require.Nil(err)
 
-	if !checkConfig(string(data)) {
-		t.Fatalf("config file missing some information")
-	}
+	checkConfig(t, string(data))
 
 	ensureFiles(t, tmpDir, "data")
 }
@@ -57,9 +55,7 @@ func TestEnsureTestRoot(t *testing.T) {
 	data, err := ioutil.ReadFile(filepath.Join(rootDir, defaultConfigFilePath))
 	require.Nil(err)
 
-	if !checkConfig(string(data)) {
-		t.Fatalf("config file missing some information")
-	}
+	checkConfig(t, string(data))
 
 	// TODO: make sure the cfg returned and testconfig are the same!
 	baseConfig := DefaultBaseConfig()
@@ -67,16 +63,15 @@ func TestEnsureTestRoot(t *testing.T) {
 	ensureFiles(t, rootDir, defaultDataDir, baseConfig.Genesis, pvConfig.Key, pvConfig.State)
 }
 
-func checkConfig(configFile string) bool {
-	var valid bool
-
+func checkConfig(t *testing.T, configFile string) {
+	t.Helper()
 	// list of words we expect in the config
 	var elems = []string{
 		"moniker",
 		"seeds",
 		"proxy-app",
-		"fast_sync",
-		"create_empty_blocks",
+		"blocksync",
+		"create-empty-blocks",
 		"peer",
 		"timeout",
 		"broadcast",
@@ -89,10 +84,7 @@ func checkConfig(configFile string) bool {
 	}
 	for _, e := range elems {
 		if !strings.Contains(configFile, e) {
-			valid = false
-		} else {
-			valid = true
+			t.Errorf("config file was expected to contain %s but did not", e)
 		}
 	}
-	return valid
 }

crypto/secp256k1/secp256k1_test.go

@@ -36,8 +36,7 @@ func TestPubKeySecp256k1Address(t *testing.T) {
 		addrBbz, _, _ := base58.CheckDecode(d.addr)
 		addrB := crypto.Address(addrBbz)
 
-		var priv secp256k1.PrivKey = secp256k1.PrivKey(privB)
-
+		priv := secp256k1.PrivKey(privB)
 		pubKey := priv.PubKey()
 		pubT, _ := pubKey.(secp256k1.PubKey)
 		pub := pubT

docs/app-dev/indexing-transactions.md

@@ -62,7 +62,7 @@ be turned off regardless of other values provided.
 #### KV
 
 The `kv` indexer type is an embedded key-value store supported by the main
-underling Tendermint database. Using the `kv` indexer type allows you to query
+underlying Tendermint database. Using the `kv` indexer type allows you to query
 for block and transaction events directly against Tendermint's RPC. However, the
 query syntax is limited and so this indexer type might be deprecated or removed
 entirely in the future.

docs/architecture/README.md

@@ -98,3 +98,5 @@ Note the context/background should be written in the present tense.
 - [ADR-045: ABCI-Evidence](./adr-045-abci-evidence.md)
 - [ADR-057: RPC](./adr-057-RPC.md)
 - [ADR-069: Node Initialization](./adr-069-flexible-node-initialization.md)
+- [ADR-071: Proposer-Based Timestamps](adr-071-proposer-based-timestamps.md)
+- [ADR-072: Restore Requests for Comments](./adr-072-request-for-comments.md)

docs/architecture/adr-065-custom-event-indexing.md

@@ -24,6 +24,7 @@
 - April 1, 2021: Initial Draft (@alexanderbez)
 - April 28, 2021: Specify search capabilities are only supported through the KV indexer (@marbar3778)
 - May 19, 2021: Update the SQL schema and the eventsink interface (@jayt106)
+- Aug 30, 2021: Update the SQL schema and the psql implementation (@creachadair)
 
 ## Status
 
@@ -145,163 +146,190 @@ The postgres eventsink will not support `tx_search`, `block_search`, `GetTxByHas
 ```sql
 -- Table Definition ----------------------------------------------
 
-CREATE TYPE block_event_type AS ENUM ('begin_block', 'end_block', '');
+-- The blocks table records metadata about each block.
+-- The block record does not include its events or transactions (see tx_results).
+CREATE TABLE blocks (
+  rowid      BIGSERIAL PRIMARY KEY,
 
-CREATE TABLE block_events (
-    id SERIAL PRIMARY KEY,
-    key VARCHAR NOT NULL,
-    value VARCHAR NOT NULL,
-    height INTEGER NOT NULL,
-    type block_event_type,
-    created_at TIMESTAMPTZ NOT NULL,
-    chain_id VARCHAR NOT NULL
+  height     BIGINT NOT NULL,
+  chain_id   VARCHAR NOT NULL,
+
+  -- When this block header was logged into the sink, in UTC.
+  created_at TIMESTAMPTZ NOT NULL,
+
+  UNIQUE (height, chain_id)
 );
 
+-- Index blocks by height and chain, since we need to resolve block IDs when
+-- indexing transaction records and transaction events.
+CREATE INDEX idx_blocks_height_chain ON blocks(height, chain_id);
+
+-- The tx_results table records metadata about transaction results.  Note that
+-- the events from a transaction are stored separately.
 CREATE TABLE tx_results (
-    id SERIAL PRIMARY KEY,
-    tx_result BYTEA NOT NULL,
-    created_at TIMESTAMPTZ NOT NULL
+  rowid BIGSERIAL PRIMARY KEY,
+
+  -- The block to which this transaction belongs.
+  block_id BIGINT NOT NULL REFERENCES blocks(rowid),
+  -- The sequential index of the transaction within the block.
+  index INTEGER NOT NULL,
+  -- When this result record was logged into the sink, in UTC.
+  created_at TIMESTAMPTZ NOT NULL,
+  -- The hex-encoded hash of the transaction.
+  tx_hash VARCHAR NOT NULL,
+  -- The protobuf wire encoding of the TxResult message.
+  tx_result BYTEA NOT NULL,
+
+  UNIQUE (block_id, index)
 );
 
-CREATE TABLE tx_events (
-    id SERIAL PRIMARY KEY,
-    key VARCHAR NOT NULL,
-    value VARCHAR NOT NULL,
-    height INTEGER NOT NULL,
-    hash VARCHAR NOT NULL,
-    tx_result_id SERIAL,
-    created_at TIMESTAMPTZ NOT NULL,
-    chain_id VARCHAR NOT NULL,
-    FOREIGN KEY (tx_result_id)
-        REFERENCES tx_results(id)
-        ON DELETE CASCADE
+-- The events table records events. All events (both block and transaction) are
+-- associated with a block ID; transaction events also have a transaction ID.
+CREATE TABLE events (
+  rowid BIGSERIAL PRIMARY KEY,
+
+  -- The block and transaction this event belongs to.
+  -- If tx_id is NULL, this is a block event.
+  block_id BIGINT NOT NULL REFERENCES blocks(rowid),
+  tx_id    BIGINT NULL REFERENCES tx_results(rowid),
+
+  -- The application-defined type label for the event.
+  type VARCHAR NOT NULL
 );
 
--- Indices -------------------------------------------------------
+-- The attributes table records event attributes.
+CREATE TABLE attributes (
+   event_id      BIGINT NOT NULL REFERENCES events(rowid),
+   key           VARCHAR NOT NULL, -- bare key
+   composite_key VARCHAR NOT NULL, -- composed type.key
+   value         VARCHAR NULL,
 
-CREATE INDEX idx_block_events_key_value ON block_events(key, value);
-CREATE INDEX idx_tx_events_key_value ON tx_events(key, value);
-CREATE INDEX idx_tx_events_hash ON tx_events(hash);
+   UNIQUE (event_id, key)
+);
+
+-- A joined view of events and their attributes. Events that do not have any
+-- attributes are represented as a single row with empty key and value fields.
+CREATE VIEW event_attributes AS
+  SELECT block_id, tx_id, type, key, composite_key, value
+  FROM events LEFT JOIN attributes ON (events.rowid = attributes.event_id);
+
+-- A joined view of all block events (those having tx_id NULL).
+CREATE VIEW block_events AS
+  SELECT blocks.rowid as block_id, height, chain_id, type, key, composite_key, value
+  FROM blocks JOIN event_attributes ON (blocks.rowid = event_attributes.block_id)
+  WHERE event_attributes.tx_id IS NULL;
+
+-- A joined view of all transaction events.
+CREATE VIEW tx_events AS
+  SELECT height, index, chain_id, type, key, composite_key, value, tx_results.created_at
+  FROM blocks JOIN tx_results ON (blocks.rowid = tx_results.block_id)
+  JOIN event_attributes ON (tx_results.rowid = event_attributes.tx_id)
+  WHERE event_attributes.tx_id IS NOT NULL;
 ```
 
 The `PSQLEventSink` will implement the `EventSink` interface as follows
 (some details omitted for brevity):
 
-
 ```go
-func NewPSQLEventSink(connStr string, chainID string) (*PSQLEventSink, error) {
-  db, err := sql.Open("postgres", connStr)
-  if err != nil {
-    return nil, err
-  }
+func NewEventSink(connStr, chainID string) (*EventSink, error) {
+	db, err := sql.Open(driverName, connStr)
+	// ...
 
-  // ...
+	return &EventSink{
+		store:   db,
+		chainID: chainID,
+	}, nil
 }
 
-func (es *PSQLEventSink) IndexBlockEvents(h types.EventDataNewBlockHeader) error {
-  sqlStmt := sq.Insert("block_events").Columns("key", "value", "height", "type", "created_at", "chain_id")
+func (es *EventSink) IndexBlockEvents(h types.EventDataNewBlockHeader) error {
+	ts := time.Now().UTC()
 
-  // index the reserved block height index
-  ts := time.Now()
-  sqlStmt = sqlStmt.Values(types.BlockHeightKey, h.Header.Height, h.Header.Height, "", ts, es.chainID)
+	return runInTransaction(es.store, func(tx *sql.Tx) error {
+		// Add the block to the blocks table and report back its row ID for use
+		// in indexing the events for the block.
+		blockID, err := queryWithID(tx, `
+INSERT INTO blocks (height, chain_id, created_at)
+  VALUES ($1, $2, $3)
+  ON CONFLICT DO NOTHING
+  RETURNING rowid;
+`, h.Header.Height, es.chainID, ts)
+		// ...
 
-  for _, event := range h.ResultBeginBlock.Events {
-    // only index events with a non-empty type
-    if len(event.Type) == 0 {
-      continue
-    }
-
-    for _, attr := range event.Attributes {
-      if len(attr.Key) == 0 {
-        continue
-      }
-
-      // index iff the event specified index:true and it's not a reserved event
-      compositeKey := fmt.Sprintf("%s.%s", event.Type, string(attr.Key))
-      if compositeKey == types.BlockHeightKey {
-        return fmt.Errorf("event type and attribute key \"%s\" is reserved; please use a different key", compositeKey)
-      }
-
-      if attr.GetIndex() {
-        sqlStmt = sqlStmt.Values(compositeKey, string(attr.Value), h.Header.Height, BlockEventTypeBeginBlock, ts, es.chainID)
-      }
-    }
-  }
-
-  // index end_block events...
-  // execute sqlStmt db query...
+		// Insert the special block meta-event for height.
+		if err := insertEvents(tx, blockID, 0, []abci.Event{
+			makeIndexedEvent(types.BlockHeightKey, fmt.Sprint(h.Header.Height)),
+		}); err != nil {
+			return fmt.Errorf("block meta-events: %w", err)
+		}
+		// Insert all the block events. Order is important here,
+		if err := insertEvents(tx, blockID, 0, h.ResultBeginBlock.Events); err != nil {
+			return fmt.Errorf("begin-block events: %w", err)
+		}
+		if err := insertEvents(tx, blockID, 0, h.ResultEndBlock.Events); err != nil {
+			return fmt.Errorf("end-block events: %w", err)
+		}
+		return nil
+	})
 }
 
-func (es *PSQLEventSink) IndexTxEvents(txr []*abci.TxResult) error {
-  sqlStmtEvents := sq.Insert("tx_events").Columns("key", "value", "height", "hash", "tx_result_id", "created_at", "chain_id")
-  sqlStmtTxResult := sq.Insert("tx_results").Columns("tx_result", "created_at")
+func (es *EventSink) IndexTxEvents(txrs []*abci.TxResult) error {
+	ts := time.Now().UTC()
 
-  ts := time.Now()
-  for _, tx := range txr {
-    // store the tx result
-    txBz, err := proto.Marshal(tx)
-    if err != nil {
-      return err
-    }
+	for _, txr := range txrs {
+		// Encode the result message in protobuf wire format for indexing.
+		resultData, err := proto.Marshal(txr)
+		// ...
 
-    sqlStmtTxResult = sqlStmtTxResult.Values(txBz, ts)
+		// Index the hash of the underlying transaction as a hex string.
+		txHash := fmt.Sprintf("%X", types.Tx(txr.Tx).Hash())
 
-    // execute sqlStmtTxResult db query...
-    var txID uint32
-    err = sqlStmtTxResult.QueryRow().Scan(&txID)
-		if err != nil {
+		if err := runInTransaction(es.store, func(tx *sql.Tx) error {
+			// Find the block associated with this transaction.
+			blockID, err := queryWithID(tx, `
+SELECT rowid FROM blocks WHERE height = $1 AND chain_id = $2;
+`, txr.Height, es.chainID)
+			// ...
+
+			// Insert a record for this tx_result and capture its ID for indexing events.
+			txID, err := queryWithID(tx, `
+INSERT INTO tx_results (block_id, index, created_at, tx_hash, tx_result)
+  VALUES ($1, $2, $3, $4, $5)
+  ON CONFLICT DO NOTHING
+  RETURNING rowid;
+`, blockID, txr.Index, ts, txHash, resultData)
+			// ...
+
+			// Insert the special transaction meta-events for hash and height.
+			if err := insertEvents(tx, blockID, txID, []abci.Event{
+				makeIndexedEvent(types.TxHashKey, txHash),
+				makeIndexedEvent(types.TxHeightKey, fmt.Sprint(txr.Height)),
+			}); err != nil {
+				return fmt.Errorf("indexing transaction meta-events: %w", err)
+			}
+			// Index any events packaged with the transaction.
+			if err := insertEvents(tx, blockID, txID, txr.Result.Events); err != nil {
+				return fmt.Errorf("indexing transaction events: %w", err)
+			}
+			return nil
+
+		}); err != nil {
 			return err
 		}
-
-    // index the reserved height and hash indices
-    hash := types.Tx(tx.Tx).Hash()
-    sqlStmtEvents = sqlStmtEvents.Values(types.TxHashKey, hash, tx.Height, hash, txID, ts, es.chainID)
-    sqlStmtEvents = sqlStmtEvents.Values(types.TxHeightKey, tx.Height, tx.Height, hash, txID, ts, es.chainID)
-
-    for _, event := range result.Result.Events {
-      // only index events with a non-empty type
-      if len(event.Type) == 0 {
-        continue
-      }
-
-      for _, attr := range event.Attributes {
-        if len(attr.Key) == 0 {
-          continue
-        }
-
-        // index if `index: true` is set
-        compositeTag := fmt.Sprintf("%s.%s", event.Type, string(attr.Key))
-			
-        // ensure event does not conflict with a reserved prefix key
-        if compositeTag == types.TxHashKey || compositeTag == types.TxHeightKey {
-          return fmt.Errorf("event type and attribute key \"%s\" is reserved; please use a different key", compositeTag)
-        }
-		
-        if attr.GetIndex() {
-          sqlStmtEvents = sqlStmtEvents.Values(compositeKey, string(attr.Value), tx.Height, hash, txID, ts, es.chainID)
-        }
-      }
-    }
-  }
-  
-  // execute sqlStmtEvents db query...
+	}
+	return nil
 }
 
-func (es *PSQLEventSink) SearchBlockEvents(ctx context.Context, q *query.Query) ([]int64, error) {
-  return nil, errors.New("block search is not supported via the postgres event sink")
-}
+// SearchBlockEvents is not implemented by this sink, and reports an error for all queries.
+func (es *EventSink) SearchBlockEvents(ctx context.Context, q *query.Query) ([]int64, error)
 
-func (es *PSQLEventSink) SearchTxEvents(ctx context.Context, q *query.Query) ([]*abci.TxResult, error) {
-  return nil, errors.New("tx search is not supported via the postgres event sink")
-}
+// SearchTxEvents is not implemented by this sink, and reports an error for all queries.
+func (es *EventSink) SearchTxEvents(ctx context.Context, q *query.Query) ([]*abci.TxResult, error)
 
-func (es *PSQLEventSink) GetTxByHash(hash []byte) (*abci.TxResult, error) {
-	return nil, errors.New("getTxByHash is not supported via the postgres event sink")
-}
+// GetTxByHash is not implemented by this sink, and reports an error for all queries.
+func (es *EventSink) GetTxByHash(hash []byte) (*abci.TxResult, error)
 
-func (es *PSQLEventSink) HasBlock(h int64) (bool, error) {
-	return false, errors.New("hasBlock is not supported via the postgres event sink")
-}
+// HasBlock is not implemented by this sink, and reports an error for all queries.
+func (es *EventSink) HasBlock(h int64) (bool, error)
 ```
 
 ### Configuration

docs/architecture/adr-072-request-for-comments.md

@@ -0,0 +1,105 @@
+# ADR 72: Restore Requests for Comments
+
+## Changelog
+
+- 20-Aug-2021: Initial draft (@creachadair)
+
+## Status
+
+Proposed
+
+## Context
+
+In the past, we kept a collection of Request for Comments (RFC) documents in `docs/rfc`.
+Prior to the creation of the ADR process, these documents were used to document
+design and implementation decisions about Tendermint Core. The RFC directory
+was removed in favor of ADRs, in commit 3761aa69 (PR
+[\#6345](https://github.com/tendermint/tendermint/pull/6345)).
+
+For issues where an explicit design decision or implementation change is
+required, an ADR is generally preferable to an open-ended RFC: An ADR is
+relatively narrowly-focused, identifies a specific design or implementation
+question, and documents the consensus answer to that question.
+
+Some discussions are more open-ended, however, or don't require a specific
+decision to be made (yet). Such conversations are still valuable to document,
+and several members of the Tendermint team have been doing so by writing gists
+or Google docs to share them around. That works well enough in the moment, but
+gists do not support any kind of collaborative editing, and both gists and docs
+are hard to discover after the fact. Google docs have much better collaborative
+editing, but are worse for discoverability, especially when contributors span
+different Google accounts.
+
+Discoverability is important, because these kinds of open-ended discussions are
+useful to people who come later -- either as new team members or as outside
+contributors seeking to use and understand the thoughts behind our designs and
+the architectural decisions that arose from those discussion.
+
+With these in mind, I propose that:
+
+-  We re-create a new, initially empty `docs/rfc` directory in the repository,
+   and use it to capture these kinds of open-ended discussions in supplement to
+   ADRs.
+
+-  Unlike in the previous RFC scheme, documents in this new directory will
+   _not_ be used directly for decision-making. This is the key difference
+   between an RFC and an ADR.
+
+   Instead, an RFC will exist to document background, articulate general
+   principles, and serve as a historical record of discussion and motivation.
+
+   In this system, an RFC may _only_ result in a decision indirectly, via ADR
+   documents created in response to the RFC.
+
+   **In short:** If a decision is required, write an ADR; otherwise if a
+   sufficiently broad discussion is needed, write an RFC.
+
+Just so that there is a consistent format, I also propose that:
+
+-  RFC files are named `rfc-XXX-title.{md,rst,txt}` and are written in plain
+   text, Markdown, or ReStructured Text.
+
+-  Like an ADR, an RFC should include a high-level change log at the top of the
+   document, and sections for:
+
+     * Abstract: A brief, high-level synopsis of the topic.
+     * Background: Any background necessary to understand the topic.
+     * Discussion: Detailed discussion of the issue being considered.
+
+-  Unlike an ADR, an RFC does _not_ include sections for Decisions, Detailed
+   Design, or evaluation of proposed solutions. If an RFC leads to a proposal
+   for an actual architectural change, that must be recorded in an ADR in the
+   usual way, and may refer back to the RFC in its References section.
+
+## Alternative Approaches
+
+Leaving aside implementation details, the main alternative to this proposal is
+to leave things as they are now, with ADRs as the only log of record and other
+discussions being held informally in whatever medium is convenient at the time.
+
+## Decision
+
+(pending)
+
+## Detailed Design
+
+- Create a new `docs/rfc` directory in the `tendermint` repository. Note that
+  this proposal intentionally does _not_ pull back the previous contents of
+  that path from Git history, as those documents were appropriately merged into
+  the ADR process.
+
+- Create a `README.md` for RFCs that explains the rules and their relationship
+  to ADRs.
+
+- Create an `rfc-template.md` file for RFC files.
+
+## Consequences
+
+### Positive
+
+- We will have a more discoverable place to record open-ended discussions that
+  do not immediately result in a design change.
+
+### Negative
+
+- Potentially some people could be confused about the RFC/ADR distinction.

docs/nodes/configuration.md

@@ -36,10 +36,6 @@ proxy-app = "tcp://127.0.0.1:26658"
 # A custom human readable name for this node
 moniker = "anonymous"
 
-# If this node is many blocks behind the tip of the chain, BlockSync
-# allows them to catchup quickly by downloading blocks in parallel
-# and verifying their commits
-fast-sync = true
 
 # Mode of Node: full | validator | seed (default: "validator")
 # * validator node (default)
@@ -356,11 +352,16 @@ temp-dir = ""
 #######################################################
 ###       BlockSync Configuration Connections       ###
 #######################################################
-[fastsync]
+[blocksync]
+
+# If this node is many blocks behind the tip of the chain, BlockSync
+# allows them to catchup quickly by downloading blocks in parallel
+# and verifying their commits
+enable = true
 
 # Block Sync version to use:
-#   1) "v0" (default) - the legacy block sync implementation
-#   2) "v2" - complete redesign of v0, optimized for testability & readability
+#   1) "v0" (default) - the standard block sync implementation
+#   2) "v2" - DEPRECATED, please use v0
 version = "v0"
 
 #######################################################

docs/rfc/README.md

@@ -0,0 +1,43 @@
+---
+order: 1
+parent:
+  order: false
+---
+
+# Requests for Comments
+
+A Request for Comments (RFC) is a record of discussion on an open-ended topic
+related to the design and implementation of Tendermint Core, for which no
+immediate decision is required.
+
+The purpose of an RFC is to serve as a historical record of a high-level
+discussion that might otherwise only be recorded in an ad hoc way (for example,
+via gists or Google docs) that are difficult to discover for someone after the
+fact. An RFC _may_ give rise to more specific architectural _decisions_ for
+Tendermint, but those decisions must be recorded separately in [Architecture
+Decision Records (ADR)](./../architecture).
+
+As a rule of thumb, if you can articulate a specific question that needs to be
+answered, write an ADR. If you need to explore the topic and get input from
+others to know what questions need to be answered, an RFC may be appropriate.
+
+## RFC Content
+
+An RFC should provide:
+
+- A **changelog**, documenting when and how the RFC has changed.
+- An **abstract**, briefly summarizing the topic so the reader can quickly tell
+  whether it is relevant to their interest.
+- Any **background** a reader will need to understand and participate in the
+  substance of the discussion (links to other documents are fine here).
+- The **discussion**, the primary content of the document.
+
+The [rfc-template.md](./rfc-template.md) file includes placeholders for these
+sections.
+
+## Table of Contents
+
+- [RFC-000: P2P Roadmap](./rfc-000-p2p-roadmap.rst)
+- [RFC-001: Storage Engines](./rfc-001-storage-engine.rst)
+
+<!-- - [RFC-NNN: Title](./rfc-NNN-title.md) -->

docs/rfc/rfc-000-p2p-roadmap.rst

@@ -0,0 +1,316 @@
+====================
+RFC 000: P2P Roadmap
+====================
+
+Changelog
+---------
+
+- 2021-08-20: Completed initial draft and distributed via a gist
+- 2021-08-25: Migrated as an RFC and changed format
+
+Abstract
+--------
+
+This document discusses the future of peer network management in Tendermint, with
+a particular focus on features, semantics, and a proposed roadmap.
+Specifically, we consider libp2p as a tool kit for implementing some fundamentals.
+
+Background
+----------
+
+For the 0.35 release cycle the switching/routing layer of Tendermint was
+replaced. This work was done "in place," and produced a version of Tendermint
+that was backward-compatible and interoperable with previous versions of the
+software. While there are new p2p/peer management constructs in the new
+version (e.g. ``PeerManager`` and ``Router``), the main effect of this change
+was to simplify the ways that other components within Tendermint interacted with
+the peer management layer, and to make it possible for higher-level components
+(specifically the reactors), to be used and tested more independently.
+
+This refactoring, which was a major undertaking, was entirely necessary to
+enable areas for future development and iteration on this aspect of
+Tendermint. There are also a number of potential user-facing features that
+depend heavily on the p2p layer: additional transport protocols, transport
+compression, improved resilience to network partitions. These improvements to
+modularity, stability, and reliability of the p2p system will also make
+ongoing maintenance and feature development easier in the rest of Tendermint.
+
+Critique of Current Peer-to-Peer Infrastructure
+---------------------------------------
+
+The current (refactored) P2P stack is an improvement on the previous iteration
+(legacy), but as of 0.35, there remains room for improvement in the design and
+implementation of the P2P layer. 
+
+Some limitations of the current stack include:
+
+- heavy reliance on buffering to avoid backups in the flow of components,
+  which is fragile to maintain and can lead to unexpected memory usage
+  patterns and forces the routing layer to make decisions about when messages
+  should be discarded. 
+  
+- the current p2p stack relies on convention (rather than the compiler) to
+  enforce the API boundaries and conventions between reactors and the router,
+  making it very easy to write "wrong" reactor code or introduce a bad
+  dependency.
+
+- the current stack is probably more complex and difficult to maintain because
+  the legacy system must coexist with the new components in 0.35. When the
+  legacy stack is removed there are some simple changes that will become
+  possible and could reduce the complexity of the new system. (e.g. `#6598
+  <https://github.com/tendermint/tendermint/issues/6598>`_.)
+
+- the current stack encapsulates a lot of information about peers, and makes it
+  difficult to expose that information to monitoring/observability tools. This
+  general opacity also makes it difficult to interact with the peer system
+  from other areas of the code base (e.g. tests, reactors).
+  
+- the legacy stack provided some control to operators to force the system to
+  dial new peers or seed nodes or manipulate the topology of the system _in
+  situ_. The current stack can't easily provide this, and while the new stack
+  may have better behavior, it does leave operators hands tied.
+
+Some of these issues will be resolved early in the 0.36 cycle, with the
+removal of the legacy components.
+
+The 0.36 release also provides the opportunity to make changes to the
+protocol, as the release will not be compatible with previous releases.
+
+Areas for Development
+---------------------
+
+These sections describe features that may make sense to include in a Phase 2 of
+a P2P project.
+
+Internal Message Passing
+~~~~~~~~~~~~~~~~~~~~~~~~
+
+Currently, there's no provision for intranode communication using the P2P
+layer, which means when two reactors need to interact with each other they
+have to have dependencies on each other's interfaces, and
+initialization. Changing these interactions (e.g. transitions between
+blocksync and consensus) from procedure calls to message passing.
+
+This is a relatively simple change and could be implemented with the following
+components:
+
+- a constant to represent "local" delivery as  the ``To``` field on
+  ``p2p.Envelope``.
+  
+- special path for routing local messages that doesn't require message
+  serialization (protobuf marshalling/unmarshaling).
+  
+Adding these semantics, particularly if in conjunction with synchronous
+semantics provides a solution to dependency graph problems currently present
+in the Tendermint codebase, which will simplify development, make it possible
+to isolate components for testing. 
+
+Eventually, this will also make it possible to have a logical Tendermint node
+running in multiple processes or in a collection of containers, although the
+usecase of this may be debatable.
+
+Synchronous Semantics (Paired Request/Response)
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+In the current system, all messages are sent with fire-and-forget semantics,
+and there's no coupling between a request sent via the p2p layer, and a
+response. These kinds of semantics would simplify the implementation of
+state and block sync reactors, and make intra-node message passing more
+powerful.
+
+For some interactions, like gossiping transactions between the mempools of
+different nodes, fire-and-forget semantics make sense, but for other
+operations the missing link between requests/responses leads to either
+inefficiency when a node fails to respond or becomes unavailable, or code that
+is just difficult to follow.
+
+To support this kind of work, the protocol would need to accommodate some kind
+of request/response ID to allow identifying out-of-order responses over a
+single connection. Additionally, expanded the programming model of the
+``p2p.Channel`` to accommodate some kind of _future_ or similar paradigm to
+make it viable to write reactor code without needing for the reactor developer
+to wrestle with lower level concurency constructs.
+
+
+Timeout Handling (QoS)
+~~~~~~~~~~~~~~~~~~~~~~
+
+Currently, all timeouts, buffering, and QoS features are handled at the router
+layer, and the reactors are implemented in ways that assume/require
+asynchronous operation. This both increases the required complexity at the
+routing layer, and means that misbehavior at the reactor level is difficult to
+detect or attribute. Additionally, the current system provides three main
+parameters to control quality of service:
+
+- buffer sizes for channels and queues.
+  
+- priorities for channels
+
+- queue implementation details for shedding load.
+
+These end up being quite coarse controls, and changing the settings are
+difficult because as the queues and channels are able to buffer large numbers
+of messages it can be hard to see the impact of a given change, particularly
+in our extant test environment. In general, we should endeavor to: 
+
+- set real timeouts, via contexts, on most message send operations, so that
+  senders rather than queues can be responsible for timeout
+  logic. Additionally, this will make it possible to avoid sending messages
+  during shutdown.
+  
+- reduce (to the greatest extent possible) the amount of buffering in
+  channels and the queues, to more readily surface backpressure and reduce the
+  potential for buildup of stale messages.
+
+Stream Based Connection Handling
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+Currently the transport layer is message based, which makes sense from a
+mental model of how the protocol works, but makes it more difficult to
+implement transports and connection types, as it forces a higher level view of
+the connection and interaction which makes it harder to implement for novel
+transport types and makes it more likely that message-based caching and rate
+limiting will be implemented at the transport layer rather than at a more
+appropriate level.
+
+The transport then, would be responsible for negitating the connection and the
+handshake and otherwise behave like a socket/file discriptor with ``Read` and
+``Write`` methods.
+
+While this was included in the initial design for the new P2P layer, it may be
+obviated entirely if the transport and peer layer is replaced with libp2p,
+which is primarily stream based.
+
+Service Discovery
+~~~~~~~~~~~~~~~~~
+
+In the current system, Tendermint assumes that all nodes in a network are
+largely equivelent, and nodes tend to be "chatty" making many requests of
+large numbers of peers and waiting for peers to (hopefully) respond. While
+this works and has allowed Tendermint to get to a certain point, this both
+produces a theoretical scaling bottle neck and makes it harder to test and
+verify components of the system.
+
+In addition to peer's identity and connection information, peers should be
+able to advertise a number of services or capabilities, and node operators or
+developers should be able to specify peer capability requirements (e.g. target
+at least <x>-percent of peers with <y> capability.)  
+
+These capabilities may be useful in selecting peers to send messages to, it
+may make sense to extend Tendermint's message addressing capability to allow
+reactors to send messages to groups of peers based on role rather than only
+allowing addressing to one or all peers.
+
+Having a good service discovery mechanism may pair well with the synchronous
+semantics (request/response) work, as it allows reactors to "make a request of
+a peer with <x> capability and wait for the response," rather force the
+reactors to need to track the capabilities or state of specific peers.
+
+Solutions
+---------
+
+Continued Homegrown Implementation
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The current peer system is homegrown and is conceptually compatible with the
+needs of the project, and while there are limitations to the system, the p2p
+layer is not (currently as of 0.35) a major source of bugs or friction during
+development. 
+
+However, the current implementation makes a number of allowances for
+interoperability, and there are a collection of iterative improvements that
+should be considered in the next couple of releases. To maintain the current
+implementation, upcoming work would include:
+
+- change the ``Transport`` mechanism to facilitate easier implementations.
+
+- implement different ``Transport`` handlers to be able to manage peer
+  connections using different protocols (e.g. QUIC, etc.)
+
+- entirely remove the constructs and implementations of the legacy peer
+  implementation. 
+  
+- establish and enforce clearer chains of responsibility for connection
+  establishment (e.g. handshaking, setup,) which is currently shared between
+  three components. 
+
+- report better metrics regarding the into the state of peers and network
+  connectivity, which are opaque outside of the system. This is constrained at
+  the moment as a side effect of the split responsibility for connection
+  establishment.
+  
+- extend the PEX system to include service information so that ndoes in the
+  network weren't necessarily homogeneous.
+
+While maintaining a bespoke peer management layer would seem to distract from
+development of core functionality, the truth is that (once the legacy code is
+removed,) the scope of the peer layer is relatively small from a maintenance
+perspective, and having control at this layer might actually afford the
+project with the ability to more rapidly iterate on some features.
+
+LibP2P
+~~~~~~
+
+LibP2P provides components that, approximately, account for the
+``PeerManager`` and ``Transport`` components of the current (new) P2P
+stack. The Go APIs seem reasonable, and being able to externalize the
+implementation details of peer and connection management seems like it could
+provide a lot of benefits, particularly in supporting a more active ecosystem.
+
+In general the API provides the kind of stream-based, multi-protocol
+supporting, and idiomatic baseline for implementing a peer layer. Additionally
+because it handles peer exchange and connection management at a lower
+level, by using libp2p it'd be possible to remove a good deal of code in favor
+of just using libp2p. Having said that, Tendermint's P2P layer covers a
+greater scope (e.g. message routing to different peers) and that layer is
+something that Tendermint might want to retain.
+
+The are a number of unknowns that require more research including how much of
+a peer database the Tendermint engine itself needs to maintain, in order to
+support higher level operations (consensus, statesync), but it might be the
+case that our internal systems need to know much less about peers than
+otherwise specified. Similarly, the current system has a notion of peer
+scoring that cannot be communicated to libp2p, which may be fine as this is
+only used to support peer exchange (PEX,) which would become a property libp2p
+and not expressed in it's current higher-level form. 
+
+In general, the effort to switch to libp2p would involve: 
+
+- timing it during an appropriate protocol-breaking window, as it doesn't seem
+  viable to support both libp2p *and* the current p2p protocol. 
+  
+- providing some in-memory testing network to support the use case that the
+  current ``p2p.MemoryNetwork`` provides.
+
+- re-homing the ``p2p.Router`` implementation on top of libp2p components to
+  be able to maintain the current reactor implementations.
+  
+Open question include: 
+
+- how much local buffering should we be doing? It sort of seems like we should
+  figure out what the expected behavior is for libp2p for QoS-type
+  functionality, and if our requirements mean that we should be implementing
+  this on top of things ourselves?
+
+- if Tendermint was going to use libp2p, how would libp2p's stability
+  guarantees (protocol, etc.) impact/constrain Tendermint's stability
+  guarantees?
+
+- what kind of introspection does libp2p provide, and to what extend would
+  this change or constrain the kind of observability that Tendermint is able
+  to provide?
+
+- how do efforts to select "the best" (healthy, close, well-behaving, etc.)
+  peers work out if Tendermint is not maintaining a local peer database?
+  
+- would adding additional higher level semantics (internal message passing,
+  request/response pairs, service discovery, etc.) facilitate removing some of
+  the direct linkages between constructs/components in the system and reduce
+  the need for Tendermint nodes to maintain state about its peers?
+
+References
+----------
+
+- `Tracking Ticket for P2P Refactor Project <https://github.com/tendermint/tendermint/issues/5670>`_ 
+- `ADR 61: P2P Refactor Scope <../architecture/adr-061-p2p-refactor-scope.md>`_
+- `ADR 62: P2P Architecture and Abstraction <../architecture/adr-061-p2p-architecture.md>`_

docs/rfc/rfc-001-storage-engine.rst

@@ -0,0 +1,179 @@
+===========================================
+RFC 001: Storage Engines and Database Layer
+===========================================
+
+Changelog
+---------
+
+- 2021-04-19: Initial Draft (gist)
+- 2021-09-02: Migrated to RFC folder, with some updates  
+
+Abstract
+--------
+
+The aspect of Tendermint that's responsible for persistence and storage (often
+"the database" internally) represents a bottle neck in the architecture of the
+platform, that the 0.36 release presents a good opportunity to correct. The
+current storage engine layer provides a great deal of flexibility that is
+difficult for users to leverage or benefit from, while also making it harder
+for Tendermint Core developers to deliver improvements on storage engine. This
+RFC discusses the possible improvements to this layer of the system.
+
+Background
+----------
+
+Tendermint has a very thin common wrapper that makes Tendermint itself
+(largely) agnostic to the data storage layer (within the realm of the popular
+key-value/embedded databases.) This flexibility is not particularly useful:
+the benefits of a specific database engine in the context of Tendermint is not
+particularly well understood, and the maintenance burden for multiple backends
+is not commensurate with the benefit provided. Additionally, because the data
+storage layer is handled generically, and most tests run with an in-memory
+framework, it's difficult to take advantage of any higher-level features of a
+database engine.
+
+Ideally, developers within Tendermint will be able to interact with persisted
+data via an interface that can function, approximately like an object
+store, and this storage interface will be able to accommodate all existing
+persistence workloads (e.g. block storage, local peer management information
+like the "address book", crash-recovery log like the WAL.) In addition to
+providing a more ergonomic interface and new semantics, by selecting a single
+storage engine tendermint can use native durability and atomicity features of
+the storage engine and simplify its own implementations. 
+
+Data Access Patterns
+~~~~~~~~~~~~~~~~~~~~
+
+Tendermint's data access patterns have the following characteristics:
+
+- aggregate data size often exceeds memory.
+
+- data is rarely mutated after it's written for most data (e.g. blocks), but
+  small amounts of working data is persisted by nodes and is frequently
+  mutated (e.g. peer information, validator information.)
+
+- read patterns can be quite random.
+
+- crash resistance and crash recovery, provided by write-ahead-logs (in
+  consensus, and potentially for the mempool) should allow the system to
+  resume work after an unexpected shut down.
+
+Project Goals
+~~~~~~~~~~~~~
+
+As we think about replacing the current persistence layer, we should consider
+the following high level goals: 
+
+- drop dependencies on storage engines that have a CGo dependency.
+
+- encapsulate data format and data storage from higher-level services
+  (e.g. reactors) within tendermint.
+
+- select a storage engine that does not incur any additional operational
+  complexity (e.g. database should be embedded.)
+
+- provide database semantics with sufficient ACID, snapshots, and
+  transactional support.
+
+Open Questions
+~~~~~~~~~~~~~~
+
+The following questions remain:
+
+- what kind of data-access concurrency does tendermint require?
+
+- would tendermint users SDK/etc. benefit from some shared database
+  infrastructure?
+  
+  - In earlier conversations it seemed as if the SDK has selected Badger and
+    RocksDB for their storage engines, and it might make sense to be able to
+    (optionally) pass a handle to a Badger instance between the libraries in
+    some cases.
+
+- what are typical data sizes, and what kinds of memory sizes can we expect
+  operators to be able to provide?
+
+- in addition to simple persistence, what kind of additional semantics would
+  tendermint like to enjoy (e.g. transactional semantics, unique constraints,
+  indexes, in-place-updates, etc.)?
+
+Decision Framework
+~~~~~~~~~~~~~~~~~~
+
+Given the constraint of removing the CGo dependency, the decision is between
+"badger" and "boltdb" (in the form of the etcd/CoreOS fork,) as low level. On
+top of this and somewhat orthogonally, we must also decide on the interface to
+the database and how the larger application will have to interact with the
+database layer. Users of the data layer shouldn't ever need to interact with
+raw byte slices from the database, and should mostly have the experience of
+interacting with Go-types.
+
+Badger is more consistently developed and has a broader feature set than
+Bolt. At the same time, Badger is likely more memory intensive and may have
+more overhead in terms of open file handles given it's model. At first glance,
+Badger is the obvious choice: it's actively developed and it has a lot of
+features that could be useful. Bolt is not without some benefits: it's stable
+and is maintained by the etcd folks, it's simpler model (single memory mapped
+file, etc,) may be easier to reason about.
+
+I propose that we consider the following specific questions about storage
+engines:
+
+- does Badger's evolving development, which may result in data file format
+  changes in the future, and could restrict our access to using the latest
+  version of the library between major upgrades, present a problem?
+
+- do we do we have goals/concerns about memory footprint that Badger may
+  prevent us from hitting, particularly as data sets grow over time?
+
+- what kind of additional tooling might we need/like to build (dump/restore,
+  etc.)?
+
+- do we want to run unit/integration tests against a data files on disk rather
+  than relying exclusively on the memory database?
+
+Project Scope
+~~~~~~~~~~~~~
+
+This project will consist of the following aspects:
+
+- selecting a storage engine, and modifying the tendermint codebase to
+  disallow any configuration of the storage engine outside of the tendermint. 
+
+- remove the dependency on the current tm-db interfaces and replace with some
+  internalized, safe, and ergonomic interface for data persistence with all
+  required database semantics.
+
+- update core tendermint code to use the new interface and data tools.
+
+Next Steps
+~~~~~~~~~~
+
+- circulate the RFC, and discuss options with appropriate stakeholders. 
+  
+- write brief ADR to summarize decisions around technical decisions reached
+  during the RFC phase. 
+
+References
+----------
+
+- `bolddb <https://github.com/etcd-io/bbolt>`_
+- `badger <https://github.com/dgraph-io/badger>`_
+- `badgerdb overview <https://dbdb.io/db/badgerdb>`_
+- `botldb overview <https://dbdb.io/db/boltdb>`_
+- `boltdb vs badger <https://tech.townsourced.com/post/boltdb-vs-badger>`_
+- `bolthold <https://github.com/timshannon/bolthold>`_
+- `badgerhold <https://github.com/timshannon/badgerhold>`_
+- `Pebble <https://github.com/cockroachdb/pebble>`_
+- `SDK Issue Regarding IVAL <https://github.com/cosmos/cosmos-sdk/issues/7100>`_
+- `SDK Discussion about SMT/IVAL <https://github.com/cosmos/cosmos-sdk/discussions/8297>`_
+
+Discussion
+----------
+
+- All things being equal, my tendency would be to use badger, with badgerhold
+  (if that makes sense) for its ergonomics and indexing capabilities, which
+  will require some small selection of wrappers for better write transaction
+  support. This is a weakly held tendency/belief and I think it would be
+  useful for the RFC process to build consensus (or not) around this basic
+  assumption.

docs/rfc/rfc-template.md

@@ -0,0 +1,35 @@
+# RFC {RFC-NUMBER}: {TITLE}
+
+## Changelog
+
+- {date}: {changelog}
+
+## Abstract
+
+> A brief high-level synopsis of the topic of discussion for this RFC, ideally
+> just a few sentences.  This should help the reader quickly decide whether the
+> rest of the discussion is relevant to their interest.
+
+## Background
+
+> Any context or orientation needed for a reader to understand and participate
+> in the substance of the Discussion. If necessary, this section may include
+> links to other documentation or sources rather than restating existing
+> material, but should provide enough detail that the reader can tell what they
+> need to read to be up-to-date.
+
+### References
+
+> Links to external materials needed to follow the discussion may be added here.
+>
+> In addition, if the discussion in a request for comments leads to any design
+> decisions, it may be helpful to add links to the ADR documents here after the
+> discussion has settled.
+
+## Discussion
+
+> This section contains the core of the discussion.
+>
+> There is no fixed format for this section, but ideally changes to this
+> section should be updated before merging to reflect any discussion that took
+> place on the PR that made those changes.

docs/tendermint-core/block-sync.md

@@ -17,9 +17,9 @@ consensus gossip protocol.
 
 ## Using Block Sync
 
-To support faster syncing, Tendermint offers a `fast-sync` mode, which
+To support faster syncing, Tendermint offers a `blocksync` mode, which
 is enabled by default, and can be toggled in the `config.toml` or via
-`--fast_sync=false`.
+`--blocksync.enable=false`.
 
 In this mode, the Tendermint daemon will sync hundreds of times faster
 than if it used the real-time consensus process. Once caught up, the
@@ -29,18 +29,23 @@ has at least one peer and it's height is at least as high as the max
 reported peer height. See [the IsCaughtUp
 method](https://github.com/tendermint/tendermint/blob/b467515719e686e4678e6da4e102f32a491b85a0/blockchain/pool.go#L128).
 
-Note: There are two versions of Block Sync. We recommend using v0 as v2 is still in beta.
+Note: There are multiple versions of Block Sync. Please use v0 as the other versions are no longer supported.
   If you would like to use a different version you can do so by changing the version in the `config.toml`:
 
 ```toml
 #######################################################
 ###       Block Sync Configuration Connections       ###
 #######################################################
-[fastsync]
+[blocksync]
+
+# If this node is many blocks behind the tip of the chain, BlockSync
+# allows them to catchup quickly by downloading blocks in parallel
+# and verifying their commits
+enable = true
 
 # Block Sync version to use:
-#   1) "v0" (default) - the legacy Block Sync implementation
-#   2) "v2" - complete redesign of v0, optimized for testability & readability
+#   1) "v0" (default) - the standard Block Sync implementation
+#   2) "v2" - DEPRECATED, please use v0
 version = "v0"
 ```
 
@@ -55,4 +60,4 @@ the network best height, it will switches to the state sync mechanism and then e
 another event for exposing the fast-sync `complete` status and the state `height`.
 
 The user can query the events by subscribing `EventQueryBlockSyncStatus`
-Please check [types](https://pkg.go.dev/github.com/tendermint/tendermint/types?utm_source=godoc#pkg-constants) for the details.
+Please check [types](https://pkg.go.dev/github.com/tendermint/tendermint/types?utm_source=godoc#pkg-constants) for the details.

docs/tools/debugging/README.md

@@ -62,3 +62,30 @@ given destination directory. Each archive will contain:
 
 Note: goroutine.out and heap.out will only be written if a profile address is
 provided and is operational. This command is blocking and will log any error.
+
+## Tendermint Inspect
+
+Tendermint includes an `inspect` command for querying Tendermint's state store and block
+store over Tendermint RPC.
+
+When the Tendermint consensus engine detects inconsistent state, it will crash the
+entire Tendermint process. 
+While in this inconsistent state, a node running Tendermint's consensus engine will not start up. 
+The `inspect` command runs only a subset of Tendermint's RPC endpoints for querying the block store
+and state store. 
+`inspect` allows operators to query a read-only view of the stage.
+`inspect` does not run the consensus engine at all and can therefore be used to debug
+processes that have crashed due to inconsistent state. 
+
+
+To start the `inspect` process, run
+```bash
+tendermint inspect
+```
+
+### RPC endpoints
+The list of available RPC endpoints can be found by making a request to the RPC port.
+For an `inspect` process running on `127.0.0.1:26657`, navigate your browser to 
+`http://127.0.0.1:26657/` to retrieve the list of enabled RPC endpoints.
+
+Additional information on the Tendermint RPC endpoints can be found in the [rpc documentation](https://docs.tendermint.com/master/rpc).

docs/tools/debugging/pro.md

@@ -64,13 +64,42 @@ It won’t kill the node, but it will gather all of the above data and package i
 
 At this point, depending on how severe the degradation is, you may want to restart the process.
 
+## Tendermint Inspect
+
+What if the Tendermint node will not start up due to inconsistent consensus state? 
+
+When a node running the Tendermint consensus engine detects an inconsistent state 
+it will crash the entire Tendermint process. 
+The Tendermint consensus engine cannot be run in this inconsistent state and the so node
+will fail to start up as a result.
+The Tendermint RPC server can provide valuable information for debugging in this situation.
+The Tendermint `inspect` command will run a subset of the Tendermint RPC server 
+that is useful for debugging inconsistent state.
+
+### Running inspect
+
+Start up the `inspect` tool on the machine where Tendermint crashed using: 
+```bash
+tendermint inspect --home=</path/to/app.d>
+```
+
+`inspect` will use the data directory specified in your Tendermint configuration file.
+`inspect` will also run the RPC server at the address specified in your Tendermint configuration file.
+
+### Using inspect
+
+With the `inspect` server running, you can access RPC endpoints that are critically important
+for debugging.
+Calling the `/status`, `/consensus_state` and `/dump_consensus_state` RPC endpoint 
+will return useful information about the Tendermint consensus state.
+
 ## Outro
 
-We’re hoping that the `tendermint debug` subcommand will become de facto the first response to any accidents.
+We’re hoping that these Tendermint tools will become de facto the first response for any accidents.
 
-Let us know what your experience has been so far! Have you had a chance to try `tendermint debug` yet?
+Let us know what your experience has been so far! Have you had a chance to try `tendermint debug` or `tendermint inspect` yet?
 
-Join our chat, where we discuss the current issues and future improvements.
+Join our [discord chat](https://discord.gg/vcExX9T), where we discuss the current issues and future improvements.
 
 —
 

docs/tutorials/go-built-in.md

@@ -388,7 +388,6 @@ func main() {
  c := make(chan os.Signal, 1)
  signal.Notify(c, os.Interrupt, syscall.SIGTERM)
  <-c
- os.Exit(0)
 }
 
 func newTendermint(app abci.Application, configFile string) (*nm.Node, error) {
@@ -564,7 +563,6 @@ defer func() {
 c := make(chan os.Signal, 1)
 signal.Notify(c, os.Interrupt, syscall.SIGTERM)
 <-c
-os.Exit(0)
 ```
 
 ## 1.5 Getting Up and Running

go.mod

@@ -4,7 +4,6 @@ go 1.16
 
 require (
 	github.com/BurntSushi/toml v0.4.1
-	github.com/Masterminds/squirrel v1.5.0
 	github.com/Workiva/go-datastructures v1.0.53
 	github.com/adlio/schema v1.1.13
 	github.com/btcsuite/btcd v0.22.0-beta
@@ -13,30 +12,33 @@ require (
 	github.com/go-kit/kit v0.11.0
 	github.com/gogo/protobuf v1.3.2
 	github.com/golang/protobuf v1.5.2
-	github.com/golangci/golangci-lint v1.42.0
+	github.com/golangci/golangci-lint v1.42.1
 	github.com/google/orderedcode v0.0.1
 	github.com/google/uuid v1.3.0
 	github.com/gorilla/websocket v1.4.2
 	github.com/grpc-ecosystem/go-grpc-middleware v1.3.0
 	github.com/grpc-ecosystem/go-grpc-prometheus v1.2.0
-	github.com/lib/pq v1.10.2
+	github.com/lib/pq v1.10.3
 	github.com/libp2p/go-buffer-pool v0.0.2
 	github.com/minio/highwayhash v1.0.2
+	github.com/mroth/weightedrand v0.4.1
 	github.com/oasisprotocol/curve25519-voi v0.0.0-20210609091139-0a56a4bca00b
 	github.com/ory/dockertest v3.3.5+incompatible
 	github.com/prometheus/client_golang v1.11.0
 	github.com/rcrowley/go-metrics v0.0.0-20200313005456-10cdbea86bc0
 	github.com/rs/cors v1.8.0
-	github.com/rs/zerolog v1.23.0
+	github.com/rs/zerolog v1.24.0
 	github.com/sasha-s/go-deadlock v0.2.1-0.20190427202633-1595213edefa
 	github.com/snikch/goodman v0.0.0-20171125024755-10e37e294daa
 	github.com/spf13/cobra v1.2.1
+	github.com/spf13/pflag v1.0.5 // indirect
 	github.com/spf13/viper v1.8.1
 	github.com/stretchr/testify v1.7.0
 	github.com/tendermint/tm-db v0.6.4
 	github.com/vektra/mockery/v2 v2.9.0
 	golang.org/x/crypto v0.0.0-20210513164829-c07d793c2f9a
 	golang.org/x/net v0.0.0-20210428140749-89ef3d95e781
+	golang.org/x/sync v0.0.0-20210220032951-036812b2e83c
 	google.golang.org/grpc v1.40.0
 	gopkg.in/check.v1 v1.0.0-20200902074654-038fdea0a05b // indirect
 	pgregory.net/rapid v0.4.7

go.sum

@@ -44,8 +44,8 @@ cloud.google.com/go/storage v1.8.0/go.mod h1:Wv1Oy7z6Yz3DshWRJFhqM/UCfaWIRTdp0RX
 cloud.google.com/go/storage v1.10.0/go.mod h1:FLPqc6j+Ki4BU591ie1oL6qBQGu2Bl/tZ9ullr3+Kg0=
 contrib.go.opencensus.io/exporter/stackdriver v0.13.4/go.mod h1:aXENhDJ1Y4lIg4EUaVTwzvYETVNZk10Pu26tevFKLUc=
 dmitri.shuralyov.com/gpu/mtl v0.0.0-20190408044501-666a987793e9/go.mod h1:H6x//7gZCb22OMCxBHrMx7a5I7Hp++hsVxbQ4BYO7hU=
-github.com/Antonboom/errname v0.1.3 h1:qKV8gSzPzBqrG/q0dgraZXJCymWt6KuD9+Y7K7xtzN8=
-github.com/Antonboom/errname v0.1.3/go.mod h1:jRXo3m0E0EuCnK3wbsSVH3X55Z4iTDLl6ZfCxwFj4TM=
+github.com/Antonboom/errname v0.1.4 h1:lGSlI42Gm4bI1e+IITtXJXvxFM8N7naWimVFKcb0McY=
+github.com/Antonboom/errname v0.1.4/go.mod h1:jRXo3m0E0EuCnK3wbsSVH3X55Z4iTDLl6ZfCxwFj4TM=
 github.com/Azure/go-ansiterm v0.0.0-20170929234023-d6e3b3328b78 h1:w+iIsaOQNcT7OZ575w+acHgRric5iCyQh+xv+KJ4HB8=
 github.com/Azure/go-ansiterm v0.0.0-20170929234023-d6e3b3328b78/go.mod h1:LmzpDX56iTiv29bbRTIsUNlaFfuhWRQBWjQdVyAevI8=
 github.com/BurntSushi/toml v0.3.1/go.mod h1:xHWCNGjB5oqiDr8zfno3MHue2Ht5sIBksp03qcyfWMU=
@@ -63,8 +63,6 @@ github.com/Masterminds/semver v1.5.0 h1:H65muMkzWKEuNDnfl9d70GUjFniHKHRbFPGBuZ3Q
 github.com/Masterminds/semver v1.5.0/go.mod h1:MB6lktGJrhw8PrUyiEoblNEGEQ+RzHPF078ddwwvV3Y=
 github.com/Masterminds/sprig v2.15.0+incompatible/go.mod h1:y6hNFY5UBTIWBxnzTeuNhlNS5hqE0NB0E6fgfo2Br3o=
 github.com/Masterminds/sprig v2.22.0+incompatible/go.mod h1:y6hNFY5UBTIWBxnzTeuNhlNS5hqE0NB0E6fgfo2Br3o=
-github.com/Masterminds/squirrel v1.5.0 h1:JukIZisrUXadA9pl3rMkjhiamxiB0cXiu+HGp/Y8cY8=
-github.com/Masterminds/squirrel v1.5.0/go.mod h1:NNaOrjSoIDfDA40n7sr2tPNZRfjzjA400rg+riTZj10=
 github.com/Microsoft/go-winio v0.4.14 h1:+hMXMk01us9KgxGb7ftKQt2Xpf5hH/yky+TDA+qxleU=
 github.com/Microsoft/go-winio v0.4.14/go.mod h1:qXqCSQ3Xa7+6tgxaGTIe4Kpcdsi+P8jBhyzoq1bpyYA=
 github.com/Nvveen/Gotty v0.0.0-20120604004816-cd527374f1e5 h1:TngWCqHvy9oXAN6lEVMRuU21PR1EtLVZJmdB18Gu3Rw=
@@ -343,8 +341,8 @@ github.com/golangci/go-misc v0.0.0-20180628070357-927a3d87b613 h1:9kfjN3AdxcbsZB
 github.com/golangci/go-misc v0.0.0-20180628070357-927a3d87b613/go.mod h1:SyvUF2NxV+sN8upjjeVYr5W7tyxaT1JVtvhKhOn2ii8=
 github.com/golangci/gofmt v0.0.0-20190930125516-244bba706f1a h1:iR3fYXUjHCR97qWS8ch1y9zPNsgXThGwjKPrYfqMPks=
 github.com/golangci/gofmt v0.0.0-20190930125516-244bba706f1a/go.mod h1:9qCChq59u/eW8im404Q2WWTrnBUQKjpNYKMbU4M7EFU=
-github.com/golangci/golangci-lint v1.42.0 h1:hqf1zo6zY3GKGjjBk3ttdH22tGwF6ZRpk6j6xyJmE8I=
-github.com/golangci/golangci-lint v1.42.0/go.mod h1:wgkGQnU9lOUFvTFo5QBSOvaSSddEV21Z1zYkJSbppZA=
+github.com/golangci/golangci-lint v1.42.1 h1:nC4WyrbdnNdohDVUoNKjy/4N4FTM1gCFaVeXecy6vzM=
+github.com/golangci/golangci-lint v1.42.1/go.mod h1:MuInrVlgg2jq4do6XI1jbkErbVHVbwdrLLtGv6p2wPI=
 github.com/golangci/lint-1 v0.0.0-20191013205115-297bf364a8e0 h1:MfyDlzVjl1hoaPzPD4Gpb/QgoRfSBR0jdhwGyAWwMSA=
 github.com/golangci/lint-1 v0.0.0-20191013205115-297bf364a8e0/go.mod h1:66R6K6P6VWk9I95jvqGxkqJxVWGFy9XlDwLwVz1RCFg=
 github.com/golangci/maligned v0.0.0-20180506175553-b1d89398deca h1:kNY3/svz5T29MYHubXix4aDDuE3RWHkPvopM/EDv/MA=
@@ -548,10 +546,6 @@ github.com/kunwardeep/paralleltest v1.0.2/go.mod h1:ZPqNm1fVHPllh5LPVujzbVz1JN2G
 github.com/kylelemons/godebug v1.1.0/go.mod h1:9/0rRGxNHcop5bhtWyNeEfOS8JIWk580+fNqagV/RAw=
 github.com/kyoh86/exportloopref v0.1.8 h1:5Ry/at+eFdkX9Vsdw3qU4YkvGtzuVfzT4X7S77LoN/M=
 github.com/kyoh86/exportloopref v0.1.8/go.mod h1:1tUcJeiioIs7VWe5gcOObrux3lb66+sBqGZrRkMwPgg=
-github.com/lann/builder v0.0.0-20180802200727-47ae307949d0 h1:SOEGU9fKiNWd/HOJuq6+3iTQz8KNCLtVX6idSoTLdUw=
-github.com/lann/builder v0.0.0-20180802200727-47ae307949d0/go.mod h1:dXGbAdH5GtBTC4WfIxhKZfyBF/HBFgRZSWwZ9g/He9o=
-github.com/lann/ps v0.0.0-20150810152359-62de8c46ede0 h1:P6pPBnrTSX3DEVR4fDembhRWSsG5rVo6hYhAB/ADZrk=
-github.com/lann/ps v0.0.0-20150810152359-62de8c46ede0/go.mod h1:vmVJ0l/dxyfGW6FmdpVm2joNMFikkuWg0EoCKLGUMNw=
 github.com/ldez/gomoddirectives v0.2.2 h1:p9/sXuNFArS2RLc+UpYZSI4KQwGMEDWC/LbtF5OPFVg=
 github.com/ldez/gomoddirectives v0.2.2/go.mod h1:cpgBogWITnCfRq2qGoDkKMEVSaarhdBr6g8G04uz6d0=
 github.com/ldez/tagliatelle v0.2.0 h1:693V8Bf1NdShJ8eu/s84QySA0J2VWBanVBa2WwXD/Wk=
@@ -562,8 +556,9 @@ github.com/lib/pq v1.0.0/go.mod h1:5WUZQaWbwv1U+lTReE5YruASi9Al49XbQIvNi/34Woo=
 github.com/lib/pq v1.2.0/go.mod h1:5WUZQaWbwv1U+lTReE5YruASi9Al49XbQIvNi/34Woo=
 github.com/lib/pq v1.8.0/go.mod h1:AlVN5x4E4T544tWzH6hKfbfQvm3HdbOxrmggDNAPY9o=
 github.com/lib/pq v1.9.0/go.mod h1:AlVN5x4E4T544tWzH6hKfbfQvm3HdbOxrmggDNAPY9o=
-github.com/lib/pq v1.10.2 h1:AqzbZs4ZoCBp+GtejcpCpcxM3zlSMx29dXbUSeVtJb8=
 github.com/lib/pq v1.10.2/go.mod h1:AlVN5x4E4T544tWzH6hKfbfQvm3HdbOxrmggDNAPY9o=
+github.com/lib/pq v1.10.3 h1:v9QZf2Sn6AmjXtQeFpdoq/eaNtYP6IN+7lcrygsIAtg=
+github.com/lib/pq v1.10.3/go.mod h1:AlVN5x4E4T544tWzH6hKfbfQvm3HdbOxrmggDNAPY9o=
 github.com/libp2p/go-buffer-pool v0.0.2 h1:QNK2iAFa8gjAe1SPz6mHSMuCcjs+X1wlHzeOSqcmlfs=
 github.com/libp2p/go-buffer-pool v0.0.2/go.mod h1:MvaB6xw5vOrDl8rYZGLFdKAuk/hRoRZd1Vi32+RXyFM=
 github.com/logrusorgru/aurora v0.0.0-20181002194514-a7b3b318ed4e/go.mod h1:7rIyQOR62GCctdiQpZ/zOJlFyk6y+94wXzv6RNZgaR4=
@@ -601,8 +596,8 @@ github.com/mbilski/exhaustivestruct v1.2.0 h1:wCBmUnSYufAHO6J4AVWY6ff+oxWxsVFrwg
 github.com/mbilski/exhaustivestruct v1.2.0/go.mod h1:OeTBVxQWoEmB2J2JCHmXWPJ0aksxSUOUy+nvtVEfzXc=
 github.com/mgechev/dots v0.0.0-20190921121421-c36f7dcfbb81 h1:QASJXOGm2RZ5Ardbc86qNFvby9AqkLDibfChMtAg5QM=
 github.com/mgechev/dots v0.0.0-20190921121421-c36f7dcfbb81/go.mod h1:KQ7+USdGKfpPjXk4Ga+5XxQM4Lm4e3gAogrreFAYpOg=
-github.com/mgechev/revive v1.1.0 h1:TvabpsolbtlzZTyJcgMRN38MHrgi8C0DhmGE5dhscGY=
-github.com/mgechev/revive v1.1.0/go.mod h1:PKqk4L74K6wVNwY2b6fr+9Qqr/3hIsHVfZCJdbvozrY=
+github.com/mgechev/revive v1.1.1 h1:mkXNHP14Y6tfq+ocnQaiKEtgJDM41yaoyQq4qn6TD/4=
+github.com/mgechev/revive v1.1.1/go.mod h1:PKqk4L74K6wVNwY2b6fr+9Qqr/3hIsHVfZCJdbvozrY=
 github.com/miekg/dns v1.0.14/go.mod h1:W1PPwlIAgtquWBMBEV9nkV9Cazfe8ScdGz/Lj7v3Nrg=
 github.com/miekg/dns v1.1.26/go.mod h1:bPDLeHnStXmXAq1m/Ch/hvfNHr14JKNPMBo3VZKjuso=
 github.com/miekg/dns v1.1.35/go.mod h1:KNUDUusw/aVsxyTYZM1oqvCicbwhgbNgztCETuNZ7xM=
@@ -636,6 +631,8 @@ github.com/moricho/tparallel v0.2.1 h1:95FytivzT6rYzdJLdtfn6m1bfFJylOJK41+lgv/EH
 github.com/moricho/tparallel v0.2.1/go.mod h1:fXEIZxG2vdfl0ZF8b42f5a78EhjjD5mX8qUplsoSU4k=
 github.com/mozilla/scribe v0.0.0-20180711195314-fb71baf557c1/go.mod h1:FIczTrinKo8VaLxe6PWTPEXRXDIHz2QAwiaBaP5/4a8=
 github.com/mozilla/tls-observatory v0.0.0-20210609171429-7bc42856d2e5/go.mod h1:FUqVoUPHSEdDR0MnFM3Dh8AU0pZHLXUD127SAJGER/s=
+github.com/mroth/weightedrand v0.4.1 h1:rHcbUBopmi/3x4nnrvwGJBhX9d0vk+KgoLUZeDP6YyI=
+github.com/mroth/weightedrand v0.4.1/go.mod h1:3p2SIcC8al1YMzGhAIoXD+r9olo/g/cdJgAD905gyNE=
 github.com/mwitkow/go-conntrack v0.0.0-20161129095857-cc309e4a2223/go.mod h1:qRWi+5nqEBWmkhHvq77mSJWrCKwh8bxhgT7d/eI7P4U=
 github.com/mwitkow/go-conntrack v0.0.0-20190716064945-2f068394615f/go.mod h1:qRWi+5nqEBWmkhHvq77mSJWrCKwh8bxhgT7d/eI7P4U=
 github.com/mwitkow/go-proto-validators v0.0.0-20180403085117-0950a7990007/go.mod h1:m2XC9Qq0AlmmVksL6FktJCdTYyLk7V3fKyp0sl1yWQo=
@@ -717,8 +714,8 @@ github.com/pkg/sftp v1.10.1/go.mod h1:lYOWFsE0bwd1+KfKJaKeuokY15vzFx25BLbzYYoAxZ
 github.com/pmezard/go-difflib v0.0.0-20151028094244-d8ed2627bdf0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
 github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZbAQM=
 github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
-github.com/polyfloyd/go-errorlint v0.0.0-20210510181950-ab96adb96fea h1:Sk6Xawg57ZkjXmFYD1xCHSKN6FtYM+km51MM7Lveyyc=
-github.com/polyfloyd/go-errorlint v0.0.0-20210510181950-ab96adb96fea/go.mod h1:wi9BfjxjF/bwiZ701TzmfKu6UKC357IOAtNr0Td0Lvw=
+github.com/polyfloyd/go-errorlint v0.0.0-20210722154253-910bb7978349 h1:Kq/3kL0k033ds3tyez5lFPrfQ74fNJ+OqCclRipubwA=
+github.com/polyfloyd/go-errorlint v0.0.0-20210722154253-910bb7978349/go.mod h1:wi9BfjxjF/bwiZ701TzmfKu6UKC357IOAtNr0Td0Lvw=
 github.com/posener/complete v1.1.1/go.mod h1:em0nMJCgc9GFtwrmVmEMR/ZL6WyhyjMBndrE9hABlRI=
 github.com/posener/complete v1.2.3/go.mod h1:WZIdtGGp+qx0sLrYKtIRAruyNpv6hFCicSgv7Sy7s/s=
 github.com/prometheus/client_golang v0.9.1/go.mod h1:7SWBe2y4D6OKWSNQJUaRYU/AaXPKyh/dDVn+NZz0KFw=
@@ -768,9 +765,10 @@ github.com/rs/cors v1.7.0/go.mod h1:gFx+x8UowdsKA9AchylcLynDq+nNFfI8FkUZdN/jGCU=
 github.com/rs/cors v1.8.0 h1:P2KMzcFwrPoSjkF1WLRPsp3UMLyql8L4v9hQpVeK5so=
 github.com/rs/cors v1.8.0/go.mod h1:EBwu+T5AvHOcXwvZIkQFjUN6s8Czyqw12GL/Y0tUyRM=
 github.com/rs/xid v1.2.1/go.mod h1:+uKXf+4Djp6Md1KODXJxgGQPKngRmWyn10oCKFzNHOQ=
+github.com/rs/xid v1.3.0/go.mod h1:trrq9SKmegXys3aeAKXMUTdJsYXVwGY3RLcfgqegfbg=
 github.com/rs/zerolog v1.18.0/go.mod h1:9nvC1axdVrAHcu/s9taAVfBuIdTZLVQmKQyvrUjF5+I=
-github.com/rs/zerolog v1.23.0 h1:UskrK+saS9P9Y789yNNulYKdARjPZuS35B8gJF2x60g=
-github.com/rs/zerolog v1.23.0/go.mod h1:6c7hFfxPOy7TacJc4Fcdi24/J0NKYGzjG8FWRI916Qo=
+github.com/rs/zerolog v1.24.0 h1:76ivFxmVSRs1u2wUwJVg5VZDYQgeH1JpoS6ndgr9Wy8=
+github.com/rs/zerolog v1.24.0/go.mod h1:7KHcEGe0QZPOm2IE4Kpb5rTh6n1h2hIgS5OOnu1rUaI=
 github.com/russross/blackfriday v1.5.2/go.mod h1:JO/DiYxRf+HjHt06OyowR9PTA263kcR/rfWxYHBV53g=
 github.com/russross/blackfriday/v2 v2.0.1/go.mod h1:+Rmxgy9KzJVeS9/2gXHxylqXiyQDYRxCVz55jmeOWTM=
 github.com/ryancurrah/gomodguard v1.2.3 h1:ww2fsjqocGCAFamzvv/b8IsRduuHHeK2MHTcTxZTQX8=
@@ -865,8 +863,8 @@ github.com/tecbot/gorocksdb v0.0.0-20191217155057-f0fad39f321c h1:g+WoO5jjkqGAzH
 github.com/tecbot/gorocksdb v0.0.0-20191217155057-f0fad39f321c/go.mod h1:ahpPrc7HpcfEWDQRZEmnXMzHY03mLDYMCxeDzy46i+8=
 github.com/tendermint/tm-db v0.6.4 h1:3N2jlnYQkXNQclQwd/eKV/NzlqPlfK21cpRRIx80XXQ=
 github.com/tendermint/tm-db v0.6.4/go.mod h1:dptYhIpJ2M5kUuenLr+Yyf3zQOv1SgBZcl8/BmWlMBw=
-github.com/tetafro/godot v1.4.8 h1:rhuUH+tBrx24yVAr6Ox3/UxcsiUPPJcGhinfLdbdew0=
-github.com/tetafro/godot v1.4.8/go.mod h1:LR3CJpxDVGlYOWn3ZZg1PgNZdTUvzsZWu8xaEohUpn8=
+github.com/tetafro/godot v1.4.9 h1:wsNd0RuUxISqqudFqchsSsMqsM188DoZVPBeKl87tP0=
+github.com/tetafro/godot v1.4.9/go.mod h1:LR3CJpxDVGlYOWn3ZZg1PgNZdTUvzsZWu8xaEohUpn8=
 github.com/timakin/bodyclose v0.0.0-20200424151742-cb6215831a94 h1:ig99OeTyDwQWhPe2iw9lwfQVF1KB3Q4fpP3X7/2VBG8=
 github.com/timakin/bodyclose v0.0.0-20200424151742-cb6215831a94/go.mod h1:Qimiffbc6q9tBWlVV6x0P9sat/ao1xEkREYPPj9hphk=
 github.com/tinylib/msgp v1.1.5/go.mod h1:eQsjooMTnV42mHu917E26IogZ2930nFyBQdofk10Udg=
@@ -1075,6 +1073,7 @@ golang.org/x/sync v0.0.0-20200317015054-43a5402ce75a/go.mod h1:RxMgew5VJxzue5/jJ
 golang.org/x/sync v0.0.0-20200625203802-6e8e738ad208/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
 golang.org/x/sync v0.0.0-20201020160332-67f06af15bc9/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
 golang.org/x/sync v0.0.0-20201207232520-09787c993a3a/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
+golang.org/x/sync v0.0.0-20210220032951-036812b2e83c h1:5KslGYwFpkhGh+Q16bwMP3cOontH8FOep7tGV86Y7SQ=
 golang.org/x/sync v0.0.0-20210220032951-036812b2e83c/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
 golang.org/x/sys v0.0.0-20180823144017-11551d06cbcc/go.mod h1:STP8DvDyc/dI5b8T5hshtkjS+E42TnysNCUPdjciGhY=
 golang.org/x/sys v0.0.0-20180830151530-49385e6e1522/go.mod h1:STP8DvDyc/dI5b8T5hshtkjS+E42TnysNCUPdjciGhY=

inspect/doc.go

@@ -0,0 +1,36 @@
+/*
+Package inspect provides a tool for investigating the state of a
+failed Tendermint node.
+
+This package provides the Inspector type. The Inspector type runs a subset of the Tendermint
+RPC endpoints that are useful for debugging issues with Tendermint consensus.
+
+When a node running the Tendermint consensus engine detects an inconsistent consensus state,
+the entire node will crash. The Tendermint consensus engine cannot run in this
+inconsistent state so the node will not be able to start up again.
+
+The RPC endpoints provided by the Inspector type allow for a node operator to inspect
+the block store and state store to better understand what may have caused the inconsistent state.
+
+
+The Inspector type's lifecycle is controlled by a context.Context
+  ins := inspect.NewFromConfig(rpcConfig)
+  ctx, cancelFunc:= context.WithCancel(context.Background())
+
+  // Run blocks until the Inspector server is shut down.
+  go ins.Run(ctx)
+  ...
+
+  // calling the cancel function will stop the running inspect server
+  cancelFunc()
+
+Inspector serves its RPC endpoints on the address configured in the RPC configuration
+
+  rpcConfig.ListenAddress = "tcp://127.0.0.1:26657"
+  ins := inspect.NewFromConfig(rpcConfig)
+  go ins.Run(ctx)
+
+The list of available RPC endpoints can then be viewed by navigating to
+http://127.0.0.1:26657/ in the web browser.
+*/
+package inspect

inspect/inspect.go

@@ -0,0 +1,149 @@
+package inspect
+
+import (
+	"context"
+	"errors"
+	"fmt"
+	"net"
+
+	"github.com/tendermint/tendermint/config"
+	"github.com/tendermint/tendermint/inspect/rpc"
+	"github.com/tendermint/tendermint/libs/log"
+	tmstrings "github.com/tendermint/tendermint/libs/strings"
+	rpccore "github.com/tendermint/tendermint/rpc/core"
+	"github.com/tendermint/tendermint/state"
+	"github.com/tendermint/tendermint/state/indexer"
+	"github.com/tendermint/tendermint/state/indexer/sink"
+	"github.com/tendermint/tendermint/store"
+	"github.com/tendermint/tendermint/types"
+
+	"golang.org/x/sync/errgroup"
+)
+
+// Inspector manages an RPC service that exports methods to debug a failed node.
+// After a node shuts down due to a consensus failure, it will no longer start
+// up its state cannot easily be inspected. An Inspector value provides a similar interface
+// to the node, using the underlying Tendermint data stores, without bringing up
+// any other components. A caller can query the Inspector service to inspect the
+// persisted state and debug the failure.
+type Inspector struct {
+	routes rpccore.RoutesMap
+
+	config *config.RPCConfig
+
+	indexerService *indexer.Service
+	eventBus       *types.EventBus
+	logger         log.Logger
+}
+
+// New returns an Inspector that serves RPC on the specified BlockStore and StateStore.
+// The Inspector type does not modify the state or block stores.
+// The sinks are used to enable block and transaction querying via the RPC server.
+// The caller is responsible for starting and stopping the Inspector service.
+///
+//nolint:lll
+func New(cfg *config.RPCConfig, bs state.BlockStore, ss state.Store, es []indexer.EventSink, logger log.Logger) *Inspector {
+	routes := rpc.Routes(*cfg, ss, bs, es, logger)
+	eb := types.NewEventBus()
+	eb.SetLogger(logger.With("module", "events"))
+	is := indexer.NewIndexerService(es, eb)
+	is.SetLogger(logger.With("module", "txindex"))
+	return &Inspector{
+		routes:         routes,
+		config:         cfg,
+		logger:         logger,
+		eventBus:       eb,
+		indexerService: is,
+	}
+}
+
+// NewFromConfig constructs an Inspector using the values defined in the passed in config.
+func NewFromConfig(cfg *config.Config) (*Inspector, error) {
+	bsDB, err := config.DefaultDBProvider(&config.DBContext{ID: "blockstore", Config: cfg})
+	if err != nil {
+		return nil, err
+	}
+	bs := store.NewBlockStore(bsDB)
+	sDB, err := config.DefaultDBProvider(&config.DBContext{ID: "state", Config: cfg})
+	if err != nil {
+		return nil, err
+	}
+	genDoc, err := types.GenesisDocFromFile(cfg.GenesisFile())
+	if err != nil {
+		return nil, err
+	}
+	sinks, err := sink.EventSinksFromConfig(cfg, config.DefaultDBProvider, genDoc.ChainID)
+	if err != nil {
+		return nil, err
+	}
+	logger := log.MustNewDefaultLogger(log.LogFormatPlain, log.LogLevelInfo, false)
+	ss := state.NewStore(sDB)
+	return New(cfg.RPC, bs, ss, sinks, logger), nil
+}
+
+// Run starts the Inspector servers and blocks until the servers shut down. The passed
+// in context is used to control the lifecycle of the servers.
+func (ins *Inspector) Run(ctx context.Context) error {
+	err := ins.eventBus.Start()
+	if err != nil {
+		return fmt.Errorf("error starting event bus: %s", err)
+	}
+	defer func() {
+		err := ins.eventBus.Stop()
+		if err != nil {
+			ins.logger.Error("event bus stopped with error", "err", err)
+		}
+	}()
+	err = ins.indexerService.Start()
+	if err != nil {
+		return fmt.Errorf("error starting indexer service: %s", err)
+	}
+	defer func() {
+		err := ins.indexerService.Stop()
+		if err != nil {
+			ins.logger.Error("indexer service stopped with error", "err", err)
+		}
+	}()
+	return startRPCServers(ctx, ins.config, ins.logger, ins.routes)
+}
+
+func startRPCServers(ctx context.Context, cfg *config.RPCConfig, logger log.Logger, routes rpccore.RoutesMap) error {
+	g, tctx := errgroup.WithContext(ctx)
+	listenAddrs := tmstrings.SplitAndTrimEmpty(cfg.ListenAddress, ",", " ")
+	rh := rpc.Handler(cfg, routes, logger)
+	for _, listenerAddr := range listenAddrs {
+		server := rpc.Server{
+			Logger:  logger,
+			Config:  cfg,
+			Handler: rh,
+			Addr:    listenerAddr,
+		}
+		if cfg.IsTLSEnabled() {
+			keyFile := cfg.KeyFile()
+			certFile := cfg.CertFile()
+			listenerAddr := listenerAddr
+			g.Go(func() error {
+				logger.Info("RPC HTTPS server starting", "address", listenerAddr,
+					"certfile", certFile, "keyfile", keyFile)
+				err := server.ListenAndServeTLS(tctx, certFile, keyFile)
+				if !errors.Is(err, net.ErrClosed) {
+					return err
+				}
+				logger.Info("RPC HTTPS server stopped", "address", listenerAddr)
+				return nil
+			})
+		} else {
+			listenerAddr := listenerAddr
+			g.Go(func() error {
+				logger.Info("RPC HTTP server starting", "address", listenerAddr)
+				err := server.ListenAndServe(tctx)
+				if !errors.Is(err, net.ErrClosed) {
+					return err
+				}
+				logger.Info("RPC HTTP server stopped", "address", listenerAddr)
+				return nil
+			})
+		}
+	}
+	return g.Wait()
+}

inspect/inspect_test.go

@@ -0,0 +1,583 @@
+package inspect_test
+
+import (
+	"context"
+	"fmt"
+	"net"
+	"os"
+	"strings"
+	"sync"
+	"testing"
+	"time"
+
+	"github.com/fortytw2/leaktest"
+	"github.com/stretchr/testify/mock"
+	"github.com/stretchr/testify/require"
+	abcitypes "github.com/tendermint/tendermint/abci/types"
+	"github.com/tendermint/tendermint/config"
+	"github.com/tendermint/tendermint/inspect"
+	"github.com/tendermint/tendermint/libs/log"
+	"github.com/tendermint/tendermint/libs/pubsub/query"
+	"github.com/tendermint/tendermint/proto/tendermint/state"
+	httpclient "github.com/tendermint/tendermint/rpc/client/http"
+	"github.com/tendermint/tendermint/state/indexer"
+	indexermocks "github.com/tendermint/tendermint/state/indexer/mocks"
+	statemocks "github.com/tendermint/tendermint/state/mocks"
+	"github.com/tendermint/tendermint/types"
+)
+
+func TestInspectConstructor(t *testing.T) {
+	cfg := config.ResetTestRoot("test")
+	t.Cleanup(leaktest.Check(t))
+	defer func() { _ = os.RemoveAll(cfg.RootDir) }()
+	t.Run("from config", func(t *testing.T) {
+		d, err := inspect.NewFromConfig(cfg)
+		require.NoError(t, err)
+		require.NotNil(t, d)
+	})
+
+}
+
+func TestInspectRun(t *testing.T) {
+	cfg := config.ResetTestRoot("test")
+	t.Cleanup(leaktest.Check(t))
+	defer func() { _ = os.RemoveAll(cfg.RootDir) }()
+	t.Run("from config", func(t *testing.T) {
+		d, err := inspect.NewFromConfig(cfg)
+		require.NoError(t, err)
+		ctx, cancel := context.WithCancel(context.Background())
+		stoppedWG := &sync.WaitGroup{}
+		stoppedWG.Add(1)
+		go func() {
+			require.NoError(t, d.Run(ctx))
+			stoppedWG.Done()
+		}()
+		cancel()
+		stoppedWG.Wait()
+	})
+
+}
+
+func TestBlock(t *testing.T) {
+	testHeight := int64(1)
+	testBlock := new(types.Block)
+	testBlock.Header.Height = testHeight
+	testBlock.Header.LastCommitHash = []byte("test hash")
+	stateStoreMock := &statemocks.Store{}
+
+	blockStoreMock := &statemocks.BlockStore{}
+	blockStoreMock.On("Height").Return(testHeight)
+	blockStoreMock.On("Base").Return(int64(0))
+	blockStoreMock.On("LoadBlockMeta", testHeight).Return(&types.BlockMeta{})
+	blockStoreMock.On("LoadBlock", testHeight).Return(testBlock)
+	eventSinkMock := &indexermocks.EventSink{}
+	eventSinkMock.On("Stop").Return(nil)
+
+	rpcConfig := config.TestRPCConfig()
+	l := log.TestingLogger()
+	d := inspect.New(rpcConfig, blockStoreMock, stateStoreMock, []indexer.EventSink{eventSinkMock}, l)
+	ctx, cancel := context.WithCancel(context.Background())
+	wg := &sync.WaitGroup{}
+	wg.Add(1)
+
+	startedWG := &sync.WaitGroup{}
+	startedWG.Add(1)
+	go func() {
+		startedWG.Done()
+		defer wg.Done()
+		require.NoError(t, d.Run(ctx))
+	}()
+	// FIXME: used to induce context switch.
+	// Determine more deterministic method for prompting a context switch
+	startedWG.Wait()
+	requireConnect(t, rpcConfig.ListenAddress, 20)
+	cli, err := httpclient.New(rpcConfig.ListenAddress)
+	require.NoError(t, err)
+	resultBlock, err := cli.Block(context.Background(), &testHeight)
+	require.NoError(t, err)
+	require.Equal(t, testBlock.Height, resultBlock.Block.Height)
+	require.Equal(t, testBlock.LastCommitHash, resultBlock.Block.LastCommitHash)
+	cancel()
+	wg.Wait()
+
+	blockStoreMock.AssertExpectations(t)
+	stateStoreMock.AssertExpectations(t)
+}
+
+func TestTxSearch(t *testing.T) {
+	testHash := []byte("test")
+	testTx := []byte("tx")
+	testQuery := fmt.Sprintf("tx.hash='%s'", string(testHash))
+	testTxResult := &abcitypes.TxResult{
+		Height: 1,
+		Index:  100,
+		Tx:     testTx,
+	}
+
+	stateStoreMock := &statemocks.Store{}
+	blockStoreMock := &statemocks.BlockStore{}
+	eventSinkMock := &indexermocks.EventSink{}
+	eventSinkMock.On("Stop").Return(nil)
+	eventSinkMock.On("Type").Return(indexer.KV)
+	eventSinkMock.On("SearchTxEvents", mock.Anything,
+		mock.MatchedBy(func(q *query.Query) bool { return testQuery == q.String() })).
+		Return([]*abcitypes.TxResult{testTxResult}, nil)
+
+	rpcConfig := config.TestRPCConfig()
+	l := log.TestingLogger()
+	d := inspect.New(rpcConfig, blockStoreMock, stateStoreMock, []indexer.EventSink{eventSinkMock}, l)
+	ctx, cancel := context.WithCancel(context.Background())
+	wg := &sync.WaitGroup{}
+	wg.Add(1)
+
+	startedWG := &sync.WaitGroup{}
+	startedWG.Add(1)
+	go func() {
+		startedWG.Done()
+		defer wg.Done()
+		require.NoError(t, d.Run(ctx))
+	}()
+	// FIXME: used to induce context switch.
+	// Determine more deterministic method for prompting a context switch
+	startedWG.Wait()
+	requireConnect(t, rpcConfig.ListenAddress, 20)
+	cli, err := httpclient.New(rpcConfig.ListenAddress)
+	require.NoError(t, err)
+
+	var page = 1
+	resultTxSearch, err := cli.TxSearch(context.Background(), testQuery, false, &page, &page, "")
+	require.NoError(t, err)
+	require.Len(t, resultTxSearch.Txs, 1)
+	require.Equal(t, types.Tx(testTx), resultTxSearch.Txs[0].Tx)
+
+	cancel()
+	wg.Wait()
+
+	eventSinkMock.AssertExpectations(t)
+	stateStoreMock.AssertExpectations(t)
+	blockStoreMock.AssertExpectations(t)
+}
+func TestTx(t *testing.T) {
+	testHash := []byte("test")
+	testTx := []byte("tx")
+
+	stateStoreMock := &statemocks.Store{}
+	blockStoreMock := &statemocks.BlockStore{}
+	eventSinkMock := &indexermocks.EventSink{}
+	eventSinkMock.On("Stop").Return(nil)
+	eventSinkMock.On("Type").Return(indexer.KV)
+	eventSinkMock.On("GetTxByHash", testHash).Return(&abcitypes.TxResult{
+		Tx: testTx,
+	}, nil)
+
+	rpcConfig := config.TestRPCConfig()
+	l := log.TestingLogger()
+	d := inspect.New(rpcConfig, blockStoreMock, stateStoreMock, []indexer.EventSink{eventSinkMock}, l)
+	ctx, cancel := context.WithCancel(context.Background())
+	wg := &sync.WaitGroup{}
+	wg.Add(1)
+
+	startedWG := &sync.WaitGroup{}
+	startedWG.Add(1)
+	go func() {
+		startedWG.Done()
+		defer wg.Done()
+		require.NoError(t, d.Run(ctx))
+	}()
+	// FIXME: used to induce context switch.
+	// Determine more deterministic method for prompting a context switch
+	startedWG.Wait()
+	requireConnect(t, rpcConfig.ListenAddress, 20)
+	cli, err := httpclient.New(rpcConfig.ListenAddress)
+	require.NoError(t, err)
+
+	res, err := cli.Tx(context.Background(), testHash, false)
+	require.NoError(t, err)
+	require.Equal(t, types.Tx(testTx), res.Tx)
+
+	cancel()
+	wg.Wait()
+
+	eventSinkMock.AssertExpectations(t)
+	stateStoreMock.AssertExpectations(t)
+	blockStoreMock.AssertExpectations(t)
+}
+func TestConsensusParams(t *testing.T) {
+	testHeight := int64(1)
+	testMaxGas := int64(55)
+	stateStoreMock := &statemocks.Store{}
+	blockStoreMock := &statemocks.BlockStore{}
+	blockStoreMock.On("Height").Return(testHeight)
+	blockStoreMock.On("Base").Return(int64(0))
+	stateStoreMock.On("LoadConsensusParams", testHeight).Return(types.ConsensusParams{
+		Block: types.BlockParams{
+			MaxGas: testMaxGas,
+		},
+	}, nil)
+	eventSinkMock := &indexermocks.EventSink{}
+	eventSinkMock.On("Stop").Return(nil)
+	rpcConfig := config.TestRPCConfig()
+	l := log.TestingLogger()
+	d := inspect.New(rpcConfig, blockStoreMock, stateStoreMock, []indexer.EventSink{eventSinkMock}, l)
+
+	ctx, cancel := context.WithCancel(context.Background())
+	wg := &sync.WaitGroup{}
+	wg.Add(1)
+
+	startedWG := &sync.WaitGroup{}
+	startedWG.Add(1)
+	go func() {
+		startedWG.Done()
+		defer wg.Done()
+		require.NoError(t, d.Run(ctx))
+	}()
+	// FIXME: used to induce context switch.
+	// Determine more deterministic method for prompting a context switch
+	startedWG.Wait()
+	requireConnect(t, rpcConfig.ListenAddress, 20)
+	cli, err := httpclient.New(rpcConfig.ListenAddress)
+	require.NoError(t, err)
+	params, err := cli.ConsensusParams(context.Background(), &testHeight)
+	require.NoError(t, err)
+	require.Equal(t, params.ConsensusParams.Block.MaxGas, testMaxGas)
+
+	cancel()
+	wg.Wait()
+
+	blockStoreMock.AssertExpectations(t)
+	stateStoreMock.AssertExpectations(t)
+}
+
+func TestBlockResults(t *testing.T) {
+	testHeight := int64(1)
+	testGasUsed := int64(100)
+	stateStoreMock := &statemocks.Store{}
+	//	tmstate "github.com/tendermint/tendermint/proto/tendermint/state"
+	stateStoreMock.On("LoadABCIResponses", testHeight).Return(&state.ABCIResponses{
+		DeliverTxs: []*abcitypes.ResponseDeliverTx{
+			{
+				GasUsed: testGasUsed,
+			},
+		},
+		EndBlock:   &abcitypes.ResponseEndBlock{},
+		BeginBlock: &abcitypes.ResponseBeginBlock{},
+	}, nil)
+	blockStoreMock := &statemocks.BlockStore{}
+	blockStoreMock.On("Base").Return(int64(0))
+	blockStoreMock.On("Height").Return(testHeight)
+	eventSinkMock := &indexermocks.EventSink{}
+	eventSinkMock.On("Stop").Return(nil)
+	rpcConfig := config.TestRPCConfig()
+	l := log.TestingLogger()
+	d := inspect.New(rpcConfig, blockStoreMock, stateStoreMock, []indexer.EventSink{eventSinkMock}, l)
+
+	ctx, cancel := context.WithCancel(context.Background())
+	wg := &sync.WaitGroup{}
+	wg.Add(1)
+
+	startedWG := &sync.WaitGroup{}
+	startedWG.Add(1)
+	go func() {
+		startedWG.Done()
+		defer wg.Done()
+		require.NoError(t, d.Run(ctx))
+	}()
+	// FIXME: used to induce context switch.
+	// Determine more deterministic method for prompting a context switch
+	startedWG.Wait()
+	requireConnect(t, rpcConfig.ListenAddress, 20)
+	cli, err := httpclient.New(rpcConfig.ListenAddress)
+	require.NoError(t, err)
+	res, err := cli.BlockResults(context.Background(), &testHeight)
+	require.NoError(t, err)
+	require.Equal(t, res.TotalGasUsed, testGasUsed)
+
+	cancel()
+	wg.Wait()
+
+	blockStoreMock.AssertExpectations(t)
+	stateStoreMock.AssertExpectations(t)
+}
+
+func TestCommit(t *testing.T) {
+	testHeight := int64(1)
+	testRound := int32(101)
+	stateStoreMock := &statemocks.Store{}
+	blockStoreMock := &statemocks.BlockStore{}
+	blockStoreMock.On("Base").Return(int64(0))
+	blockStoreMock.On("Height").Return(testHeight)
+	blockStoreMock.On("LoadBlockMeta", testHeight).Return(&types.BlockMeta{}, nil)
+	blockStoreMock.On("LoadSeenCommit").Return(&types.Commit{
+		Height: testHeight,
+		Round:  testRound,
+	}, nil)
+	eventSinkMock := &indexermocks.EventSink{}
+	eventSinkMock.On("Stop").Return(nil)
+	rpcConfig := config.TestRPCConfig()
+	l := log.TestingLogger()
+	d := inspect.New(rpcConfig, blockStoreMock, stateStoreMock, []indexer.EventSink{eventSinkMock}, l)
+
+	ctx, cancel := context.WithCancel(context.Background())
+	wg := &sync.WaitGroup{}
+	wg.Add(1)
+
+	startedWG := &sync.WaitGroup{}
+	startedWG.Add(1)
+	go func() {
+		startedWG.Done()
+		defer wg.Done()
+		require.NoError(t, d.Run(ctx))
+	}()
+	// FIXME: used to induce context switch.
+	// Determine more deterministic method for prompting a context switch
+	startedWG.Wait()
+	requireConnect(t, rpcConfig.ListenAddress, 20)
+	cli, err := httpclient.New(rpcConfig.ListenAddress)
+	require.NoError(t, err)
+	res, err := cli.Commit(context.Background(), &testHeight)
+	require.NoError(t, err)
+	require.NotNil(t, res)
+	require.Equal(t, res.SignedHeader.Commit.Round, testRound)
+
+	cancel()
+	wg.Wait()
+
+	blockStoreMock.AssertExpectations(t)
+	stateStoreMock.AssertExpectations(t)
+}
+
+func TestBlockByHash(t *testing.T) {
+	testHeight := int64(1)
+	testHash := []byte("test hash")
+	testBlock := new(types.Block)
+	testBlock.Header.Height = testHeight
+	testBlock.Header.LastCommitHash = testHash
+	stateStoreMock := &statemocks.Store{}
+	blockStoreMock := &statemocks.BlockStore{}
+	blockStoreMock.On("LoadBlockMeta", testHeight).Return(&types.BlockMeta{
+		BlockID: types.BlockID{
+			Hash: testHash,
+		},
+		Header: types.Header{
+			Height: testHeight,
+		},
+	}, nil)
+	blockStoreMock.On("LoadBlockByHash", testHash).Return(testBlock, nil)
+	eventSinkMock := &indexermocks.EventSink{}
+	eventSinkMock.On("Stop").Return(nil)
+	rpcConfig := config.TestRPCConfig()
+	l := log.TestingLogger()
+	d := inspect.New(rpcConfig, blockStoreMock, stateStoreMock, []indexer.EventSink{eventSinkMock}, l)
+
+	ctx, cancel := context.WithCancel(context.Background())
+	wg := &sync.WaitGroup{}
+	wg.Add(1)
+
+	startedWG := &sync.WaitGroup{}
+	startedWG.Add(1)
+	go func() {
+		startedWG.Done()
+		defer wg.Done()
+		require.NoError(t, d.Run(ctx))
+	}()
+	// FIXME: used to induce context switch.
+	// Determine more deterministic method for prompting a context switch
+	startedWG.Wait()
+	requireConnect(t, rpcConfig.ListenAddress, 20)
+	cli, err := httpclient.New(rpcConfig.ListenAddress)
+	require.NoError(t, err)
+	res, err := cli.BlockByHash(context.Background(), testHash)
+	require.NoError(t, err)
+	require.NotNil(t, res)
+	require.Equal(t, []byte(res.BlockID.Hash), testHash)
+
+	cancel()
+	wg.Wait()
+
+	blockStoreMock.AssertExpectations(t)
+	stateStoreMock.AssertExpectations(t)
+}
+
+func TestBlockchain(t *testing.T) {
+	testHeight := int64(1)
+	testBlock := new(types.Block)
+	testBlockHash := []byte("test hash")
+	testBlock.Header.Height = testHeight
+	testBlock.Header.LastCommitHash = testBlockHash
+	stateStoreMock := &statemocks.Store{}
+
+	blockStoreMock := &statemocks.BlockStore{}
+	blockStoreMock.On("Height").Return(testHeight)
+	blockStoreMock.On("Base").Return(int64(0))
+	blockStoreMock.On("LoadBlockMeta", testHeight).Return(&types.BlockMeta{
+		BlockID: types.BlockID{
+			Hash: testBlockHash,
+		},
+	})
+	eventSinkMock := &indexermocks.EventSink{}
+	eventSinkMock.On("Stop").Return(nil)
+	rpcConfig := config.TestRPCConfig()
+	l := log.TestingLogger()
+	d := inspect.New(rpcConfig, blockStoreMock, stateStoreMock, []indexer.EventSink{eventSinkMock}, l)
+
+	ctx, cancel := context.WithCancel(context.Background())
+	wg := &sync.WaitGroup{}
+	wg.Add(1)
+
+	startedWG := &sync.WaitGroup{}
+	startedWG.Add(1)
+	go func() {
+		startedWG.Done()
+		defer wg.Done()
+		require.NoError(t, d.Run(ctx))
+	}()
+	// FIXME: used to induce context switch.
+	// Determine more deterministic method for prompting a context switch
+	startedWG.Wait()
+	requireConnect(t, rpcConfig.ListenAddress, 20)
+	cli, err := httpclient.New(rpcConfig.ListenAddress)
+	require.NoError(t, err)
+	res, err := cli.BlockchainInfo(context.Background(), 0, 100)
+	require.NoError(t, err)
+	require.NotNil(t, res)
+	require.Equal(t, testBlockHash, []byte(res.BlockMetas[0].BlockID.Hash))
+
+	cancel()
+	wg.Wait()
+
+	blockStoreMock.AssertExpectations(t)
+	stateStoreMock.AssertExpectations(t)
+}
+
+func TestValidators(t *testing.T) {
+	testHeight := int64(1)
+	testVotingPower := int64(100)
+	testValidators := types.ValidatorSet{
+		Validators: []*types.Validator{
+			{
+				VotingPower: testVotingPower,
+			},
+		},
+	}
+	stateStoreMock := &statemocks.Store{}
+	stateStoreMock.On("LoadValidators", testHeight).Return(&testValidators, nil)
+
+	blockStoreMock := &statemocks.BlockStore{}
+	blockStoreMock.On("Height").Return(testHeight)
+	blockStoreMock.On("Base").Return(int64(0))
+	eventSinkMock := &indexermocks.EventSink{}
+	eventSinkMock.On("Stop").Return(nil)
+	rpcConfig := config.TestRPCConfig()
+	l := log.TestingLogger()
+	d := inspect.New(rpcConfig, blockStoreMock, stateStoreMock, []indexer.EventSink{eventSinkMock}, l)
+
+	ctx, cancel := context.WithCancel(context.Background())
+	wg := &sync.WaitGroup{}
+	wg.Add(1)
+
+	startedWG := &sync.WaitGroup{}
+	startedWG.Add(1)
+	go func() {
+		startedWG.Done()
+		defer wg.Done()
+		require.NoError(t, d.Run(ctx))
+	}()
+	// FIXME: used to induce context switch.
+	// Determine more deterministic method for prompting a context switch
+	startedWG.Wait()
+	requireConnect(t, rpcConfig.ListenAddress, 20)
+	cli, err := httpclient.New(rpcConfig.ListenAddress)
+	require.NoError(t, err)
+
+	testPage := 1
+	testPerPage := 100
+	res, err := cli.Validators(context.Background(), &testHeight, &testPage, &testPerPage)
+	require.NoError(t, err)
+	require.NotNil(t, res)
+	require.Equal(t, testVotingPower, res.Validators[0].VotingPower)
+
+	cancel()
+	wg.Wait()
+
+	blockStoreMock.AssertExpectations(t)
+	stateStoreMock.AssertExpectations(t)
+}
+
+func TestBlockSearch(t *testing.T) {
+	testHeight := int64(1)
+	testBlockHash := []byte("test hash")
+	testQuery := "block.height = 1"
+	stateStoreMock := &statemocks.Store{}
+
+	blockStoreMock := &statemocks.BlockStore{}
+	eventSinkMock := &indexermocks.EventSink{}
+	eventSinkMock.On("Stop").Return(nil)
+	eventSinkMock.On("Type").Return(indexer.KV)
+	blockStoreMock.On("LoadBlock", testHeight).Return(&types.Block{
+		Header: types.Header{
+			Height: testHeight,
+		},
+	}, nil)
+	blockStoreMock.On("LoadBlockMeta", testHeight).Return(&types.BlockMeta{
+		BlockID: types.BlockID{
+			Hash: testBlockHash,
+		},
+	})
+	eventSinkMock.On("SearchBlockEvents", mock.Anything,
+		mock.MatchedBy(func(q *query.Query) bool { return testQuery == q.String() })).
+		Return([]int64{testHeight}, nil)
+	rpcConfig := config.TestRPCConfig()
+	l := log.TestingLogger()
+	d := inspect.New(rpcConfig, blockStoreMock, stateStoreMock, []indexer.EventSink{eventSinkMock}, l)
+
+	ctx, cancel := context.WithCancel(context.Background())
+	wg := &sync.WaitGroup{}
+	wg.Add(1)
+
+	startedWG := &sync.WaitGroup{}
+	startedWG.Add(1)
+	go func() {
+		startedWG.Done()
+		defer wg.Done()
+		require.NoError(t, d.Run(ctx))
+	}()
+	// FIXME: used to induce context switch.
+	// Determine more deterministic method for prompting a context switch
+	startedWG.Wait()
+	requireConnect(t, rpcConfig.ListenAddress, 20)
+	cli, err := httpclient.New(rpcConfig.ListenAddress)
+	require.NoError(t, err)
+
+	testPage := 1
+	testPerPage := 100
+	testOrderBy := "desc"
+	res, err := cli.BlockSearch(context.Background(), testQuery, &testPage, &testPerPage, testOrderBy)
+	require.NoError(t, err)
+	require.NotNil(t, res)
+	require.Equal(t, testBlockHash, []byte(res.Blocks[0].BlockID.Hash))
+
+	cancel()
+	wg.Wait()
+
+	blockStoreMock.AssertExpectations(t)
+	stateStoreMock.AssertExpectations(t)
+}
+
+func requireConnect(t testing.TB, addr string, retries int) {
+	parts := strings.SplitN(addr, "://", 2)
+	if len(parts) != 2 {
+		t.Fatalf("malformed address to dial: %s", addr)
+	}
+	var err error
+	for i := 0; i < retries; i++ {
+		var conn net.Conn
+		conn, err = net.Dial(parts[0], parts[1])
+		if err == nil {
+			conn.Close()
+			return
+		}
+		// FIXME attempt to yield and let the other goroutine continue execution.
+		time.Sleep(time.Microsecond * 100)
+	}
+	t.Fatalf("unable to connect to server %s after %d tries: %s", addr, retries, err)
+}

inspect/rpc/rpc.go

@@ -0,0 +1,143 @@
+package rpc
+
+import (
+	"context"
+	"net/http"
+	"time"
+
+	"github.com/rs/cors"
+
+	"github.com/tendermint/tendermint/config"
+	"github.com/tendermint/tendermint/internal/consensus"
+	"github.com/tendermint/tendermint/libs/log"
+	"github.com/tendermint/tendermint/libs/pubsub"
+	"github.com/tendermint/tendermint/rpc/core"
+	"github.com/tendermint/tendermint/rpc/jsonrpc/server"
+	"github.com/tendermint/tendermint/state"
+	"github.com/tendermint/tendermint/state/indexer"
+	"github.com/tendermint/tendermint/types"
+)
+
+// Server defines parameters for running an Inspector rpc server.
+type Server struct {
+	Addr    string // TCP address to listen on, ":http" if empty
+	Handler http.Handler
+	Logger  log.Logger
+	Config  *config.RPCConfig
+}
+
+// Routes returns the set of routes used by the Inspector server.
+//
+//nolint: lll
+func Routes(cfg config.RPCConfig, s state.Store, bs state.BlockStore, es []indexer.EventSink, logger log.Logger) core.RoutesMap {
+	env := &core.Environment{
+		Config:           cfg,
+		EventSinks:       es,
+		StateStore:       s,
+		BlockStore:       bs,
+		ConsensusReactor: waitSyncCheckerImpl{},
+		Logger:           logger,
+	}
+	return core.RoutesMap{
+		"blockchain":       server.NewRPCFunc(env.BlockchainInfo, "minHeight,maxHeight", true),
+		"consensus_params": server.NewRPCFunc(env.ConsensusParams, "height", true),
+		"block":            server.NewRPCFunc(env.Block, "height", true),
+		"block_by_hash":    server.NewRPCFunc(env.BlockByHash, "hash", true),
+		"block_results":    server.NewRPCFunc(env.BlockResults, "height", true),
+		"commit":           server.NewRPCFunc(env.Commit, "height", true),
+		"validators":       server.NewRPCFunc(env.Validators, "height,page,per_page", true),
+		"tx":               server.NewRPCFunc(env.Tx, "hash,prove", true),
+		"tx_search":        server.NewRPCFunc(env.TxSearch, "query,prove,page,per_page,order_by", false),
+		"block_search":     server.NewRPCFunc(env.BlockSearch, "query,page,per_page,order_by", false),
+	}
+}
+
+// Handler returns the http.Handler configured for use with an Inspector server. Handler
+// registers the routes on the http.Handler and also registers the websocket handler
+// and the CORS handler if specified by the configuration options.
+func Handler(rpcConfig *config.RPCConfig, routes core.RoutesMap, logger log.Logger) http.Handler {
+	mux := http.NewServeMux()
+	wmLogger := logger.With("protocol", "websocket")
+
+	var eventBus types.EventBusSubscriber
+
+	websocketDisconnectFn := func(remoteAddr string) {
+		err := eventBus.UnsubscribeAll(context.Background(), remoteAddr)
+		if err != nil && err != pubsub.ErrSubscriptionNotFound {
+			wmLogger.Error("Failed to unsubscribe addr from events", "addr", remoteAddr, "err", err)
+		}
+	}
+	wm := server.NewWebsocketManager(routes,
+		server.OnDisconnect(websocketDisconnectFn),
+		server.ReadLimit(rpcConfig.MaxBodyBytes))
+	wm.SetLogger(wmLogger)
+	mux.HandleFunc("/websocket", wm.WebsocketHandler)
+
+	server.RegisterRPCFuncs(mux, routes, logger)
+	var rootHandler http.Handler = mux
+	if rpcConfig.IsCorsEnabled() {
+		rootHandler = addCORSHandler(rpcConfig, mux)
+	}
+	return rootHandler
+}
+
+func addCORSHandler(rpcConfig *config.RPCConfig, h http.Handler) http.Handler {
+	corsMiddleware := cors.New(cors.Options{
+		AllowedOrigins: rpcConfig.CORSAllowedOrigins,
+		AllowedMethods: rpcConfig.CORSAllowedMethods,
+		AllowedHeaders: rpcConfig.CORSAllowedHeaders,
+	})
+	h = corsMiddleware.Handler(h)
+	return h
+}
+
+type waitSyncCheckerImpl struct{}
+
+func (waitSyncCheckerImpl) WaitSync() bool {
+	return false
+}
+
+func (waitSyncCheckerImpl) GetPeerState(peerID types.NodeID) (*consensus.PeerState, bool) {
+	return nil, false
+}
+
+// ListenAndServe listens on the address specified in srv.Addr and handles any
+// incoming requests over HTTP using the Inspector rpc handler specified on the server.
+func (srv *Server) ListenAndServe(ctx context.Context) error {
+	listener, err := server.Listen(srv.Addr, srv.Config.MaxOpenConnections)
+	if err != nil {
+		return err
+	}
+	go func() {
+		<-ctx.Done()
+		listener.Close()
+	}()
+	return server.Serve(listener, srv.Handler, srv.Logger, serverRPCConfig(srv.Config))
+}
+
+// ListenAndServeTLS listens on the address specified in srv.Addr. ListenAndServeTLS handles
+// incoming requests over HTTPS using the Inspector rpc handler specified on the server.
+func (srv *Server) ListenAndServeTLS(ctx context.Context, certFile, keyFile string) error {
+	listener, err := server.Listen(srv.Addr, srv.Config.MaxOpenConnections)
+	if err != nil {
+		return err
+	}
+	go func() {
+		<-ctx.Done()
+		listener.Close()
+	}()
+	return server.ServeTLS(listener, srv.Handler, certFile, keyFile, srv.Logger, serverRPCConfig(srv.Config))
+}
+
+func serverRPCConfig(r *config.RPCConfig) *server.Config {
+	cfg := server.DefaultConfig()
+	cfg.MaxBodyBytes = r.MaxBodyBytes
+	cfg.MaxHeaderBytes = r.MaxHeaderBytes
+	// If necessary adjust global WriteTimeout to ensure it's greater than
+	// TimeoutBroadcastTxCommit.
+	// See https://github.com/tendermint/tendermint/issues/3435
+	if cfg.WriteTimeout <= r.TimeoutBroadcastTxCommit {
+		cfg.WriteTimeout = r.TimeoutBroadcastTxCommit + 1*time.Second
+	}
+	return cfg
+}

internal/consensus/reactor.go

@@ -1096,7 +1096,7 @@ func (r *Reactor) handleDataMessage(envelope p2p.Envelope, msgI Message) error {
 	}
 
 	if r.WaitSync() {
-		logger.Info("ignoring message received during sync", "msg", msgI)
+		logger.Info("ignoring message received during sync", "msg", fmt.Sprintf("%T", msgI))
 		return nil
 	}
 

internal/consensus/state.go

@@ -916,8 +916,8 @@ func (cs *State) handleMsg(mi msgInfo) {
 			"height", cs.Height,
 			"round", cs.Round,
 			"peer", peerID,
+			"msg_type", fmt.Sprintf("%T", msg),
 			"err", err,
-			"msg", msg,
 		)
 	}
 }

internal/evidence/pool.go

@@ -516,10 +516,13 @@ func (evpool *Pool) processConsensusBuffer(state sm.State) {
 
 		// Check the height of the conflicting votes and fetch the corresponding time and validator set
 		// to produce the valid evidence
-		var dve *types.DuplicateVoteEvidence
+		var (
+			dve *types.DuplicateVoteEvidence
+			err error
+		)
 		switch {
 		case voteSet.VoteA.Height == state.LastBlockHeight:
-			dve = types.NewDuplicateVoteEvidence(
+			dve, err = types.NewDuplicateVoteEvidence(
 				voteSet.VoteA,
 				voteSet.VoteB,
 				state.LastBlockTime,
@@ -527,8 +530,8 @@ func (evpool *Pool) processConsensusBuffer(state sm.State) {
 			)
 
 		case voteSet.VoteA.Height < state.LastBlockHeight:
-			valSet, err := evpool.stateDB.LoadValidators(voteSet.VoteA.Height)
-			if err != nil {
+			valSet, dbErr := evpool.stateDB.LoadValidators(voteSet.VoteA.Height)
+			if dbErr != nil {
 				evpool.logger.Error("failed to load validator set for conflicting votes",
 					"height", voteSet.VoteA.Height, "err", err)
 				continue
@@ -538,7 +541,7 @@ func (evpool *Pool) processConsensusBuffer(state sm.State) {
 				evpool.logger.Error("failed to load block time for conflicting votes", "height", voteSet.VoteA.Height)
 				continue
 			}
-			dve = types.NewDuplicateVoteEvidence(
+			dve, err = types.NewDuplicateVoteEvidence(
 				voteSet.VoteA,
 				voteSet.VoteB,
 				blockMeta.Header.Time,
@@ -554,6 +557,10 @@ func (evpool *Pool) processConsensusBuffer(state sm.State) {
 				"state.LastBlockHeight", state.LastBlockHeight)
 			continue
 		}
+		if err != nil {
+			evpool.logger.Error("error in generating evidence from votes", "err", err)
+			continue
+		}
 
 		// check if we already have this evidence
 		if evpool.isPending(dve) {
@@ -608,7 +615,7 @@ func prefixToBytes(prefix int64) []byte {
 }
 
 func keyCommitted(evidence types.Evidence) []byte {
-	var height int64 = evidence.Height()
+	height := evidence.Height()
 	key, err := orderedcode.Append(nil, prefixCommitted, height, string(evidence.Hash()))
 	if err != nil {
 		panic(err)
@@ -617,7 +624,7 @@ func keyCommitted(evidence types.Evidence) []byte {
 }
 
 func keyPending(evidence types.Evidence) []byte {
-	var height int64 = evidence.Height()
+	height := evidence.Height()
 	key, err := orderedcode.Append(nil, prefixPending, height, string(evidence.Hash()))
 	if err != nil {
 		panic(err)

internal/evidence/reactor_test.go

@@ -534,12 +534,13 @@ func TestEvidenceListSerialization(t *testing.T) {
 
 	valSet := types.NewValidatorSet([]*types.Validator{val})
 
-	dupl := types.NewDuplicateVoteEvidence(
+	dupl, err := types.NewDuplicateVoteEvidence(
 		exampleVote(1),
 		exampleVote(2),
 		defaultEvidenceTime,
 		valSet,
 	)
+	require.NoError(t, err)
 
 	testCases := map[string]struct {
 		evidenceList []types.Evidence

internal/mempool/v0/clist_mempool.go

@@ -14,7 +14,6 @@ import (
 	"github.com/tendermint/tendermint/internal/mempool"
 	"github.com/tendermint/tendermint/libs/log"
 	tmmath "github.com/tendermint/tendermint/libs/math"
-	pubmempool "github.com/tendermint/tendermint/pkg/mempool"
 	"github.com/tendermint/tendermint/proxy"
 	"github.com/tendermint/tendermint/types"
 )
@@ -217,7 +216,7 @@ func (mem *CListMempool) CheckTx(
 	}
 
 	if txSize > mem.config.MaxTxBytes {
-		return pubmempool.ErrTxTooLarge{
+		return types.ErrTxTooLarge{
 			Max:    mem.config.MaxTxBytes,
 			Actual: txSize,
 		}
@@ -225,7 +224,7 @@ func (mem *CListMempool) CheckTx(
 
 	if mem.preCheck != nil {
 		if err := mem.preCheck(tx); err != nil {
-			return pubmempool.ErrPreCheck{
+			return types.ErrPreCheck{
 				Reason: err,
 			}
 		}
@@ -248,7 +247,7 @@ func (mem *CListMempool) CheckTx(
 			// its non-trivial since invalid txs can become valid,
 			// but they can spam the same tx with little cost to them atm.
 			if loaded {
-				return pubmempool.ErrTxInCache
+				return types.ErrTxInCache
 			}
 		}
 
@@ -364,7 +363,7 @@ func (mem *CListMempool) isFull(txSize int) error {
 	)
 
 	if memSize >= mem.config.Size || int64(txSize)+txsBytes > mem.config.MaxTxsBytes {
-		return pubmempool.ErrMempoolIsFull{
+		return types.ErrMempoolIsFull{
 			NumTxs:      memSize,
 			MaxTxs:      mem.config.Size,
 			TxsBytes:    txsBytes,

internal/mempool/v0/clist_mempool_test.go

@@ -23,7 +23,6 @@ import (
 	"github.com/tendermint/tendermint/libs/log"
 	tmrand "github.com/tendermint/tendermint/libs/rand"
 	"github.com/tendermint/tendermint/libs/service"
-	pubmempool "github.com/tendermint/tendermint/pkg/mempool"
 	"github.com/tendermint/tendermint/proxy"
 	"github.com/tendermint/tendermint/types"
 )
@@ -82,7 +81,7 @@ func checkTxs(t *testing.T, mp mempool.Mempool, count int, peerID uint16) types.
 			// Skip invalid txs.
 			// TestMempoolFilters will fail otherwise. It asserts a number of txs
 			// returned.
-			if pubmempool.IsPreCheckError(err) {
+			if types.IsPreCheckError(err) {
 				continue
 			}
 			t.Fatalf("CheckTx failed: %v while checking #%d tx", err, i)
@@ -455,7 +454,7 @@ func TestMempool_CheckTxChecksTxSize(t *testing.T) {
 		if !testCase.err {
 			require.NoError(t, err, caseString)
 		} else {
-			require.Equal(t, err, pubmempool.ErrTxTooLarge{
+			require.Equal(t, err, types.ErrTxTooLarge{
 				Max:    maxTxSize,
 				Actual: testCase.len,
 			}, caseString)
@@ -503,7 +502,7 @@ func TestMempoolTxsBytes(t *testing.T) {
 
 	err = mp.CheckTx(context.Background(), []byte{0x05}, nil, mempool.TxInfo{})
 	if assert.Error(t, err) {
-		assert.IsType(t, pubmempool.ErrMempoolIsFull{}, err)
+		assert.IsType(t, types.ErrMempoolIsFull{}, err)
 	}
 
 	// 6. zero after tx is rechecked and removed due to not being valid anymore

internal/mempool/v1/mempool.go

@@ -14,7 +14,6 @@ import (
 	"github.com/tendermint/tendermint/internal/mempool"
 	"github.com/tendermint/tendermint/libs/log"
 	tmmath "github.com/tendermint/tendermint/libs/math"
-	pubmempool "github.com/tendermint/tendermint/pkg/mempool"
 	"github.com/tendermint/tendermint/proxy"
 	"github.com/tendermint/tendermint/types"
 )
@@ -239,7 +238,7 @@ func (txmp *TxMempool) CheckTx(
 
 	txSize := len(tx)
 	if txSize > txmp.config.MaxTxBytes {
-		return pubmempool.ErrTxTooLarge{
+		return types.ErrTxTooLarge{
 			Max:    txmp.config.MaxTxBytes,
 			Actual: txSize,
 		}
@@ -247,7 +246,7 @@ func (txmp *TxMempool) CheckTx(
 
 	if txmp.preCheck != nil {
 		if err := txmp.preCheck(tx); err != nil {
-			return pubmempool.ErrPreCheck{
+			return types.ErrPreCheck{
 				Reason: err,
 			}
 		}
@@ -267,7 +266,7 @@ func (txmp *TxMempool) CheckTx(
 		if wtx != nil && ok {
 			// We already have the transaction stored and the we've already seen this
 			// transaction from txInfo.SenderID.
-			return pubmempool.ErrTxInCache
+			return types.ErrTxInCache
 		}
 
 		txmp.logger.Debug("tx exists already in cache", "tx_hash", tx.Hash())
@@ -728,7 +727,7 @@ func (txmp *TxMempool) canAddTx(wtx *WrappedTx) error {
 	)
 
 	if numTxs >= txmp.config.Size || int64(wtx.Size())+sizeBytes > txmp.config.MaxTxsBytes {
-		return pubmempool.ErrMempoolIsFull{
+		return types.ErrMempoolIsFull{
 			NumTxs:      numTxs,
 			MaxTxs:      txmp.config.Size,
 			TxsBytes:    sizeBytes,

internal/p2p/wdrr_queue.go

@@ -16,7 +16,7 @@ type wrappedEnvelope struct {
 	size     uint
 }
 
-// assert the WDDR scheduler implements the queue interface at compile-time
+// assert the WDRR scheduler implements the queue interface at compile-time
 var _ queue = (*wdrrScheduler)(nil)
 
 // wdrrQueue implements a Weighted Deficit Round Robin (WDRR) scheduling

internal/statesync/block_queue_test.go

@@ -274,8 +274,10 @@ loop:
 }
 
 func mockLBResp(t *testing.T, peer types.NodeID, height int64, time time.Time) lightBlockResponse {
+	vals, pv := factory.RandValidatorSet(3, 10)
+	_, _, lb := mockLB(t, height, time, factory.MakeBlockID(), vals, pv)
 	return lightBlockResponse{
-		block: mockLB(t, height, time, factory.MakeBlockID()),
+		block: lb,
 		peer:  peer,
 	}
 }

internal/statesync/dispatcher.go

@@ -5,7 +5,6 @@ import (
 	"errors"
 	"fmt"
 	"sync"
-	"time"
 
 	"github.com/tendermint/tendermint/internal/p2p"
 	"github.com/tendermint/tendermint/light/provider"
@@ -17,169 +16,79 @@ import (
 var (
 	errNoConnectedPeers    = errors.New("no available peers to dispatch request to")
 	errUnsolicitedResponse = errors.New("unsolicited light block response")
-	errNoResponse          = errors.New("peer failed to respond within timeout")
 	errPeerAlreadyBusy     = errors.New("peer is already processing a request")
-	errDisconnected        = errors.New("dispatcher has been disconnected")
+	errDisconnected        = errors.New("dispatcher disconnected")
 )
 
-// dispatcher keeps a list of peers and allows concurrent requests for light
-// blocks. NOTE: It is not the responsibility of the dispatcher to verify the
-// light blocks.
-type dispatcher struct {
-	availablePeers *peerlist
-	requestCh      chan<- p2p.Envelope
-	timeout        time.Duration
+// A Dispatcher multiplexes concurrent requests by multiple peers for light blocks.
+// Only one request per peer can be sent at a time. Subsequent concurrent requests will
+// report an error from the LightBlock method.
+// NOTE: It is not the responsibility of the dispatcher to verify the light blocks.
+type Dispatcher struct {
+	// the channel with which to send light block requests on
+	requestCh chan<- p2p.Envelope
+	closeCh   chan struct{}
 
-	mtx     sync.Mutex
-	calls   map[types.NodeID]chan *types.LightBlock
-	running bool
+	mtx sync.Mutex
+	// all pending calls that have been dispatched and are awaiting an answer
+	calls map[types.NodeID]chan *types.LightBlock
 }
 
-func newDispatcher(requestCh chan<- p2p.Envelope, timeout time.Duration) *dispatcher {
-	return &dispatcher{
-		availablePeers: newPeerList(),
-		timeout:        timeout,
-		requestCh:      requestCh,
-		calls:          make(map[types.NodeID]chan *types.LightBlock),
-		running:        true,
+func NewDispatcher(requestCh chan<- p2p.Envelope) *Dispatcher {
+	return &Dispatcher{
+		requestCh: requestCh,
+		closeCh:   make(chan struct{}),
+		calls:     make(map[types.NodeID]chan *types.LightBlock),
 	}
 }
 
-// LightBlock uses the request channel to fetch a light block from the next peer
-// in a list, tracks the call and waits for the reactor to pass along the response
-func (d *dispatcher) LightBlock(ctx context.Context, height int64) (*types.LightBlock, types.NodeID, error) {
-	d.mtx.Lock()
-	// check to see that the dispatcher is connected to at least one peer
-	if d.availablePeers.Len() == 0 && len(d.calls) == 0 {
-		d.mtx.Unlock()
-		return nil, "", errNoConnectedPeers
-	}
-	d.mtx.Unlock()
-
-	// fetch the next peer id in the list and request a light block from that
-	// peer
-	peer := d.availablePeers.Pop(ctx)
-	lb, err := d.lightBlock(ctx, height, peer)
-	return lb, peer, err
-}
-
-// Providers turns the dispatcher into a set of providers (per peer) which can
-// be used by a light client
-func (d *dispatcher) Providers(chainID string, timeout time.Duration) []provider.Provider {
-	d.mtx.Lock()
-	defer d.mtx.Unlock()
-
-	providers := make([]provider.Provider, d.availablePeers.Len())
-	peers := d.availablePeers.Peers()
-	for index, peer := range peers {
-		providers[index] = &blockProvider{
-			peer:       peer,
-			dispatcher: d,
-			chainID:    chainID,
-			timeout:    timeout,
-		}
-	}
-	return providers
-}
-
-func (d *dispatcher) stop() {
-	d.mtx.Lock()
-	defer d.mtx.Unlock()
-	d.running = false
-	for peer, call := range d.calls {
-		close(call)
-		delete(d.calls, peer)
-	}
-}
-
-func (d *dispatcher) start() {
-	d.mtx.Lock()
-	defer d.mtx.Unlock()
-	d.running = true
-}
-
-func (d *dispatcher) lightBlock(ctx context.Context, height int64, peer types.NodeID) (*types.LightBlock, error) {
+// LightBlock uses the request channel to fetch a light block from a given peer
+// tracking, the call and waiting for the reactor to pass back the response. A nil
+// LightBlock response is used to signal that the peer doesn't have the requested LightBlock.
+func (d *Dispatcher) LightBlock(ctx context.Context, height int64, peer types.NodeID) (*types.LightBlock, error) {
 	// dispatch the request to the peer
 	callCh, err := d.dispatch(peer, height)
 	if err != nil {
 		return nil, err
 	}
 
+	// clean up the call after a response is returned
+	defer func() {
+		d.mtx.Lock()
+		defer d.mtx.Unlock()
+		if call, ok := d.calls[peer]; ok {
+			delete(d.calls, peer)
+			close(call)
+		}
+	}()
+
 	// wait for a response, cancel or timeout
 	select {
 	case resp := <-callCh:
 		return resp, nil
 
 	case <-ctx.Done():
-		d.release(peer)
-		return nil, nil
+		return nil, ctx.Err()
 
-	case <-time.After(d.timeout):
-		d.release(peer)
-		return nil, errNoResponse
-	}
-}
-
-// respond allows the underlying process which receives requests on the
-// requestCh to respond with the respective light block
-func (d *dispatcher) respond(lb *proto.LightBlock, peer types.NodeID) error {
-	d.mtx.Lock()
-	defer d.mtx.Unlock()
-
-	// check that the response came from a request
-	answerCh, ok := d.calls[peer]
-	if !ok {
-		// this can also happen if the response came in after the timeout
-		return errUnsolicitedResponse
-	}
-	// release the peer after returning the response
-	defer d.availablePeers.Append(peer)
-	defer close(answerCh)
-	defer delete(d.calls, peer)
-
-	if lb == nil {
-		answerCh <- nil
-		return nil
-	}
-
-	block, err := types.LightBlockFromProto(lb)
-	if err != nil {
-		fmt.Println("error with converting light block")
-		return err
-	}
-
-	answerCh <- block
-	return nil
-}
-
-func (d *dispatcher) addPeer(peer types.NodeID) {
-	d.availablePeers.Append(peer)
-}
-
-func (d *dispatcher) removePeer(peer types.NodeID) {
-	d.mtx.Lock()
-	defer d.mtx.Unlock()
-	if _, ok := d.calls[peer]; ok {
-		delete(d.calls, peer)
-	} else {
-		d.availablePeers.Remove(peer)
+	case <-d.closeCh:
+		return nil, errDisconnected
 	}
 }
 
 // dispatch takes a peer and allocates it a channel so long as it's not already
 // busy and the receiving channel is still running. It then dispatches the message
-func (d *dispatcher) dispatch(peer types.NodeID, height int64) (chan *types.LightBlock, error) {
+func (d *Dispatcher) dispatch(peer types.NodeID, height int64) (chan *types.LightBlock, error) {
 	d.mtx.Lock()
 	defer d.mtx.Unlock()
-	ch := make(chan *types.LightBlock, 1)
-
-	// check if the dispatcher is running or not
-	if !d.running {
-		close(ch)
-		return ch, errDisconnected
+	select {
+	case <-d.closeCh:
+		return nil, errDisconnected
+	default:
 	}
 
-	// this should happen only if we add the same peer twice (somehow)
+	ch := make(chan *types.LightBlock, 1)
+
+	// check if a request for the same peer has already been made
 	if _, ok := d.calls[peer]; ok {
 		close(ch)
 		return ch, errPeerAlreadyBusy
@@ -193,47 +102,107 @@ func (d *dispatcher) dispatch(peer types.NodeID, height int64) (chan *types.Ligh
 			Height: uint64(height),
 		},
 	}
+
 	return ch, nil
 }
 
-// release appends the peer back to the list and deletes the allocated call so
-// that a new call can be made to that peer
-func (d *dispatcher) release(peer types.NodeID) {
+// Respond allows the underlying process which receives requests on the
+// requestCh to respond with the respective light block. A nil response is used to
+// represent that the receiver of the request does not have a light block at that height.
+func (d *Dispatcher) Respond(lb *proto.LightBlock, peer types.NodeID) error {
 	d.mtx.Lock()
 	defer d.mtx.Unlock()
-	if call, ok := d.calls[peer]; ok {
-		close(call)
-		delete(d.calls, peer)
+
+	// check that the response came from a request
+	answerCh, ok := d.calls[peer]
+	if !ok {
+		// this can also happen if the response came in after the timeout
+		return errUnsolicitedResponse
 	}
-	d.availablePeers.Append(peer)
+
+	// If lb is nil we take that to mean that the peer didn't have the requested light
+	// block and thus pass on the nil to the caller.
+	if lb == nil {
+		answerCh <- nil
+		return nil
+	}
+
+	block, err := types.LightBlockFromProto(lb)
+	if err != nil {
+		return err
+	}
+
+	answerCh <- block
+	return nil
+}
+
+// Close shuts down the dispatcher and cancels any pending calls awaiting responses.
+// Peers awaiting responses that have not arrived are delivered a nil block.
+func (d *Dispatcher) Close() {
+	d.mtx.Lock()
+	defer d.mtx.Unlock()
+	close(d.closeCh)
+	for peer, call := range d.calls {
+		delete(d.calls, peer)
+		close(call)
+	}
+}
+
+func (d *Dispatcher) Done() <-chan struct{} {
+	return d.closeCh
 }
 
 //----------------------------------------------------------------
 
-// blockProvider is a p2p based light provider which uses a dispatcher connected
+// BlockProvider is a p2p based light provider which uses a dispatcher connected
 // to the state sync reactor to serve light blocks to the light client
 //
 // TODO: This should probably be moved over to the light package but as we're
 // not yet officially supporting p2p light clients we'll leave this here for now.
-type blockProvider struct {
+//
+// NOTE: BlockProvider will return an error with concurrent calls. However, we don't
+// need a mutex because a light client (and the backfill process) will never call a
+// method more than once at the same time
+type BlockProvider struct {
 	peer       types.NodeID
 	chainID    string
-	timeout    time.Duration
-	dispatcher *dispatcher
+	dispatcher *Dispatcher
 }
 
-func (p *blockProvider) LightBlock(ctx context.Context, height int64) (*types.LightBlock, error) {
-	// FIXME: The provider doesn't know if the dispatcher is still connected to
-	// that peer. If the connection is dropped for whatever reason the
-	// dispatcher needs to be able to relay this back to the provider so it can
-	// return ErrConnectionClosed instead of ErrNoResponse
-	ctx, cancel := context.WithTimeout(ctx, p.timeout)
-	defer cancel()
-	lb, _ := p.dispatcher.lightBlock(ctx, height, p.peer)
-	if lb == nil {
-		return nil, provider.ErrNoResponse
+// Creates a block provider which implements the light client Provider interface.
+func NewBlockProvider(peer types.NodeID, chainID string, dispatcher *Dispatcher) *BlockProvider {
+	return &BlockProvider{
+		peer:       peer,
+		chainID:    chainID,
+		dispatcher: dispatcher,
+	}
+}
+
+// LightBlock fetches a light block from the peer at a specified height returning either a
+// light block or an appropriate error.
+func (p *BlockProvider) LightBlock(ctx context.Context, height int64) (*types.LightBlock, error) {
+	lb, err := p.dispatcher.LightBlock(ctx, height, p.peer)
+	switch err {
+	case nil:
+		if lb == nil {
+			return nil, provider.ErrLightBlockNotFound
+		}
+	case context.DeadlineExceeded, context.Canceled:
+		return nil, err
+	case errPeerAlreadyBusy:
+		return nil, provider.ErrLightBlockNotFound
+	default:
+		return nil, provider.ErrUnreliableProvider{Reason: err.Error()}
 	}
 
+	// check that the height requested is the same one returned
+	if lb.Height != height {
+		return nil, provider.ErrBadLightBlock{
+			Reason: fmt.Errorf("expected height %d, got height %d", height, lb.Height),
+		}
+	}
+
+	// perform basic validation
 	if err := lb.ValidateBasic(p.chainID); err != nil {
 		return nil, provider.ErrBadLightBlock{Reason: err}
 	}
@@ -245,37 +214,37 @@ func (p *blockProvider) LightBlock(ctx context.Context, height int64) (*types.Li
 // attacks. This is a no op as there currently isn't a way to wire this up to
 // the evidence reactor (we should endeavor to do this in the future but for now
 // it's not critical for backwards verification)
-func (p *blockProvider) ReportEvidence(ctx context.Context, ev types.Evidence) error {
+func (p *BlockProvider) ReportEvidence(ctx context.Context, ev types.Evidence) error {
 	return nil
 }
 
 // String implements stringer interface
-func (p *blockProvider) String() string { return string(p.peer) }
+func (p *BlockProvider) String() string { return string(p.peer) }
 
 //----------------------------------------------------------------
 
 // peerList is a rolling list of peers. This is used to distribute the load of
 // retrieving blocks over all the peers the reactor is connected to
-type peerlist struct {
+type peerList struct {
 	mtx     sync.Mutex
 	peers   []types.NodeID
 	waiting []chan types.NodeID
 }
 
-func newPeerList() *peerlist {
-	return &peerlist{
+func newPeerList() *peerList {
+	return &peerList{
 		peers:   make([]types.NodeID, 0),
 		waiting: make([]chan types.NodeID, 0),
 	}
 }
 
-func (l *peerlist) Len() int {
+func (l *peerList) Len() int {
 	l.mtx.Lock()
 	defer l.mtx.Unlock()
 	return len(l.peers)
 }
 
-func (l *peerlist) Pop(ctx context.Context) types.NodeID {
+func (l *peerList) Pop(ctx context.Context) types.NodeID {
 	l.mtx.Lock()
 	if len(l.peers) == 0 {
 		// if we don't have any peers in the list we block until a peer is
@@ -299,7 +268,7 @@ func (l *peerlist) Pop(ctx context.Context) types.NodeID {
 	return peer
 }
 
-func (l *peerlist) Append(peer types.NodeID) {
+func (l *peerList) Append(peer types.NodeID) {
 	l.mtx.Lock()
 	defer l.mtx.Unlock()
 	if len(l.waiting) > 0 {
@@ -312,7 +281,7 @@ func (l *peerlist) Append(peer types.NodeID) {
 	}
 }
 
-func (l *peerlist) Remove(peer types.NodeID) {
+func (l *peerList) Remove(peer types.NodeID) {
 	l.mtx.Lock()
 	defer l.mtx.Unlock()
 	for i, p := range l.peers {
@@ -323,7 +292,7 @@ func (l *peerlist) Remove(peer types.NodeID) {
 	}
 }
 
-func (l *peerlist) Peers() []types.NodeID {
+func (l *peerList) All() []types.NodeID {
 	l.mtx.Lock()
 	defer l.mtx.Unlock()
 	return l.peers

internal/statesync/dispatcher_test.go

@@ -13,145 +13,102 @@ import (
 	"github.com/stretchr/testify/require"
 
 	"github.com/tendermint/tendermint/internal/p2p"
+	"github.com/tendermint/tendermint/internal/test/factory"
 	ssproto "github.com/tendermint/tendermint/proto/tendermint/statesync"
 	"github.com/tendermint/tendermint/types"
 )
 
 func TestDispatcherBasic(t *testing.T) {
 	t.Cleanup(leaktest.Check(t))
+	const numPeers = 5
 
 	ch := make(chan p2p.Envelope, 100)
 	closeCh := make(chan struct{})
 	defer close(closeCh)
 
-	d := newDispatcher(ch, 1*time.Second)
-
+	d := NewDispatcher(ch)
 	go handleRequests(t, d, ch, closeCh)
 
-	peers := createPeerSet(5)
-	for _, peer := range peers {
-		d.addPeer(peer)
-	}
-
+	peers := createPeerSet(numPeers)
 	wg := sync.WaitGroup{}
 
 	// make a bunch of async requests and require that the correct responses are
 	// given
-	for i := 1; i < 10; i++ {
+	for i := 0; i < numPeers; i++ {
 		wg.Add(1)
 		go func(height int64) {
 			defer wg.Done()
-			lb, peer, err := d.LightBlock(context.Background(), height)
+			lb, err := d.LightBlock(context.Background(), height, peers[height-1])
 			require.NoError(t, err)
 			require.NotNil(t, lb)
 			require.Equal(t, lb.Height, height)
-			require.Contains(t, peers, peer)
-		}(int64(i))
+		}(int64(i + 1))
 	}
 	wg.Wait()
+
+	// assert that all calls were responded to
+	assert.Empty(t, d.calls)
 }
 
 func TestDispatcherReturnsNoBlock(t *testing.T) {
 	t.Cleanup(leaktest.Check(t))
 	ch := make(chan p2p.Envelope, 100)
-	d := newDispatcher(ch, 1*time.Second)
-	peerFromSet := createPeerSet(1)[0]
-	d.addPeer(peerFromSet)
+	d := NewDispatcher(ch)
 	doneCh := make(chan struct{})
+	peer := factory.NodeID("a")
 
 	go func() {
 		<-ch
-		require.NoError(t, d.respond(nil, peerFromSet))
+		require.NoError(t, d.Respond(nil, peer))
 		close(doneCh)
 	}()
 
-	lb, peerResult, err := d.LightBlock(context.Background(), 1)
+	lb, err := d.LightBlock(context.Background(), 1, peer)
 	<-doneCh
 
 	require.Nil(t, lb)
 	require.Nil(t, err)
-	require.Equal(t, peerFromSet, peerResult)
 }
 
-func TestDispatcherErrorsWhenNoPeers(t *testing.T) {
+func TestDispatcherTimeOutWaitingOnLightBlock(t *testing.T) {
 	t.Cleanup(leaktest.Check(t))
 	ch := make(chan p2p.Envelope, 100)
-	d := newDispatcher(ch, 1*time.Second)
+	d := NewDispatcher(ch)
+	peer := factory.NodeID("a")
 
-	lb, peerResult, err := d.LightBlock(context.Background(), 1)
+	ctx, cancelFunc := context.WithTimeout(context.Background(), 10*time.Millisecond)
+	defer cancelFunc()
 
+	lb, err := d.LightBlock(ctx, 1, peer)
+
+	require.Error(t, err)
+	require.Equal(t, context.DeadlineExceeded, err)
 	require.Nil(t, lb)
-	require.Empty(t, peerResult)
-	require.Equal(t, errNoConnectedPeers, err)
-}
-
-func TestDispatcherReturnsBlockOncePeerAvailable(t *testing.T) {
-	t.Cleanup(leaktest.Check(t))
-	dispatcherRequestCh := make(chan p2p.Envelope, 100)
-	d := newDispatcher(dispatcherRequestCh, 1*time.Second)
-	peerFromSet := createPeerSet(1)[0]
-	d.addPeer(peerFromSet)
-	ctx := context.Background()
-	wrapped, cancelFunc := context.WithCancel(ctx)
-
-	doneCh := make(chan struct{})
-	go func() {
-		lb, peerResult, err := d.LightBlock(wrapped, 1)
-		require.Nil(t, lb)
-		require.Equal(t, peerFromSet, peerResult)
-		require.Nil(t, err)
-
-		// calls to dispatcher.Lightblock write into the dispatcher's requestCh.
-		// we read from the requestCh here to unblock the requestCh for future
-		// calls.
-		<-dispatcherRequestCh
-		close(doneCh)
-	}()
-	cancelFunc()
-	<-doneCh
-
-	go func() {
-		<-dispatcherRequestCh
-		lb := &types.LightBlock{}
-		asProto, err := lb.ToProto()
-		require.Nil(t, err)
-		err = d.respond(asProto, peerFromSet)
-		require.Nil(t, err)
-	}()
-
-	lb, peerResult, err := d.LightBlock(context.Background(), 1)
-
-	require.NotNil(t, lb)
-	require.Equal(t, peerFromSet, peerResult)
-	require.Nil(t, err)
 }
 
 func TestDispatcherProviders(t *testing.T) {
 	t.Cleanup(leaktest.Check(t))
 
 	ch := make(chan p2p.Envelope, 100)
-	chainID := "state-sync-test"
+	chainID := "test-chain"
 	closeCh := make(chan struct{})
 	defer close(closeCh)
 
-	d := newDispatcher(ch, 1*time.Second)
-
+	d := NewDispatcher(ch)
 	go handleRequests(t, d, ch, closeCh)
 
 	peers := createPeerSet(5)
-	for _, peer := range peers {
-		d.addPeer(peer)
+	providers := make([]*BlockProvider, len(peers))
+	for idx, peer := range peers {
+		providers[idx] = NewBlockProvider(peer, chainID, d)
 	}
-
-	providers := d.Providers(chainID, 5*time.Second)
 	require.Len(t, providers, 5)
+
 	for i, p := range providers {
-		bp, ok := p.(*blockProvider)
-		require.True(t, ok)
-		assert.Equal(t, bp.String(), string(peers[i]))
+		assert.Equal(t, string(peers[i]), p.String(), i)
 		lb, err := p.LightBlock(context.Background(), 10)
-		assert.Error(t, err)
-		assert.Nil(t, lb)
+		assert.NoError(t, err)
+		assert.NotNil(t, lb)
 	}
 }
 
@@ -166,7 +123,7 @@ func TestPeerListBasic(t *testing.T) {
 		peerList.Append(peer)
 	}
 
-	for idx, peer := range peerList.Peers() {
+	for idx, peer := range peerList.All() {
 		assert.Equal(t, peer, peerSet[idx])
 	}
 
@@ -178,13 +135,22 @@ func TestPeerListBasic(t *testing.T) {
 	}
 	assert.Equal(t, half, peerList.Len())
 
+	// removing a peer that doesn't exist should not change the list
 	peerList.Remove(types.NodeID("lp"))
 	assert.Equal(t, half, peerList.Len())
 
+	// removing a peer that exists should decrease the list size by one
 	peerList.Remove(peerSet[half])
-	half++
-	assert.Equal(t, peerSet[half], peerList.Pop(ctx))
+	assert.Equal(t, numPeers-half-1, peerList.Len())
 
+	// popping the next peer should work as expected
+	assert.Equal(t, peerSet[half+1], peerList.Pop(ctx))
+	assert.Equal(t, numPeers-half-2, peerList.Len())
+
+	// append the two peers back
+	peerList.Append(peerSet[half])
+	peerList.Append(peerSet[half+1])
+	assert.Equal(t, half, peerList.Len())
 }
 
 func TestPeerListBlocksWhenEmpty(t *testing.T) {
@@ -277,9 +243,28 @@ func TestPeerListConcurrent(t *testing.T) {
 	}
 }
 
+func TestPeerListRemove(t *testing.T) {
+	peerList := newPeerList()
+	numPeers := 10
+
+	peerSet := createPeerSet(numPeers)
+	for _, peer := range peerSet {
+		peerList.Append(peer)
+	}
+
+	for _, peer := range peerSet {
+		peerList.Remove(peer)
+		for _, p := range peerList.All() {
+			require.NotEqual(t, p, peer)
+		}
+		numPeers--
+		require.Equal(t, numPeers, peerList.Len())
+	}
+}
+
 // handleRequests is a helper function usually run in a separate go routine to
 // imitate the expected responses of the reactor wired to the dispatcher
-func handleRequests(t *testing.T, d *dispatcher, ch chan p2p.Envelope, closeCh chan struct{}) {
+func handleRequests(t *testing.T, d *Dispatcher, ch chan p2p.Envelope, closeCh chan struct{}) {
 	t.Helper()
 	for {
 		select {
@@ -288,7 +273,7 @@ func handleRequests(t *testing.T, d *dispatcher, ch chan p2p.Envelope, closeCh c
 			peer := request.To
 			resp := mockLBResp(t, peer, int64(height), time.Now())
 			block, _ := resp.block.ToProto()
-			require.NoError(t, d.respond(block, resp.peer))
+			require.NoError(t, d.Respond(block, resp.peer))
 		case <-closeCh:
 			return
 		}

internal/statesync/mock_sync_reactor.go

@@ -1,50 +0,0 @@
-package statesync
-
-import (
-	"context"
-	"time"
-
-	mock "github.com/stretchr/testify/mock"
-	state "github.com/tendermint/tendermint/state"
-)
-
-// MockSyncReactor is an autogenerated mock type for the SyncReactor type.
-// Because of the stateprovider uses in Sync(), we use package statesync instead of mocks.
-type MockSyncReactor struct {
-	mock.Mock
-}
-
-// Backfill provides a mock function with given fields: _a0
-func (_m *MockSyncReactor) Backfill(_a0 state.State) error {
-	ret := _m.Called(_a0)
-
-	var r0 error
-	if rf, ok := ret.Get(0).(func(state.State) error); ok {
-		r0 = rf(_a0)
-	} else {
-		r0 = ret.Error(0)
-	}
-
-	return r0
-}
-
-// Sync provides a mock function with given fields: _a0, _a1, _a2
-func (_m *MockSyncReactor) Sync(_a0 context.Context, _a1 StateProvider, _a2 time.Duration) (state.State, error) {
-	ret := _m.Called(_a0, _a1, _a2)
-
-	var r0 state.State
-	if rf, ok := ret.Get(0).(func(context.Context, StateProvider, time.Duration) state.State); ok {
-		r0 = rf(_a0, _a1, _a2)
-	} else {
-		r0 = ret.Get(0).(state.State)
-	}
-
-	var r1 error
-	if rf, ok := ret.Get(1).(func(context.Context, StateProvider, time.Duration) error); ok {
-		r1 = rf(_a0, _a1, _a2)
-	} else {
-		r1 = ret.Error(1)
-	}
-
-	return r0, r1
-}

internal/statesync/reactor.go

@@ -16,6 +16,8 @@ import (
 	"github.com/tendermint/tendermint/internal/p2p"
 	"github.com/tendermint/tendermint/libs/log"
 	"github.com/tendermint/tendermint/libs/service"
+	"github.com/tendermint/tendermint/light"
+	"github.com/tendermint/tendermint/light/provider"
 	ssproto "github.com/tendermint/tendermint/proto/tendermint/statesync"
 	"github.com/tendermint/tendermint/proxy"
 	sm "github.com/tendermint/tendermint/state"
@@ -61,13 +63,24 @@ var (
 			MsgType: new(ssproto.Message),
 			Descriptor: &p2p.ChannelDescriptor{
 				ID:                  byte(LightBlockChannel),
-				Priority:            2,
+				Priority:            5,
 				SendQueueCapacity:   10,
 				RecvMessageCapacity: lightBlockMsgSize,
 				RecvBufferCapacity:  128,
 				MaxSendBytes:        400,
 			},
 		},
+		ParamsChannel: {
+			MsgType: new(ssproto.Message),
+			Descriptor: &p2p.ChannelDescriptor{
+				ID:                  byte(ParamsChannel),
+				Priority:            2,
+				SendQueueCapacity:   10,
+				RecvMessageCapacity: paramMsgSize,
+				RecvBufferCapacity:  128,
+				MaxSendBytes:        400,
+			},
+		},
 	}
 )
 
@@ -81,6 +94,9 @@ const (
 	// LightBlockChannel exchanges light blocks
 	LightBlockChannel = p2p.ChannelID(0x62)
 
+	// ParamsChannel exchanges consensus params
+	ParamsChannel = p2p.ChannelID(0x63)
+
 	// recentSnapshots is the number of recent snapshots to send and receive per peer.
 	recentSnapshots = 10
 
@@ -91,31 +107,34 @@ const (
 	chunkMsgSize = int(16e6) // ~16MB
 
 	// lightBlockMsgSize is the maximum size of a lightBlockResponseMessage
-	lightBlockMsgSize = int(1e7) // ~10MB
+	lightBlockMsgSize = int(1e7) // ~1MB
+
+	// paramMsgSize is the maximum size of a paramsResponseMessage
+	paramMsgSize = int(1e5) // ~100kb
 
 	// lightBlockResponseTimeout is how long the dispatcher waits for a peer to
 	// return a light block
-	lightBlockResponseTimeout = 30 * time.Second
+	lightBlockResponseTimeout = 10 * time.Second
+
+	// consensusParamsResponseTimeout is the time the p2p state provider waits
+	// before performing a secondary call
+	consensusParamsResponseTimeout = 5 * time.Second
 
 	// maxLightBlockRequestRetries is the amount of retries acceptable before
 	// the backfill process aborts
 	maxLightBlockRequestRetries = 20
 )
 
-// SyncReactor defines an interface used for testing abilities of node.startStateSync.
-type SyncReactor interface {
-	Sync(context.Context, StateProvider, time.Duration) (sm.State, error)
-	Backfill(sm.State) error
-}
-
 // Reactor handles state sync, both restoring snapshots for the local node and
 // serving snapshots for other nodes.
 type Reactor struct {
 	service.BaseService
 
-	cfg        config.StateSyncConfig
-	stateStore sm.Store
-	blockStore *store.BlockStore
+	chainID       string
+	initialHeight int64
+	cfg           config.StateSyncConfig
+	stateStore    sm.Store
+	blockStore    *store.BlockStore
 
 	conn        proxy.AppConnSnapshot
 	connQuery   proxy.AppConnQuery
@@ -123,15 +142,22 @@ type Reactor struct {
 	snapshotCh  *p2p.Channel
 	chunkCh     *p2p.Channel
 	blockCh     *p2p.Channel
+	paramsCh    *p2p.Channel
 	peerUpdates *p2p.PeerUpdates
 	closeCh     chan struct{}
 
-	dispatcher *dispatcher
+	// Dispatcher is used to multiplex light block requests and responses over multiple
+	// peers used by the p2p state provider and in reverse sync.
+	dispatcher *Dispatcher
+	peers      *peerList
 
-	// This will only be set when a state sync is in progress. It is used to feed
-	// received snapshots and chunks into the sync.
-	mtx    tmsync.RWMutex
-	syncer *syncer
+	// These will only be set when a state sync is in progress. It is used to feed
+	// received snapshots and chunks into the syncer and manage incoming and outgoing
+	// providers.
+	mtx           tmsync.RWMutex
+	syncer        *syncer
+	providers     map[types.NodeID]*BlockProvider
+	stateProvider StateProvider
 }
 
 // NewReactor returns a reference to a new state sync reactor, which implements
@@ -139,29 +165,36 @@ type Reactor struct {
 // and querying, references to p2p Channels and a channel to listen for peer
 // updates on. Note, the reactor will close all p2p Channels when stopping.
 func NewReactor(
+	chainID string,
+	initialHeight int64,
 	cfg config.StateSyncConfig,
 	logger log.Logger,
 	conn proxy.AppConnSnapshot,
 	connQuery proxy.AppConnQuery,
-	snapshotCh, chunkCh, blockCh *p2p.Channel,
+	snapshotCh, chunkCh, blockCh, paramsCh *p2p.Channel,
 	peerUpdates *p2p.PeerUpdates,
 	stateStore sm.Store,
 	blockStore *store.BlockStore,
 	tempDir string,
 ) *Reactor {
 	r := &Reactor{
-		cfg:         cfg,
-		conn:        conn,
-		connQuery:   connQuery,
-		snapshotCh:  snapshotCh,
-		chunkCh:     chunkCh,
-		blockCh:     blockCh,
-		peerUpdates: peerUpdates,
-		closeCh:     make(chan struct{}),
-		tempDir:     tempDir,
-		dispatcher:  newDispatcher(blockCh.Out, lightBlockResponseTimeout),
-		stateStore:  stateStore,
-		blockStore:  blockStore,
+		chainID:       chainID,
+		initialHeight: initialHeight,
+		cfg:           cfg,
+		conn:          conn,
+		connQuery:     connQuery,
+		snapshotCh:    snapshotCh,
+		chunkCh:       chunkCh,
+		blockCh:       blockCh,
+		paramsCh:      paramsCh,
+		peerUpdates:   peerUpdates,
+		closeCh:       make(chan struct{}),
+		tempDir:       tempDir,
+		stateStore:    stateStore,
+		blockStore:    blockStore,
+		peers:         newPeerList(),
+		dispatcher:    NewDispatcher(blockCh.Out),
+		providers:     make(map[types.NodeID]*BlockProvider),
 	}
 
 	r.BaseService = *service.NewBaseService(logger, "StateSync", r)
@@ -170,26 +203,20 @@ func NewReactor(
 
 // OnStart starts separate go routines for each p2p Channel and listens for
 // envelopes on each. In addition, it also listens for peer updates and handles
-// messages on that p2p channel accordingly. The caller must be sure to execute
-// OnStop to ensure the outbound p2p Channels are closed. No error is returned.
+// messages on that p2p channel accordingly. Note, we do not launch a go-routine to
+// handle individual envelopes as to not have to deal with bounding workers or pools.
+// The caller must be sure to execute OnStop to ensure the outbound p2p Channels are
+// closed. No error is returned.
 func (r *Reactor) OnStart() error {
-	// Listen for envelopes on the snapshot p2p Channel in a separate go-routine
-	// as to not block or cause IO contention with the chunk p2p Channel. Note,
-	// we do not launch a go-routine to handle individual envelopes as to not
-	// have to deal with bounding workers or pools.
 	go r.processSnapshotCh()
 
-	// Listen for envelopes on the chunk p2p Channel in a separate go-routine
-	// as to not block or cause IO contention with the snapshot p2p Channel. Note,
-	// we do not launch a go-routine to handle individual envelopes as to not
-	// have to deal with bounding workers or pools.
 	go r.processChunkCh()
 
 	go r.processBlockCh()
 
-	go r.processPeerUpdates()
+	go r.processParamsCh()
 
-	r.dispatcher.start()
+	go r.processPeerUpdates()
 
 	return nil
 }
@@ -198,7 +225,9 @@ func (r *Reactor) OnStart() error {
 // blocking until they all exit.
 func (r *Reactor) OnStop() {
 	// tell the dispatcher to stop sending any more requests
-	r.dispatcher.stop()
+	r.dispatcher.Close()
+	// wait for any remaining requests to complete
+	<-r.dispatcher.Done()
 
 	// Close closeCh to signal to all spawned goroutines to gracefully exit. All
 	// p2p Channels should execute Close().
@@ -210,27 +239,27 @@ func (r *Reactor) OnStop() {
 	<-r.snapshotCh.Done()
 	<-r.chunkCh.Done()
 	<-r.blockCh.Done()
+	<-r.paramsCh.Done()
 	<-r.peerUpdates.Done()
 }
 
 // Sync runs a state sync, fetching snapshots and providing chunks to the
-// application. It also saves tendermint state and runs a backfill process to
-// retrieve the necessary amount of headers, commits and validators sets to be
-// able to process evidence and participate in consensus.
-func (r *Reactor) Sync(
-	ctx context.Context,
-	stateProvider StateProvider,
-	discoveryTime time.Duration,
-) (sm.State, error) {
+// application. At the close of the operation, Sync will bootstrap the state
+// store and persist the commit at that height so that either consensus or
+// blocksync can commence. It will then proceed to backfill the necessary amount
+// of historical blocks before participating in consensus
+func (r *Reactor) Sync(ctx context.Context) (sm.State, error) {
+	// We need at least two peers (for cross-referencing of light blocks) before we can
+	// begin state sync
+	r.waitForEnoughPeers(ctx, 2)
 	r.mtx.Lock()
 	if r.syncer != nil {
 		r.mtx.Unlock()
 		return sm.State{}, errors.New("a state sync is already in progress")
 	}
 
-	if stateProvider == nil {
-		r.mtx.Unlock()
-		return sm.State{}, errors.New("the stateProvider should not be nil when doing the state sync")
+	if err := r.initStateProvider(ctx, r.chainID, r.initialHeight); err != nil {
+		return sm.State{}, err
 	}
 
 	r.syncer = newSyncer(
@@ -238,12 +267,19 @@ func (r *Reactor) Sync(
 		r.Logger,
 		r.conn,
 		r.connQuery,
-		stateProvider,
+		r.stateProvider,
 		r.snapshotCh.Out,
 		r.chunkCh.Out,
 		r.tempDir,
 	)
 	r.mtx.Unlock()
+	defer func() {
+		r.mtx.Lock()
+		// reset syncing objects at the close of Sync
+		r.syncer = nil
+		r.stateProvider = nil
+		r.mtx.Unlock()
+	}()
 
 	requestSnapshotsHook := func() {
 		// request snapshots from all currently connected peers
@@ -253,15 +289,11 @@ func (r *Reactor) Sync(
 		}
 	}
 
-	state, commit, err := r.syncer.SyncAny(ctx, discoveryTime, requestSnapshotsHook)
+	state, commit, err := r.syncer.SyncAny(ctx, r.cfg.DiscoveryTime, requestSnapshotsHook)
 	if err != nil {
 		return sm.State{}, err
 	}
 
-	r.mtx.Lock()
-	r.syncer = nil
-	r.mtx.Unlock()
-
 	err = r.stateStore.Bootstrap(state)
 	if err != nil {
 		return sm.State{}, fmt.Errorf("failed to bootstrap node with new state: %w", err)
@@ -272,6 +304,11 @@ func (r *Reactor) Sync(
 		return sm.State{}, fmt.Errorf("failed to store last seen commit: %w", err)
 	}
 
+	err = r.Backfill(ctx, state)
+	if err != nil {
+		r.Logger.Error("backfill failed. Proceeding optimistically...", "err", err)
+	}
+
 	return state, nil
 }
 
@@ -279,7 +316,7 @@ func (r *Reactor) Sync(
 // order. It does not stop verifying blocks until reaching a block with a height
 // and time that is less or equal to the stopHeight and stopTime. The
 // trustedBlockID should be of the header at startHeight.
-func (r *Reactor) Backfill(state sm.State) error {
+func (r *Reactor) Backfill(ctx context.Context, state sm.State) error {
 	params := state.ConsensusParams.Evidence
 	stopHeight := state.LastBlockHeight - params.MaxAgeNumBlocks
 	stopTime := state.LastBlockTime.Add(-params.MaxAgeDuration)
@@ -290,7 +327,7 @@ func (r *Reactor) Backfill(state sm.State) error {
 		stopTime = state.LastBlockTime
 	}
 	return r.backfill(
-		context.Background(),
+		ctx,
 		state.ChainID,
 		state.LastBlockHeight,
 		stopHeight,
@@ -308,12 +345,12 @@ func (r *Reactor) backfill(
 	stopTime time.Time,
 ) error {
 	r.Logger.Info("starting backfill process...", "startHeight", startHeight,
-		"stopHeight", stopHeight, "trustedBlockID", trustedBlockID)
+		"stopHeight", stopHeight, "stopTime", stopTime, "trustedBlockID", trustedBlockID)
 
 	const sleepTime = 1 * time.Second
 	var (
 		lastValidatorSet *types.ValidatorSet
-		lastChangeHeight int64 = startHeight
+		lastChangeHeight = startHeight
 	)
 
 	queue := newBlockQueue(startHeight, stopHeight, initialHeight, stopTime, maxLightBlockRequestRetries)
@@ -330,8 +367,18 @@ func (r *Reactor) backfill(
 			for {
 				select {
 				case height := <-queue.nextHeight():
-					r.Logger.Debug("fetching next block", "height", height)
-					lb, peer, err := r.dispatcher.LightBlock(ctxWithCancel, height)
+					// pop the next peer of the list to send a request to
+					peer := r.peers.Pop(ctx)
+					r.Logger.Debug("fetching next block", "height", height, "peer", peer)
+					subCtx, cancel := context.WithTimeout(ctxWithCancel, lightBlockResponseTimeout)
+					defer cancel()
+					lb, err := func() (*types.LightBlock, error) {
+						defer cancel()
+						// request the light block with a timeout
+						return r.dispatcher.LightBlock(subCtx, height, peer)
+					}()
+					// once the peer has returned a value, add it back to the peer list to be used again
+					r.peers.Append(peer)
 					if errors.Is(err, context.Canceled) {
 						return
 					}
@@ -353,7 +400,7 @@ func (r *Reactor) backfill(
 						queue.retry(height)
 						// As we are fetching blocks backwards, if this node doesn't have the block it likely doesn't
 						// have any prior ones, thus we remove it from the peer list.
-						r.dispatcher.removePeer(peer)
+						r.peers.Remove(peer)
 						continue
 					}
 
@@ -450,12 +497,6 @@ func (r *Reactor) backfill(
 	}
 }
 
-// Dispatcher exposes the dispatcher so that a state provider can use it for
-// light client verification
-func (r *Reactor) Dispatcher() *dispatcher { //nolint:golint
-	return r.dispatcher
-}
-
 // handleSnapshotMessage handles envelopes sent from peers on the
 // SnapshotChannel. It returns an error only if the Envelope.Message is unknown
 // for this channel. This should never be called outside of handleMessage.
@@ -498,7 +539,7 @@ func (r *Reactor) handleSnapshotMessage(envelope p2p.Envelope) error {
 			return nil
 		}
 
-		logger.Debug("received snapshot", "height", msg.Height, "format", msg.Format)
+		logger.Info("received snapshot", "height", msg.Height, "format", msg.Format)
 		_, err := r.syncer.AddSnapshot(envelope.From, &snapshot{
 			Height:   msg.Height,
 			Format:   msg.Format,
@@ -516,6 +557,7 @@ func (r *Reactor) handleSnapshotMessage(envelope p2p.Envelope) error {
 			)
 			return nil
 		}
+		logger.Info("added snapshot", "height", msg.Height, "format", msg.Format)
 
 	default:
 		return fmt.Errorf("received unknown message: %T", msg)
@@ -623,6 +665,15 @@ func (r *Reactor) handleLightBlockMessage(envelope p2p.Envelope) error {
 			r.Logger.Error("failed to retrieve light block", "err", err, "height", msg.Height)
 			return err
 		}
+		if lb == nil {
+			r.blockCh.Out <- p2p.Envelope{
+				To: envelope.From,
+				Message: &ssproto.LightBlockResponse{
+					LightBlock: nil,
+				},
+			}
+			return nil
+		}
 
 		lbproto, err := lb.ToProto()
 		if err != nil {
@@ -640,8 +691,55 @@ func (r *Reactor) handleLightBlockMessage(envelope p2p.Envelope) error {
 		}
 
 	case *ssproto.LightBlockResponse:
-		if err := r.dispatcher.respond(msg.LightBlock, envelope.From); err != nil {
-			r.Logger.Error("error processing light block response", "err", err)
+		var height int64 = 0
+		if msg.LightBlock != nil {
+			height = msg.LightBlock.SignedHeader.Header.Height
+		}
+		r.Logger.Info("received light block response", "peer", envelope.From, "height", height)
+		if err := r.dispatcher.Respond(msg.LightBlock, envelope.From); err != nil {
+			r.Logger.Error("error processing light block response", "err", err, "height", height)
+		}
+
+	default:
+		return fmt.Errorf("received unknown message: %T", msg)
+	}
+
+	return nil
+}
+
+func (r *Reactor) handleParamsMessage(envelope p2p.Envelope) error {
+	switch msg := envelope.Message.(type) {
+	case *ssproto.ParamsRequest:
+		r.Logger.Debug("received consensus params request", "height", msg.Height)
+		cp, err := r.stateStore.LoadConsensusParams(int64(msg.Height))
+		if err != nil {
+			r.Logger.Error("failed to fetch requested consensus params", "err", err, "height", msg.Height)
+			return nil
+		}
+
+		cpproto := cp.ToProto()
+		r.paramsCh.Out <- p2p.Envelope{
+			To: envelope.From,
+			Message: &ssproto.ParamsResponse{
+				Height:          msg.Height,
+				ConsensusParams: cpproto,
+			},
+		}
+
+	case *ssproto.ParamsResponse:
+		r.mtx.RLock()
+		defer r.mtx.RUnlock()
+		r.Logger.Debug("received consensus params response", "height", msg.Height)
+
+		cp := types.ConsensusParamsFromProto(msg.ConsensusParams)
+
+		if sp, ok := r.stateProvider.(*stateProviderP2P); ok {
+			select {
+			case sp.paramsRecvCh <- cp:
+			default:
+			}
+		} else {
+			r.Logger.Debug("received unexpected params response; using RPC state provider", "peer", envelope.From)
 		}
 
 	default:
@@ -678,6 +776,9 @@ func (r *Reactor) handleMessage(chID p2p.ChannelID, envelope p2p.Envelope) (err
 	case LightBlockChannel:
 		err = r.handleLightBlockMessage(envelope)
 
+	case ParamsChannel:
+		err = r.handleParamsMessage(envelope)
+
 	default:
 		err = fmt.Errorf("unknown channel ID (%d) for envelope (%v)", chID, envelope)
 	}
@@ -703,6 +804,10 @@ func (r *Reactor) processBlockCh() {
 	r.processCh(r.blockCh, "light block")
 }
 
+func (r *Reactor) processParamsCh() {
+	r.processCh(r.paramsCh, "consensus params")
+}
+
 // processCh routes state sync messages to their respective handlers. Any error
 // encountered during message execution will result in a PeerError being sent on
 // the respective channel. When the reactor is stopped, we will catch the signal
@@ -732,24 +837,38 @@ func (r *Reactor) processCh(ch *p2p.Channel, chName string) {
 // processPeerUpdate processes a PeerUpdate, returning an error upon failing to
 // handle the PeerUpdate or if a panic is recovered.
 func (r *Reactor) processPeerUpdate(peerUpdate p2p.PeerUpdate) {
-	r.Logger.Debug("received peer update", "peer", peerUpdate.NodeID, "status", peerUpdate.Status)
-
-	r.mtx.RLock()
-	defer r.mtx.RUnlock()
+	r.Logger.Info("received peer update", "peer", peerUpdate.NodeID, "status", peerUpdate.Status)
 
 	switch peerUpdate.Status {
 	case p2p.PeerStatusUp:
-		if r.syncer != nil {
-			r.syncer.AddPeer(peerUpdate.NodeID)
+		r.peers.Append(peerUpdate.NodeID)
+	case p2p.PeerStatusDown:
+		r.peers.Remove(peerUpdate.NodeID)
+	}
+
+	r.mtx.Lock()
+	if r.syncer == nil {
+		r.mtx.Unlock()
+		return
+	}
+	defer r.mtx.Unlock()
+
+	switch peerUpdate.Status {
+	case p2p.PeerStatusUp:
+		newProvider := NewBlockProvider(peerUpdate.NodeID, r.chainID, r.dispatcher)
+		r.providers[peerUpdate.NodeID] = newProvider
+		r.syncer.AddPeer(peerUpdate.NodeID)
+		if sp, ok := r.stateProvider.(*stateProviderP2P); ok {
+			// we do this in a separate routine to not block whilst waiting for the light client to finish
+			// whatever call it's currently executing
+			go sp.addProvider(newProvider)
 		}
-		r.dispatcher.addPeer(peerUpdate.NodeID)
 
 	case p2p.PeerStatusDown:
-		if r.syncer != nil {
-			r.syncer.RemovePeer(peerUpdate.NodeID)
-		}
-		r.dispatcher.removePeer(peerUpdate.NodeID)
+		delete(r.providers, peerUpdate.NodeID)
+		r.syncer.RemovePeer(peerUpdate.NodeID)
 	}
+	r.Logger.Info("processed peer update", "peer", peerUpdate.NodeID, "status", peerUpdate.Status)
 }
 
 // processPeerUpdates initiates a blocking process where we listen for and handle
@@ -839,5 +958,50 @@ func (r *Reactor) fetchLightBlock(height uint64) (*types.LightBlock, error) {
 		},
 		ValidatorSet: vals,
 	}, nil
-
+}
+
+func (r *Reactor) waitForEnoughPeers(ctx context.Context, numPeers int) {
+	t := time.NewTicker(200 * time.Millisecond)
+	defer t.Stop()
+	for {
+		select {
+		case <-ctx.Done():
+			return
+		case <-t.C:
+			if r.peers.Len() >= numPeers {
+				return
+			}
+		}
+	}
+}
+
+func (r *Reactor) initStateProvider(ctx context.Context, chainID string, initialHeight int64) error {
+	var err error
+	to := light.TrustOptions{
+		Period: r.cfg.TrustPeriod,
+		Height: r.cfg.TrustHeight,
+		Hash:   r.cfg.TrustHashBytes(),
+	}
+	spLogger := r.Logger.With("module", "stateprovider")
+	spLogger.Info("initializing state provider", "trustPeriod", to.Period,
+		"trustHeight", to.Height, "useP2P", r.cfg.UseP2P)
+
+	if r.cfg.UseP2P {
+		peers := r.peers.All()
+		providers := make([]provider.Provider, len(peers))
+		for idx, p := range peers {
+			providers[idx] = NewBlockProvider(p, chainID, r.dispatcher)
+		}
+
+		r.stateProvider, err = NewP2PStateProvider(ctx, chainID, initialHeight, providers, to, r.paramsCh.Out, spLogger)
+		if err != nil {
+			return fmt.Errorf("failed to initialize P2P state provider: %w", err)
+		}
+	} else {
+		r.stateProvider, err = NewRPCStateProvider(ctx, chainID, initialHeight, r.cfg.RPCServers, to, spLogger)
+		if err != nil {
+			return fmt.Errorf("failed to initialize RPC state provider: %w", err)
+		}
+	}
+	return nil
 }

internal/statesync/reactor_test.go

@@ -3,6 +3,7 @@ package statesync
 import (
 	"context"
 	"fmt"
+	"strings"
 	"sync"
 	"testing"
 	"time"
@@ -21,6 +22,7 @@ import (
 	"github.com/tendermint/tendermint/light/provider"
 	ssproto "github.com/tendermint/tendermint/proto/tendermint/statesync"
 	tmproto "github.com/tendermint/tendermint/proto/tendermint/types"
+	"github.com/tendermint/tendermint/proxy"
 	proxymocks "github.com/tendermint/tendermint/proxy/mocks"
 	smmocks "github.com/tendermint/tendermint/state/mocks"
 	"github.com/tendermint/tendermint/store"
@@ -50,6 +52,11 @@ type reactorTestSuite struct {
 	blockOutCh     chan p2p.Envelope
 	blockPeerErrCh chan p2p.PeerError
 
+	paramsChannel   *p2p.Channel
+	paramsInCh      chan p2p.Envelope
+	paramsOutCh     chan p2p.Envelope
+	paramsPeerErrCh chan p2p.PeerError
+
 	peerUpdateCh chan p2p.PeerUpdate
 	peerUpdates  *p2p.PeerUpdates
 
@@ -86,6 +93,9 @@ func setup(
 		blockInCh:         make(chan p2p.Envelope, chBuf),
 		blockOutCh:        make(chan p2p.Envelope, chBuf),
 		blockPeerErrCh:    make(chan p2p.PeerError, chBuf),
+		paramsInCh:        make(chan p2p.Envelope, chBuf),
+		paramsOutCh:       make(chan p2p.Envelope, chBuf),
+		paramsPeerErrCh:   make(chan p2p.PeerError, chBuf),
 		conn:              conn,
 		connQuery:         connQuery,
 		stateProvider:     stateProvider,
@@ -118,12 +128,22 @@ func setup(
 		rts.blockPeerErrCh,
 	)
 
+	rts.paramsChannel = p2p.NewChannel(
+		ParamsChannel,
+		new(ssproto.Message),
+		rts.paramsInCh,
+		rts.paramsOutCh,
+		rts.paramsPeerErrCh,
+	)
+
 	rts.stateStore = &smmocks.Store{}
 	rts.blockStore = store.NewBlockStore(dbm.NewMemDB())
 
 	cfg := config.DefaultStateSyncConfig()
 
 	rts.reactor = NewReactor(
+		factory.DefaultTestChainID,
+		1,
 		*cfg,
 		log.TestingLogger(),
 		conn,
@@ -131,15 +151,13 @@ func setup(
 		rts.snapshotChannel,
 		rts.chunkChannel,
 		rts.blockChannel,
+		rts.paramsChannel,
 		rts.peerUpdates,
 		rts.stateStore,
 		rts.blockStore,
 		"",
 	)
 
-	// override the dispatcher with one with a shorter timeout
-	rts.reactor.dispatcher = newDispatcher(rts.blockChannel.Out, 1*time.Second)
-
 	rts.syncer = newSyncer(
 		*cfg,
 		log.NewNopLogger(),
@@ -162,6 +180,58 @@ func setup(
 	return rts
 }
 
+func TestReactor_Sync(t *testing.T) {
+	const snapshotHeight = 7
+	rts := setup(t, nil, nil, nil, 2)
+	chain := buildLightBlockChain(t, 1, 10, time.Now())
+	// app accepts any snapshot
+	rts.conn.On("OfferSnapshotSync", ctx, mock.AnythingOfType("types.RequestOfferSnapshot")).
+		Return(&abci.ResponseOfferSnapshot{Result: abci.ResponseOfferSnapshot_ACCEPT}, nil)
+
+	// app accepts every chunk
+	rts.conn.On("ApplySnapshotChunkSync", ctx, mock.AnythingOfType("types.RequestApplySnapshotChunk")).
+		Return(&abci.ResponseApplySnapshotChunk{Result: abci.ResponseApplySnapshotChunk_ACCEPT}, nil)
+
+	// app query returns valid state app hash
+	rts.connQuery.On("InfoSync", ctx, proxy.RequestInfo).Return(&abci.ResponseInfo{
+		AppVersion:       9,
+		LastBlockHeight:  snapshotHeight,
+		LastBlockAppHash: chain[snapshotHeight+1].AppHash,
+	}, nil)
+
+	// store accepts state and validator sets
+	rts.stateStore.On("Bootstrap", mock.AnythingOfType("state.State")).Return(nil)
+	rts.stateStore.On("SaveValidatorSets", mock.AnythingOfType("int64"), mock.AnythingOfType("int64"),
+		mock.AnythingOfType("*types.ValidatorSet")).Return(nil)
+
+	closeCh := make(chan struct{})
+	defer close(closeCh)
+	go handleLightBlockRequests(t, chain, rts.blockOutCh,
+		rts.blockInCh, closeCh, 0)
+	go graduallyAddPeers(rts.peerUpdateCh, closeCh, 1*time.Second)
+	go handleSnapshotRequests(t, rts.snapshotOutCh, rts.snapshotInCh, closeCh, []snapshot{
+		{
+			Height: uint64(snapshotHeight),
+			Format: 1,
+			Chunks: 1,
+		},
+	})
+
+	go handleChunkRequests(t, rts.chunkOutCh, rts.chunkInCh, closeCh, []byte("abc"))
+
+	go handleConsensusParamsRequest(t, rts.paramsOutCh, rts.paramsInCh, closeCh)
+
+	// update the config to use the p2p provider
+	rts.reactor.cfg.UseP2P = true
+	rts.reactor.cfg.TrustHeight = 1
+	rts.reactor.cfg.TrustHash = fmt.Sprintf("%X", chain[1].Hash())
+	rts.reactor.cfg.DiscoveryTime = 1 * time.Second
+
+	// Run state sync
+	_, err := rts.reactor.Sync(context.Background())
+	require.NoError(t, err)
+}
+
 func TestReactor_ChunkRequest_InvalidRequest(t *testing.T) {
 	rts := setup(t, nil, nil, nil, 2)
 
@@ -370,7 +440,7 @@ func TestReactor_LightBlockResponse(t *testing.T) {
 	}
 }
 
-func TestReactor_Dispatcher(t *testing.T) {
+func TestReactor_BlockProviders(t *testing.T) {
 	rts := setup(t, nil, nil, nil, 2)
 	rts.peerUpdateCh <- p2p.PeerUpdate{
 		NodeID: types.NodeID("aa"),
@@ -387,9 +457,13 @@ func TestReactor_Dispatcher(t *testing.T) {
 	chain := buildLightBlockChain(t, 1, 10, time.Now())
 	go handleLightBlockRequests(t, chain, rts.blockOutCh, rts.blockInCh, closeCh, 0)
 
-	dispatcher := rts.reactor.Dispatcher()
-	providers := dispatcher.Providers(factory.DefaultTestChainID, 5*time.Second)
-	require.Len(t, providers, 2)
+	peers := rts.reactor.peers.All()
+	require.Len(t, peers, 2)
+
+	providers := make([]provider.Provider, len(peers))
+	for idx, peer := range peers {
+		providers[idx] = NewBlockProvider(peer, factory.DefaultTestChainID, rts.reactor.dispatcher)
+	}
 
 	wg := sync.WaitGroup{}
 
@@ -416,6 +490,59 @@ func TestReactor_Dispatcher(t *testing.T) {
 		t.Fail()
 	case <-ctx.Done():
 	}
+
+}
+
+func TestReactor_StateProviderP2P(t *testing.T) {
+	rts := setup(t, nil, nil, nil, 2)
+	// make syncer non nil else test won't think we are state syncing
+	rts.reactor.syncer = rts.syncer
+	peerA := types.NodeID(strings.Repeat("a", 2*types.NodeIDByteLength))
+	peerB := types.NodeID(strings.Repeat("b", 2*types.NodeIDByteLength))
+	rts.peerUpdateCh <- p2p.PeerUpdate{
+		NodeID: peerA,
+		Status: p2p.PeerStatusUp,
+	}
+	rts.peerUpdateCh <- p2p.PeerUpdate{
+		NodeID: peerB,
+		Status: p2p.PeerStatusUp,
+	}
+
+	closeCh := make(chan struct{})
+	defer close(closeCh)
+
+	chain := buildLightBlockChain(t, 1, 10, time.Now())
+	go handleLightBlockRequests(t, chain, rts.blockOutCh, rts.blockInCh, closeCh, 0)
+	go handleConsensusParamsRequest(t, rts.paramsOutCh, rts.paramsInCh, closeCh)
+
+	rts.reactor.cfg.UseP2P = true
+	rts.reactor.cfg.TrustHeight = 1
+	rts.reactor.cfg.TrustHash = fmt.Sprintf("%X", chain[1].Hash())
+	ctx := context.Background()
+	rts.reactor.mtx.Lock()
+	err := rts.reactor.initStateProvider(ctx, factory.DefaultTestChainID, 1)
+	rts.reactor.mtx.Unlock()
+	require.NoError(t, err)
+	rts.reactor.syncer.stateProvider = rts.reactor.stateProvider
+
+	appHash, err := rts.reactor.stateProvider.AppHash(ctx, 5)
+	require.NoError(t, err)
+	require.Len(t, appHash, 32)
+
+	state, err := rts.reactor.stateProvider.State(ctx, 5)
+	require.NoError(t, err)
+	require.Equal(t, appHash, state.AppHash)
+	require.Equal(t, types.DefaultConsensusParams(), &state.ConsensusParams)
+
+	commit, err := rts.reactor.stateProvider.Commit(ctx, 5)
+	require.NoError(t, err)
+	require.Equal(t, commit.BlockID, state.LastBlockID)
+
+	added, err := rts.reactor.syncer.AddSnapshot(peerA, &snapshot{
+		Height: 1, Format: 2, Chunks: 7, Hash: []byte{1, 2}, Metadata: []byte{1},
+	})
+	require.NoError(t, err)
+	require.True(t, added)
 }
 
 func TestReactor_Backfill(t *testing.T) {
@@ -494,7 +621,6 @@ func retryUntil(t *testing.T, fn func() bool, timeout time.Duration) {
 		if fn() {
 			return
 		}
-
 		require.NoError(t, ctx.Err())
 	}
 }
@@ -523,7 +649,9 @@ func handleLightBlockRequests(t *testing.T,
 				} else {
 					switch errorCount % 3 {
 					case 0: // send a different block
-						differntLB, err := mockLB(t, int64(msg.Height), factory.DefaultTestTime, factory.MakeBlockID()).ToProto()
+						vals, pv := factory.RandValidatorSet(3, 10)
+						_, _, lb := mockLB(t, int64(msg.Height), factory.DefaultTestTime, factory.MakeBlockID(), vals, pv)
+						differntLB, err := lb.ToProto()
 						require.NoError(t, err)
 						sending <- p2p.Envelope{
 							From: envelope.To,
@@ -550,37 +678,147 @@ func handleLightBlockRequests(t *testing.T,
 	}
 }
 
+func handleConsensusParamsRequest(t *testing.T, receiving, sending chan p2p.Envelope, closeCh chan struct{}) {
+	t.Helper()
+	params := types.DefaultConsensusParams()
+	paramsProto := params.ToProto()
+	for {
+		select {
+		case envelope := <-receiving:
+			t.Log("received consensus params request")
+			msg, ok := envelope.Message.(*ssproto.ParamsRequest)
+			require.True(t, ok)
+			sending <- p2p.Envelope{
+				From: envelope.To,
+				Message: &ssproto.ParamsResponse{
+					Height:          msg.Height,
+					ConsensusParams: paramsProto,
+				},
+			}
+
+		case <-closeCh:
+			return
+		}
+	}
+}
+
 func buildLightBlockChain(t *testing.T, fromHeight, toHeight int64, startTime time.Time) map[int64]*types.LightBlock {
 	chain := make(map[int64]*types.LightBlock, toHeight-fromHeight)
 	lastBlockID := factory.MakeBlockID()
-	blockTime := startTime.Add(-5 * time.Minute)
+	blockTime := startTime.Add(time.Duration(fromHeight-toHeight) * time.Minute)
+	vals, pv := factory.RandValidatorSet(3, 10)
 	for height := fromHeight; height < toHeight; height++ {
-		chain[height] = mockLB(t, height, blockTime, lastBlockID)
+		vals, pv, chain[height] = mockLB(t, height, blockTime, lastBlockID, vals, pv)
 		lastBlockID = factory.MakeBlockIDWithHash(chain[height].Header.Hash())
 		blockTime = blockTime.Add(1 * time.Minute)
 	}
 	return chain
 }
 
-func mockLB(t *testing.T, height int64, time time.Time,
-	lastBlockID types.BlockID) *types.LightBlock {
+func mockLB(t *testing.T, height int64, time time.Time, lastBlockID types.BlockID,
+	currentVals *types.ValidatorSet, currentPrivVals []types.PrivValidator,
+) (*types.ValidatorSet, []types.PrivValidator, *types.LightBlock) {
 	header, err := factory.MakeHeader(&types.Header{
 		Height:      height,
 		LastBlockID: lastBlockID,
 		Time:        time,
 	})
 	require.NoError(t, err)
-	vals, pv := factory.RandValidatorSet(3, 10)
-	header.ValidatorsHash = vals.Hash()
+	nextVals, nextPrivVals := factory.RandValidatorSet(3, 10)
+	header.ValidatorsHash = currentVals.Hash()
+	header.NextValidatorsHash = nextVals.Hash()
+	header.ConsensusHash = types.DefaultConsensusParams().HashConsensusParams()
 	lastBlockID = factory.MakeBlockIDWithHash(header.Hash())
-	voteSet := types.NewVoteSet(factory.DefaultTestChainID, height, 0, tmproto.PrecommitType, vals)
-	commit, err := factory.MakeCommit(lastBlockID, height, 0, voteSet, pv, time)
+	voteSet := types.NewVoteSet(factory.DefaultTestChainID, height, 0, tmproto.PrecommitType, currentVals)
+	commit, err := factory.MakeCommit(lastBlockID, height, 0, voteSet, currentPrivVals, time)
 	require.NoError(t, err)
-	return &types.LightBlock{
+	return nextVals, nextPrivVals, &types.LightBlock{
 		SignedHeader: &types.SignedHeader{
 			Header: header,
 			Commit: commit,
 		},
-		ValidatorSet: vals,
+		ValidatorSet: currentVals,
+	}
+}
+
+// graduallyAddPeers delivers a new randomly-generated peer update on peerUpdateCh once
+// per interval, until closeCh is closed. Each peer update is assigned a random node ID.
+func graduallyAddPeers(
+	peerUpdateCh chan p2p.PeerUpdate,
+	closeCh chan struct{},
+	interval time.Duration,
+) {
+	ticker := time.NewTicker(interval)
+	for {
+		select {
+		case <-ticker.C:
+			peerUpdateCh <- p2p.PeerUpdate{
+				NodeID: factory.RandomNodeID(),
+				Status: p2p.PeerStatusUp,
+			}
+		case <-closeCh:
+			return
+		}
+	}
+}
+
+func handleSnapshotRequests(
+	t *testing.T,
+	receivingCh chan p2p.Envelope,
+	sendingCh chan p2p.Envelope,
+	closeCh chan struct{},
+	snapshots []snapshot,
+) {
+	t.Helper()
+	for {
+		select {
+		case envelope := <-receivingCh:
+			_, ok := envelope.Message.(*ssproto.SnapshotsRequest)
+			require.True(t, ok)
+			for _, snapshot := range snapshots {
+				sendingCh <- p2p.Envelope{
+					From: envelope.To,
+					Message: &ssproto.SnapshotsResponse{
+						Height:   snapshot.Height,
+						Format:   snapshot.Format,
+						Chunks:   snapshot.Chunks,
+						Hash:     snapshot.Hash,
+						Metadata: snapshot.Metadata,
+					},
+				}
+			}
+		case <-closeCh:
+			return
+		}
+	}
+}
+
+func handleChunkRequests(
+	t *testing.T,
+	receivingCh chan p2p.Envelope,
+	sendingCh chan p2p.Envelope,
+	closeCh chan struct{},
+	chunk []byte,
+) {
+	t.Helper()
+	for {
+		select {
+		case envelope := <-receivingCh:
+			msg, ok := envelope.Message.(*ssproto.ChunkRequest)
+			require.True(t, ok)
+			sendingCh <- p2p.Envelope{
+				From: envelope.To,
+				Message: &ssproto.ChunkResponse{
+					Height:  msg.Height,
+					Format:  msg.Format,
+					Index:   msg.Index,
+					Chunk:   chunk,
+					Missing: false,
+				},
+			}
+
+		case <-closeCh:
+			return
+		}
 	}
 }

internal/statesync/snapshots.go

@@ -1,13 +1,11 @@
 package statesync
 
 import (
-	"context"
 	"crypto/sha256"
 	"fmt"
 	"math/rand"
 	"sort"
 	"strings"
-	"time"
 
 	tmsync "github.com/tendermint/tendermint/internal/libs/sync"
 	"github.com/tendermint/tendermint/types"
@@ -43,8 +41,6 @@ func (s *snapshot) Key() snapshotKey {
 
 // snapshotPool discovers and aggregates snapshots across peers.
 type snapshotPool struct {
-	stateProvider StateProvider
-
 	tmsync.Mutex
 	snapshots     map[snapshotKey]*snapshot
 	snapshotPeers map[snapshotKey]map[types.NodeID]types.NodeID
@@ -60,10 +56,9 @@ type snapshotPool struct {
 	snapshotBlacklist map[snapshotKey]bool
 }
 
-// newSnapshotPool creates a new snapshot pool. The state source is used for
-func newSnapshotPool(stateProvider StateProvider) *snapshotPool {
+// newSnapshotPool creates a new empty snapshot pool.
+func newSnapshotPool() *snapshotPool {
 	return &snapshotPool{
-		stateProvider:     stateProvider,
 		snapshots:         make(map[snapshotKey]*snapshot),
 		snapshotPeers:     make(map[snapshotKey]map[types.NodeID]types.NodeID),
 		formatIndex:       make(map[uint32]map[snapshotKey]bool),
@@ -80,14 +75,6 @@ func newSnapshotPool(stateProvider StateProvider) *snapshotPool {
 // snapshot height is verified using the light client, and the expected app hash
 // is set for the snapshot.
 func (p *snapshotPool) Add(peerID types.NodeID, snapshot *snapshot) (bool, error) {
-	ctx, cancel := context.WithTimeout(context.TODO(), 30*time.Second)
-	defer cancel()
-
-	appHash, err := p.stateProvider.AppHash(ctx, snapshot.Height)
-	if err != nil {
-		return false, fmt.Errorf("failed to get app hash: %w", err)
-	}
-	snapshot.trustedAppHash = appHash
 	key := snapshot.Key()
 
 	p.Lock()

internal/statesync/snapshots_test.go

@@ -3,10 +3,8 @@ package statesync
 import (
 	"testing"
 
-	"github.com/stretchr/testify/mock"
 	"github.com/stretchr/testify/require"
 
-	"github.com/tendermint/tendermint/internal/statesync/mocks"
 	"github.com/tendermint/tendermint/types"
 )
 
@@ -39,13 +37,10 @@ func TestSnapshot_Key(t *testing.T) {
 }
 
 func TestSnapshotPool_Add(t *testing.T) {
-	stateProvider := &mocks.StateProvider{}
-	stateProvider.On("AppHash", mock.Anything, uint64(1)).Return([]byte("app_hash"), nil)
-
 	peerID := types.NodeID("aa")
 
 	// Adding to the pool should work
-	pool := newSnapshotPool(stateProvider)
+	pool := newSnapshotPool()
 	added, err := pool.Add(peerID, &snapshot{
 		Height: 1,
 		Format: 1,
@@ -66,18 +61,12 @@ func TestSnapshotPool_Add(t *testing.T) {
 	require.NoError(t, err)
 	require.False(t, added)
 
-	// The pool should have populated the snapshot with the trusted app hash
 	snapshot := pool.Best()
 	require.NotNil(t, snapshot)
-	require.Equal(t, []byte("app_hash"), snapshot.trustedAppHash)
-
-	stateProvider.AssertExpectations(t)
 }
 
 func TestSnapshotPool_GetPeer(t *testing.T) {
-	stateProvider := &mocks.StateProvider{}
-	stateProvider.On("AppHash", mock.Anything, mock.Anything).Return([]byte("app_hash"), nil)
-	pool := newSnapshotPool(stateProvider)
+	pool := newSnapshotPool()
 
 	s := &snapshot{Height: 1, Format: 1, Chunks: 1, Hash: []byte{1}}
 
@@ -112,9 +101,7 @@ func TestSnapshotPool_GetPeer(t *testing.T) {
 }
 
 func TestSnapshotPool_GetPeers(t *testing.T) {
-	stateProvider := &mocks.StateProvider{}
-	stateProvider.On("AppHash", mock.Anything, mock.Anything).Return([]byte("app_hash"), nil)
-	pool := newSnapshotPool(stateProvider)
+	pool := newSnapshotPool()
 
 	s := &snapshot{Height: 1, Format: 1, Chunks: 1, Hash: []byte{1}}
 
@@ -137,9 +124,7 @@ func TestSnapshotPool_GetPeers(t *testing.T) {
 }
 
 func TestSnapshotPool_Ranked_Best(t *testing.T) {
-	stateProvider := &mocks.StateProvider{}
-	stateProvider.On("AppHash", mock.Anything, mock.Anything).Return([]byte("app_hash"), nil)
-	pool := newSnapshotPool(stateProvider)
+	pool := newSnapshotPool()
 
 	// snapshots in expected order (best to worst). Highest height wins, then highest format.
 	// Snapshots with different chunk hashes are considered different, and the most peers is
@@ -182,9 +167,7 @@ func TestSnapshotPool_Ranked_Best(t *testing.T) {
 }
 
 func TestSnapshotPool_Reject(t *testing.T) {
-	stateProvider := &mocks.StateProvider{}
-	stateProvider.On("AppHash", mock.Anything, mock.Anything).Return([]byte("app_hash"), nil)
-	pool := newSnapshotPool(stateProvider)
+	pool := newSnapshotPool()
 
 	peerID := types.NodeID("aa")
 
@@ -212,9 +195,7 @@ func TestSnapshotPool_Reject(t *testing.T) {
 }
 
 func TestSnapshotPool_RejectFormat(t *testing.T) {
-	stateProvider := &mocks.StateProvider{}
-	stateProvider.On("AppHash", mock.Anything, mock.Anything).Return([]byte("app_hash"), nil)
-	pool := newSnapshotPool(stateProvider)
+	pool := newSnapshotPool()
 
 	peerID := types.NodeID("aa")
 
@@ -243,9 +224,7 @@ func TestSnapshotPool_RejectFormat(t *testing.T) {
 }
 
 func TestSnapshotPool_RejectPeer(t *testing.T) {
-	stateProvider := &mocks.StateProvider{}
-	stateProvider.On("AppHash", mock.Anything, mock.Anything).Return([]byte("app_hash"), nil)
-	pool := newSnapshotPool(stateProvider)
+	pool := newSnapshotPool()
 
 	peerAID := types.NodeID("aa")
 	peerBID := types.NodeID("bb")
@@ -285,9 +264,7 @@ func TestSnapshotPool_RejectPeer(t *testing.T) {
 }
 
 func TestSnapshotPool_RemovePeer(t *testing.T) {
-	stateProvider := &mocks.StateProvider{}
-	stateProvider.On("AppHash", mock.Anything, mock.Anything).Return([]byte("app_hash"), nil)
-	pool := newSnapshotPool(stateProvider)
+	pool := newSnapshotPool()
 
 	peerAID := types.NodeID("aa")
 	peerBID := types.NodeID("bb")

internal/statesync/stateprovider.go

@@ -1,7 +1,9 @@
 package statesync
 
 import (
+	"bytes"
 	"context"
+	"errors"
 	"fmt"
 	"strings"
 	"time"
@@ -9,21 +11,25 @@ import (
 	dbm "github.com/tendermint/tm-db"
 
 	tmsync "github.com/tendermint/tendermint/internal/libs/sync"
+	"github.com/tendermint/tendermint/internal/p2p"
 	"github.com/tendermint/tendermint/libs/log"
 	"github.com/tendermint/tendermint/light"
 	lightprovider "github.com/tendermint/tendermint/light/provider"
 	lighthttp "github.com/tendermint/tendermint/light/provider/http"
 	lightrpc "github.com/tendermint/tendermint/light/rpc"
 	lightdb "github.com/tendermint/tendermint/light/store/db"
+	ssproto "github.com/tendermint/tendermint/proto/tendermint/statesync"
 	rpchttp "github.com/tendermint/tendermint/rpc/client/http"
 	sm "github.com/tendermint/tendermint/state"
 	"github.com/tendermint/tendermint/types"
+	"github.com/tendermint/tendermint/version"
 )
 
 //go:generate ../../scripts/mockery_generate.sh StateProvider
 
 // StateProvider is a provider of trusted state data for bootstrapping a node. This refers
-// to the state.State object, not the state machine.
+// to the state.State object, not the state machine. There are two implementations. One
+// uses the P2P layer and the other uses the RPC layer. Both use light client verification.
 type StateProvider interface {
 	// AppHash returns the app hash after the given height has been committed.
 	AppHash(ctx context.Context, height uint64) ([]byte, error)
@@ -33,20 +39,17 @@ type StateProvider interface {
 	State(ctx context.Context, height uint64) (sm.State, error)
 }
 
-// lightClientStateProvider is a state provider using the light client.
-type lightClientStateProvider struct {
+type stateProviderRPC struct {
 	tmsync.Mutex  // light.Client is not concurrency-safe
 	lc            *light.Client
-	version       sm.Version
 	initialHeight int64
 	providers     map[lightprovider.Provider]string
 }
 
-// NewLightClientStateProvider creates a new StateProvider using a light client and RPC clients.
-func NewLightClientStateProvider(
+// NewRPCStateProvider creates a new StateProvider using a light client and RPC clients.
+func NewRPCStateProvider(
 	ctx context.Context,
 	chainID string,
-	version sm.Version,
 	initialHeight int64,
 	servers []string,
 	trustOptions light.TrustOptions,
@@ -75,51 +78,17 @@ func NewLightClientStateProvider(
 	if err != nil {
 		return nil, err
 	}
-	return &lightClientStateProvider{
+	return &stateProviderRPC{
 		lc:            lc,
-		version:       version,
 		initialHeight: initialHeight,
 		providers:     providerRemotes,
 	}, nil
 }
 
-// NewLightClientStateProviderFromDispatcher creates a light client state
-// provider but uses a p2p connected dispatched instead of RPC endpoints
-func NewLightClientStateProviderFromDispatcher(
-	ctx context.Context,
-	chainID string,
-	version sm.Version,
-	initialHeight int64,
-	dispatcher *dispatcher,
-	trustOptions light.TrustOptions,
-	logger log.Logger,
-) (StateProvider, error) {
-	providers := dispatcher.Providers(chainID, 30*time.Second)
-	if len(providers) < 2 {
-		return nil, fmt.Errorf("at least 2 peers are required, got %d", len(providers))
-	}
-
-	providersMap := make(map[lightprovider.Provider]string)
-	for _, p := range providers {
-		providersMap[p] = p.(*blockProvider).String()
-	}
-
-	lc, err := light.NewClient(ctx, chainID, trustOptions, providers[0], providers[1:],
-		lightdb.New(dbm.NewMemDB()), light.Logger(logger))
-	if err != nil {
-		return nil, err
-	}
-
-	return &lightClientStateProvider{
-		lc:            lc,
-		version:       version,
-		initialHeight: initialHeight,
-		providers:     providersMap,
-	}, nil
-}
-
-// AppHash implements StateProvider.
-func (s *lightClientStateProvider) AppHash(ctx context.Context, height uint64) ([]byte, error) {
+// AppHash implements part of StateProvider. It calls the application to verify the
+// light blocks at heights h+1 and h+2 and, if verification succeeds, reports the app
+// hash for the block at height h+1 which correlates to the state at height h.
+func (s *stateProviderRPC) AppHash(ctx context.Context, height uint64) ([]byte, error) {
 	s.Lock()
 	defer s.Unlock()
 
@@ -128,27 +97,19 @@ func (s *lightClientStateProvider) AppHash(ctx context.Context, height uint64) (
 	if err != nil {
 		return nil, err
 	}
-	// We also try to fetch the blocks at height H and H+2, since we need these
+
+	// We also try to fetch the blocks at H+2, since we need these
 	// when building the state while restoring the snapshot. This avoids the race
 	// condition where we try to restore a snapshot before H+2 exists.
-	//
-	// FIXME This is a hack, since we can't add new methods to the interface without
-	// breaking it. We should instead have a Has(ctx, height) method which checks
-	// that the state provider has access to the necessary data for the height.
-	// We piggyback on AppHash() since it's called when adding snapshots to the pool.
 	_, err = s.lc.VerifyLightBlockAtHeight(ctx, int64(height+2), time.Now())
 	if err != nil {
 		return nil, err
 	}
-	_, err = s.lc.VerifyLightBlockAtHeight(ctx, int64(height), time.Now())
-	if err != nil {
-		return nil, err
-	}
 	return header.AppHash, nil
 }
 
 // Commit implements StateProvider.
-func (s *lightClientStateProvider) Commit(ctx context.Context, height uint64) (*types.Commit, error) {
+func (s *stateProviderRPC) Commit(ctx context.Context, height uint64) (*types.Commit, error) {
 	s.Lock()
 	defer s.Unlock()
 	header, err := s.lc.VerifyLightBlockAtHeight(ctx, int64(height), time.Now())
@@ -159,13 +120,12 @@ func (s *lightClientStateProvider) Commit(ctx context.Context, height uint64) (*
 }
 
 // State implements StateProvider.
-func (s *lightClientStateProvider) State(ctx context.Context, height uint64) (sm.State, error) {
+func (s *stateProviderRPC) State(ctx context.Context, height uint64) (sm.State, error) {
 	s.Lock()
 	defer s.Unlock()
 
 	state := sm.State{
 		ChainID:       s.lc.ChainID(),
-		Version:       s.version,
 		InitialHeight: s.initialHeight,
 	}
 	if state.InitialHeight == 0 {
@@ -193,6 +153,10 @@ func (s *lightClientStateProvider) State(ctx context.Context, height uint64) (sm
 		return sm.State{}, err
 	}
 
+	state.Version = sm.Version{
+		Consensus: currentLightBlock.Version,
+		Software:  version.TMVersion,
+	}
 	state.LastBlockHeight = lastLightBlock.Height
 	state.LastBlockTime = lastLightBlock.Time
 	state.LastBlockID = lastLightBlock.Commit.BlockID
@@ -229,9 +193,188 @@ func rpcClient(server string) (*rpchttp.HTTP, error) {
 	if !strings.Contains(server, "://") {
 		server = "http://" + server
 	}
-	c, err := rpchttp.New(server)
+	return rpchttp.New(server)
+}
+
+type stateProviderP2P struct {
+	tmsync.Mutex  // light.Client is not concurrency-safe
+	lc            *light.Client
+	initialHeight int64
+	paramsSendCh  chan<- p2p.Envelope
+	paramsRecvCh  chan types.ConsensusParams
+}
+
+// NewP2PStateProvider creates a light client state
+// provider but uses a dispatcher connected to the P2P layer
+func NewP2PStateProvider(
+	ctx context.Context,
+	chainID string,
+	initialHeight int64,
+	providers []lightprovider.Provider,
+	trustOptions light.TrustOptions,
+	paramsSendCh chan<- p2p.Envelope,
+	logger log.Logger,
+) (StateProvider, error) {
+	if len(providers) < 2 {
+		return nil, fmt.Errorf("at least 2 peers are required, got %d", len(providers))
+	}
+
+	lc, err := light.NewClient(ctx, chainID, trustOptions, providers[0], providers[1:],
+		lightdb.New(dbm.NewMemDB()), light.Logger(logger))
 	if err != nil {
 		return nil, err
 	}
-	return c, nil
+
+	return &stateProviderP2P{
+		lc:            lc,
+		initialHeight: initialHeight,
+		paramsSendCh:  paramsSendCh,
+		paramsRecvCh:  make(chan types.ConsensusParams),
+	}, nil
+}
+
+// AppHash implements StateProvider.
+func (s *stateProviderP2P) AppHash(ctx context.Context, height uint64) ([]byte, error) {
+	s.Lock()
+	defer s.Unlock()
+
+	// We have to fetch the next height, which contains the app hash for the previous height.
+	header, err := s.lc.VerifyLightBlockAtHeight(ctx, int64(height+1), time.Now())
+	if err != nil {
+		return nil, err
+	}
+
+	// We also try to fetch the blocks at H+2, since we need these
+	// when building the state while restoring the snapshot. This avoids the race
+	// condition where we try to restore a snapshot before H+2 exists.
+	_, err = s.lc.VerifyLightBlockAtHeight(ctx, int64(height+2), time.Now())
+	if err != nil {
+		return nil, err
+	}
+	return header.AppHash, nil
+}
+
+// Commit implements StateProvider.
+func (s *stateProviderP2P) Commit(ctx context.Context, height uint64) (*types.Commit, error) {
+	s.Lock()
+	defer s.Unlock()
+	header, err := s.lc.VerifyLightBlockAtHeight(ctx, int64(height), time.Now())
+	if err != nil {
+		return nil, err
+	}
+	return header.Commit, nil
+}
+
+// State implements StateProvider.
+func (s *stateProviderP2P) State(ctx context.Context, height uint64) (sm.State, error) {
+	s.Lock()
+	defer s.Unlock()
+
+	state := sm.State{
+		ChainID:       s.lc.ChainID(),
+		InitialHeight: s.initialHeight,
+	}
+	if state.InitialHeight == 0 {
+		state.InitialHeight = 1
+	}
+
+	// The snapshot height maps onto the state heights as follows:
+	//
+	// height: last block, i.e. the snapshotted height
+	// height+1: current block, i.e. the first block we'll process after the snapshot
+	// height+2: next block, i.e. the second block after the snapshot
+	//
+	// We need to fetch the NextValidators from height+2 because if the application changed
+	// the validator set at the snapshot height then this only takes effect at height+2.
+	lastLightBlock, err := s.lc.VerifyLightBlockAtHeight(ctx, int64(height), time.Now())
+	if err != nil {
+		return sm.State{}, err
+	}
+	currentLightBlock, err := s.lc.VerifyLightBlockAtHeight(ctx, int64(height+1), time.Now())
+	if err != nil {
+		return sm.State{}, err
+	}
+	nextLightBlock, err := s.lc.VerifyLightBlockAtHeight(ctx, int64(height+2), time.Now())
+	if err != nil {
+		return sm.State{}, err
+	}
+
+	state.Version = sm.Version{
+		Consensus: currentLightBlock.Version,
+		Software:  version.TMVersion,
+	}
+	state.LastBlockHeight = lastLightBlock.Height
+	state.LastBlockTime = lastLightBlock.Time
+	state.LastBlockID = lastLightBlock.Commit.BlockID
+	state.AppHash = currentLightBlock.AppHash
+	state.LastResultsHash = currentLightBlock.LastResultsHash
+	state.LastValidators = lastLightBlock.ValidatorSet
+	state.Validators = currentLightBlock.ValidatorSet
+	state.NextValidators = nextLightBlock.ValidatorSet
+	state.LastHeightValidatorsChanged = nextLightBlock.Height
+
+	// We'll also need to fetch consensus params via P2P.
+	state.ConsensusParams, err = s.consensusParams(ctx, currentLightBlock.Height)
+	if err != nil {
+		return sm.State{}, err
+	}
+	// validate the consensus params
+	if !bytes.Equal(nextLightBlock.ConsensusHash, state.ConsensusParams.HashConsensusParams()) {
+		return sm.State{}, fmt.Errorf("consensus params hash mismatch at height %d. Expected %v, got %v",
+			currentLightBlock.Height, nextLightBlock.ConsensusHash, state.ConsensusParams.HashConsensusParams())
+	}
+	// set the last height changed to the current height
+	state.LastHeightConsensusParamsChanged = currentLightBlock.Height
+
+	return state, nil
+}
+
+// addProvider dynamically adds a peer as a new witness. A limit of 6 providers is kept as a
+// heuristic. Too many overburdens the network and too little compromises the second layer of security.
+func (s *stateProviderP2P) addProvider(p lightprovider.Provider) {
+	if len(s.lc.Witnesses()) < 6 {
+		s.lc.AddProvider(p)
+	}
+}
+
+// consensusParams sends out a request for consensus params blocking until one is returned.
+// If it fails to get a valid set of consensus params from any of the providers it returns an error.
+func (s *stateProviderP2P) consensusParams(ctx context.Context, height int64) (types.ConsensusParams, error) {
+	for _, provider := range s.lc.Witnesses() {
+		p, ok := provider.(*BlockProvider)
+		if !ok {
+			panic("expected p2p state provider to use p2p block providers")
+		}
+
+		// extract the nodeID of the provider
+		peer, err := types.NewNodeID(p.String())
+		if err != nil {
+			return types.ConsensusParams{}, fmt.Errorf("invalid provider (%s) node id: %w", p.String(), err)
+		}
+
+		select {
+		case s.paramsSendCh <- p2p.Envelope{
+			To: peer,
+			Message: &ssproto.ParamsRequest{
+				Height: uint64(height),
+			},
+		}:
+		case <-ctx.Done():
+			return types.ConsensusParams{}, ctx.Err()
+		}
+
+		select {
+		// if we get no response from this provider we move on to the next one
+		case <-time.After(consensusParamsResponseTimeout):
+			continue
+		case <-ctx.Done():
+			return types.ConsensusParams{}, ctx.Err()
+		case params, ok := <-s.paramsRecvCh:
+			if !ok {
+				return types.ConsensusParams{}, errors.New("params channel closed")
+			}
+			return params, nil
+		}
+	}
+	return types.ConsensusParams{}, errors.New("unable to fetch consensus params from connected providers")
 }

internal/statesync/syncer.go

@@ -12,6 +12,7 @@ import (
 	tmsync "github.com/tendermint/tendermint/internal/libs/sync"
 	"github.com/tendermint/tendermint/internal/p2p"
 	"github.com/tendermint/tendermint/libs/log"
+	"github.com/tendermint/tendermint/light"
 	ssproto "github.com/tendermint/tendermint/proto/tendermint/statesync"
 	"github.com/tendermint/tendermint/proxy"
 	sm "github.com/tendermint/tendermint/state"
@@ -40,14 +41,11 @@ var (
 	errRejectSender = errors.New("snapshot sender was rejected")
 	// errVerifyFailed is returned by Sync() when app hash or last height
 	// verification fails.
-	errVerifyFailed = errors.New("verification failed")
+	errVerifyFailed = errors.New("verification with app failed")
 	// errTimeout is returned by Sync() when we've waited too long to receive a chunk.
 	errTimeout = errors.New("timed out waiting for chunk")
 	// errNoSnapshots is returned by SyncAny() if no snapshots are found and discovery is disabled.
 	errNoSnapshots = errors.New("no suitable snapshots found")
-	// errStateCommitTimeout is returned by Sync() when the timeout for retrieving
-	// tendermint state or the commit is exceeded
-	errStateCommitTimeout = errors.New("timed out trying to retrieve state and commit")
 )
 
 // syncer runs a state sync against an ABCI app. Use either SyncAny() to automatically attempt to
@@ -84,7 +82,7 @@ func newSyncer(
 		stateProvider: stateProvider,
 		conn:          conn,
 		connQuery:     connQuery,
-		snapshots:     newSnapshotPool(stateProvider),
+		snapshots:     newSnapshotPool(),
 		snapshotCh:    snapshotCh,
 		chunkCh:       chunkCh,
 		tempDir:       tempDir,
@@ -153,7 +151,6 @@ func (s *syncer) SyncAny(
 	discoveryTime time.Duration,
 	requestSnapshots func(),
 ) (sm.State, *types.Commit, error) {
-
 	if discoveryTime != 0 && discoveryTime < minimumDiscoveryTime {
 		discoveryTime = minimumDiscoveryTime
 	}
@@ -181,7 +178,6 @@ func (s *syncer) SyncAny(
 			if discoveryTime == 0 {
 				return sm.State{}, nil, errNoSnapshots
 			}
-			requestSnapshots()
 			s.logger.Info(fmt.Sprintf("Discovering snapshots for %v", discoveryTime))
 			time.Sleep(discoveryTime)
 			continue
@@ -230,10 +226,6 @@ func (s *syncer) SyncAny(
 				s.logger.Info("Snapshot sender rejected", "peer", peer)
 			}
 
-		case errors.Is(err, errStateCommitTimeout):
-			s.logger.Info("Timed out retrieving state and commit, rejecting and retrying...", "height", snapshot.Height)
-			s.snapshots.Reject(snapshot)
-
 		default:
 			return sm.State{}, nil, fmt.Errorf("snapshot restoration failed: %w", err)
 		}
@@ -264,8 +256,29 @@ func (s *syncer) Sync(ctx context.Context, snapshot *snapshot, chunks *chunkQueu
 		s.mtx.Unlock()
 	}()
 
+	hctx, hcancel := context.WithTimeout(ctx, 30*time.Second)
+	defer hcancel()
+
+	// Fetch the app hash corresponding to the snapshot
+	appHash, err := s.stateProvider.AppHash(hctx, snapshot.Height)
+	if err != nil {
+		// check if the main context was triggered
+		if ctx.Err() != nil {
+			return sm.State{}, nil, ctx.Err()
+		}
+		// catch the case where all the light client providers have been exhausted
+		if err == light.ErrNoWitnesses {
+			return sm.State{}, nil,
+				fmt.Errorf("failed to get app hash at height %d. No witnesses remaining", snapshot.Height)
+		}
+		s.logger.Info("failed to get and verify tendermint state. Dropping snapshot and trying again",
+			"err", err, "height", snapshot.Height)
+		return sm.State{}, nil, errRejectSnapshot
+	}
+	snapshot.trustedAppHash = appHash
+
 	// Offer snapshot to ABCI app.
-	err := s.offerSnapshot(ctx, snapshot)
+	err = s.offerSnapshot(ctx, snapshot)
 	if err != nil {
 		return sm.State{}, nil, err
 	}
@@ -277,27 +290,37 @@ func (s *syncer) Sync(ctx context.Context, snapshot *snapshot, chunks *chunkQueu
 		go s.fetchChunks(fetchCtx, snapshot, chunks)
 	}
 
-	pctx, pcancel := context.WithTimeout(ctx, 30*time.Second)
+	pctx, pcancel := context.WithTimeout(ctx, 1*time.Minute)
 	defer pcancel()
 
 	// Optimistically build new state, so we don't discover any light client failures at the end.
 	state, err := s.stateProvider.State(pctx, snapshot.Height)
 	if err != nil {
-		// check if the provider context exceeded the 10 second deadline
-		if err == context.DeadlineExceeded && ctx.Err() == nil {
-			return sm.State{}, nil, errStateCommitTimeout
+		// check if the main context was triggered
+		if ctx.Err() != nil {
+			return sm.State{}, nil, ctx.Err()
 		}
-
-		return sm.State{}, nil, fmt.Errorf("failed to build new state: %w", err)
+		if err == light.ErrNoWitnesses {
+			return sm.State{}, nil,
+				fmt.Errorf("failed to get tendermint state at height %d. No witnesses remaining", snapshot.Height)
+		}
+		s.logger.Info("failed to get and verify tendermint state. Dropping snapshot and trying again",
+			"err", err, "height", snapshot.Height)
+		return sm.State{}, nil, errRejectSnapshot
 	}
 	commit, err := s.stateProvider.Commit(pctx, snapshot.Height)
 	if err != nil {
 		// check if the provider context exceeded the 10 second deadline
-		if err == context.DeadlineExceeded && ctx.Err() == nil {
-			return sm.State{}, nil, errStateCommitTimeout
+		if ctx.Err() != nil {
+			return sm.State{}, nil, ctx.Err()
 		}
-
-		return sm.State{}, nil, fmt.Errorf("failed to fetch commit: %w", err)
+		if err == light.ErrNoWitnesses {
+			return sm.State{}, nil,
+				fmt.Errorf("failed to get commit at height %d. No witnesses remaining", snapshot.Height)
+		}
+		s.logger.Info("failed to get and verify commit. Dropping snapshot and trying again",
+			"err", err, "height", snapshot.Height)
+		return sm.State{}, nil, errRejectSnapshot
 	}
 
 	// Restore snapshot

internal/test/factory/doc.go

@@ -0,0 +1,6 @@
+/*
+Package factory provides generation code for common structs in Tendermint.
+It is used primarily for the testing of internal components such as statesync,
+consensus, blocksync etc..
+*/
+package factory

internal/test/factory/factory_test.go

@@ -12,3 +12,7 @@ func TestMakeHeader(t *testing.T) {
 	_, err := MakeHeader(&types.Header{})
 	assert.NoError(t, err)
 }
+
+func TestRandomNodeID(t *testing.T) {
+	assert.NotPanics(t, func() { RandomNodeID() })
+}

internal/test/factory/p2p.go

@@ -0,0 +1,27 @@
+package factory
+
+import (
+	"encoding/hex"
+	"strings"
+
+	"github.com/tendermint/tendermint/libs/rand"
+	"github.com/tendermint/tendermint/types"
+)
+
+// NodeID returns a valid NodeID based on an inputted string
+func NodeID(str string) types.NodeID {
+	id, err := types.NewNodeID(strings.Repeat(str, 2*types.NodeIDByteLength))
+	if err != nil {
+		panic(err)
+	}
+	return id
+}
+
+// RandomNodeID returns a randomly generated valid NodeID
+func RandomNodeID() types.NodeID {
+	id, err := types.NewNodeID(hex.EncodeToString(rand.Bytes(types.NodeIDByteLength)))
+	if err != nil {
+		panic(err)
+	}
+	return id
+}

libs/time/time.go

@@ -1,7 +1,6 @@
 package time
 
 import (
-	"sort"
 	"time"
 )
 
@@ -16,43 +15,3 @@ func Now() time.Time {
 func Canonical(t time.Time) time.Time {
 	return t.Round(0).UTC()
 }
-
-// WeightedTime for computing a median.
-type WeightedTime struct {
-	Time   time.Time
-	Weight int64
-}
-
-// NewWeightedTime with time and weight.
-func NewWeightedTime(time time.Time, weight int64) *WeightedTime {
-	return &WeightedTime{
-		Time:   time,
-		Weight: weight,
-	}
-}
-
-// WeightedMedian computes weighted median time for a given array of WeightedTime and the total voting power.
-func WeightedMedian(weightedTimes []*WeightedTime, totalVotingPower int64) (res time.Time) {
-	median := totalVotingPower / 2
-
-	sort.Slice(weightedTimes, func(i, j int) bool {
-		if weightedTimes[i] == nil {
-			return false
-		}
-		if weightedTimes[j] == nil {
-			return true
-		}
-		return weightedTimes[i].Time.UnixNano() < weightedTimes[j].Time.UnixNano()
-	})
-
-	for _, weightedTime := range weightedTimes {
-		if weightedTime != nil {
-			if median <= weightedTime.Weight {
-				res = weightedTime.Time
-				break
-			}
-			median -= weightedTime.Weight
-		}
-	}
-	return
-}

light/client.go

@@ -52,6 +52,8 @@ const (
 
 	// 10s is sufficient for most networks.
 	defaultMaxBlockLag = 10 * time.Second
+
+	defaultProviderTimeout = 10 * time.Second
 )
 
 // Option sets a parameter for the light client.
@@ -61,9 +63,7 @@ type Option func(*Client)
 // check the blocks (every block, in ascending height order). Note this is
 // much slower than SkippingVerification, albeit more secure.
 func SequentialVerification() Option {
-	return func(c *Client) {
-		c.verificationMode = sequential
-	}
+	return func(c *Client) { c.verificationMode = sequential }
 }
 
 // SkippingVerification option configures the light client to skip blocks as
@@ -87,24 +87,18 @@ func SkippingVerification(trustLevel tmmath.Fraction) Option {
 // the h amount of light blocks will be removed from the store.
 // Default: 1000. A pruning size of 0 will not prune the light client at all.
 func PruningSize(h uint16) Option {
-	return func(c *Client) {
-		c.pruningSize = h
-	}
+	return func(c *Client) { c.pruningSize = h }
 }
 
 // Logger option can be used to set a logger for the client.
 func Logger(l log.Logger) Option {
-	return func(c *Client) {
-		c.logger = l
-	}
+	return func(c *Client) { c.logger = l }
 }
 
 // MaxClockDrift defines how much new header's time can drift into
 // the future relative to the light clients local time. Default: 10s.
 func MaxClockDrift(d time.Duration) Option {
-	return func(c *Client) {
-		c.maxClockDrift = d
-	}
+	return func(c *Client) { c.maxClockDrift = d }
 }
 
 // MaxBlockLag represents the maximum time difference between the realtime
@@ -116,9 +110,13 @@ func MaxClockDrift(d time.Duration) Option {
 // was 12:00. Then the lag here is 5 minutes.
 // Default: 10s
 func MaxBlockLag(d time.Duration) Option {
-	return func(c *Client) {
-		c.maxBlockLag = d
-	}
+	return func(c *Client) { c.maxBlockLag = d }
+}
+
+// Provider timeout is the maximum time that the light client will wait for a
+// provider to respond with a light block.
+func ProviderTimeout(d time.Duration) Option {
+	return func(c *Client) { c.providerTimeout = d }
 }
 
 // Client represents a light client, connected to a single chain, which gets
@@ -133,6 +131,7 @@ type Client struct {
 	trustLevel       tmmath.Fraction
 	maxClockDrift    time.Duration
 	maxBlockLag      time.Duration
+	providerTimeout  time.Duration
 
 	// Mutex for locking during changes of the light clients providers
 	providerMutex tmsync.Mutex
@@ -197,12 +196,13 @@ func NewClient(
 		chainID:          chainID,
 		trustingPeriod:   trustOptions.Period,
 		verificationMode: skipping,
-		trustLevel:       DefaultTrustLevel,
-		maxClockDrift:    defaultMaxClockDrift,
-		maxBlockLag:      defaultMaxBlockLag,
 		primary:          primary,
 		witnesses:        witnesses,
 		trustedStore:     trustedStore,
+		trustLevel:       DefaultTrustLevel,
+		maxClockDrift:    defaultMaxClockDrift,
+		maxBlockLag:      defaultMaxBlockLag,
+		providerTimeout:  defaultProviderTimeout,
 		pruningSize:      defaultPruningSize,
 		logger:           log.NewNopLogger(),
 	}
@@ -693,7 +693,9 @@ func (c *Client) verifySkipping(
 			if depth == len(blockCache)-1 {
 				// schedule what the next height we need to fetch is
 				pivotHeight := c.schedule(verifiedBlock.Height, blockCache[depth].Height)
-				interimBlock, providerErr := source.LightBlock(ctx, pivotHeight)
+				subCtx, cancel := context.WithTimeout(ctx, c.providerTimeout)
+				defer cancel()
+				interimBlock, providerErr := c.getLightBlock(subCtx, source, pivotHeight)
 				if providerErr != nil {
 					return nil, ErrVerificationFailed{From: verifiedBlock.Height, To: pivotHeight, Reason: providerErr}
 				}
@@ -930,7 +932,7 @@ func (c *Client) backwards(
 //    any other error, the primary is permanently dropped and is replaced by a witness.
 func (c *Client) lightBlockFromPrimary(ctx context.Context, height int64) (*types.LightBlock, error) {
 	c.providerMutex.Lock()
-	l, err := c.primary.LightBlock(ctx, height)
+	l, err := c.getLightBlock(ctx, c.primary, height)
 	c.providerMutex.Unlock()
 
 	switch err {
@@ -957,6 +959,16 @@ func (c *Client) lightBlockFromPrimary(ctx context.Context, height int64) (*type
 	}
 }
 
+func (c *Client) getLightBlock(ctx context.Context, p provider.Provider, height int64) (*types.LightBlock, error) {
+	subCtx, cancel := context.WithTimeout(ctx, c.providerTimeout)
+	defer cancel()
+	l, err := p.LightBlock(subCtx, height)
+	if err == context.DeadlineExceeded || ctx.Err() != nil {
+		return nil, provider.ErrNoResponse
+	}
+	return l, err
+}
+
 // NOTE: requires a providerMutex lock
 func (c *Client) removeWitnesses(indexes []int) error {
 	// check that we will still have witnesses remaining
@@ -989,7 +1001,7 @@ func (c *Client) findNewPrimary(ctx context.Context, height int64, remove bool)
 	c.providerMutex.Lock()
 	defer c.providerMutex.Unlock()
 
-	if len(c.witnesses) <= 1 {
+	if len(c.witnesses) < 1 {
 		return nil, ErrNoWitnesses
 	}
 
@@ -1001,7 +1013,7 @@ func (c *Client) findNewPrimary(ctx context.Context, height int64, remove bool)
 	)
 
 	// send out a light block request to all witnesses
-	subctx, cancel := context.WithCancel(ctx)
+	subctx, cancel := context.WithTimeout(ctx, c.providerTimeout)
 	defer cancel()
 	for index := range c.witnesses {
 		wg.Add(1)

light/client_test.go

@@ -724,51 +724,32 @@ func TestClient_BackwardsVerification(t *testing.T) {
 
 	}
 	{
-		testCases := []struct {
-			headers map[int64]*types.SignedHeader
-			vals    map[int64]*types.ValidatorSet
-		}{
-			{
-				// 7) provides incorrect height
-				headers: map[int64]*types.SignedHeader{
-					2: keys.GenSignedHeader(chainID, 1, bTime.Add(30*time.Minute), nil, vals, vals,
-						hash("app_hash"), hash("cons_hash"), hash("results_hash"), 0, len(keys)),
-					3: h3,
-				},
-				vals: valSet,
-			},
-			{
-				// 8) provides incorrect hash
-				headers: map[int64]*types.SignedHeader{
-					2: keys.GenSignedHeader(chainID, 2, bTime.Add(30*time.Minute), nil, vals, vals,
-						hash("app_hash2"), hash("cons_hash23"), hash("results_hash30"), 0, len(keys)),
-					3: h3,
-				},
-				vals: valSet,
-			},
+		// 8) provides incorrect hash
+		headers := map[int64]*types.SignedHeader{
+			2: keys.GenSignedHeader(chainID, 2, bTime.Add(30*time.Minute), nil, vals, vals,
+				hash("app_hash2"), hash("cons_hash23"), hash("results_hash30"), 0, len(keys)),
+			3: h3,
 		}
+		vals := valSet
+		mockNode := mockNodeFromHeadersAndVals(headers, vals)
+		c, err := light.NewClient(
+			ctx,
+			chainID,
+			light.TrustOptions{
+				Period: 1 * time.Hour,
+				Height: 3,
+				Hash:   h3.Hash(),
+			},
+			mockNode,
+			[]provider.Provider{mockNode},
+			dbs.New(dbm.NewMemDB()),
+			light.Logger(log.TestingLogger()),
+		)
+		require.NoError(t, err)
 
-		for idx, tc := range testCases {
-			mockNode := mockNodeFromHeadersAndVals(tc.headers, tc.vals)
-			c, err := light.NewClient(
-				ctx,
-				chainID,
-				light.TrustOptions{
-					Period: 1 * time.Hour,
-					Height: 3,
-					Hash:   h3.Hash(),
-				},
-				mockNode,
-				[]provider.Provider{mockNode},
-				dbs.New(dbm.NewMemDB()),
-				light.Logger(log.TestingLogger()),
-			)
-			require.NoError(t, err, idx)
-
-			_, err = c.VerifyLightBlockAtHeight(ctx, 2, bTime.Add(1*time.Hour).Add(1*time.Second))
-			assert.Error(t, err, idx)
-			mockNode.AssertExpectations(t)
-		}
+		_, err = c.VerifyLightBlockAtHeight(ctx, 2, bTime.Add(1*time.Hour).Add(1*time.Second))
+		assert.Error(t, err)
+		mockNode.AssertExpectations(t)
 	}
 }
 

light/detector.go

@@ -110,7 +110,7 @@ func (c *Client) detectDivergence(ctx context.Context, primaryTrace []*types.Lig
 func (c *Client) compareNewHeaderWithWitness(ctx context.Context, errc chan error, h *types.SignedHeader,
 	witness provider.Provider, witnessIndex int) {
 
-	lightBlock, err := witness.LightBlock(ctx, h.Height)
+	lightBlock, err := c.getLightBlock(ctx, witness, h.Height)
 	switch err {
 	// no error means we move on to checking the hash of the two headers
 	case nil:
@@ -331,7 +331,7 @@ func (c *Client) examineConflictingHeaderAgainstTrace(
 		if traceBlock.Height == targetBlock.Height {
 			sourceBlock = targetBlock
 		} else {
-			sourceBlock, err = source.LightBlock(ctx, traceBlock.Height)
+			sourceBlock, err = c.getLightBlock(ctx, source, traceBlock.Height)
 			if err != nil {
 				return nil, nil, fmt.Errorf("failed to examine trace: %w", err)
 			}
@@ -379,7 +379,7 @@ func (c *Client) getTargetBlockOrLatest(
 	height int64,
 	witness provider.Provider,
 ) (bool, *types.LightBlock, error) {
-	lightBlock, err := witness.LightBlock(ctx, 0)
+	lightBlock, err := c.getLightBlock(ctx, witness, 0)
 	if err != nil {
 		return false, nil, err
 	}
@@ -394,7 +394,7 @@ func (c *Client) getTargetBlockOrLatest(
 		// the witness has caught up. We recursively call the function again. However in order
 		// to avoud a wild goose chase where the witness sends us one header below and one header
 		// above the height we set a timeout to the context
-		lightBlock, err := witness.LightBlock(ctx, height)
+		lightBlock, err := c.getLightBlock(ctx, witness, height)
 		return true, lightBlock, err
 	}
 

light/proxy/proxy.go

@@ -113,7 +113,7 @@ func (p *Proxy) listen() (net.Listener, *http.ServeMux, error) {
 	}
 
 	// 4) Start listening for new connections.
-	listener, err := rpcserver.Listen(p.Addr, p.Config)
+	listener, err := rpcserver.Listen(p.Addr, p.Config.MaxOpenConnections)
 	if err != nil {
 		return nil, mux, err
 	}

networks/remote/ansible/inventory/digital_ocean.py

@@ -1,6 +1,6 @@
 #!/usr/bin/env python
 
-'''
+"""
 DigitalOcean external inventory script
 ======================================
 
@@ -22,7 +22,7 @@ found.  You can force this script to use the cache with --force-cache.
 
 ----
 Configuration is read from `digital_ocean.ini`, then from environment variables,
-then and command-line arguments.
+and then from command-line arguments.
 
 Most notably, the DigitalOcean API Token must be specified. It can be specified
 in the INI file or with the following environment variables:
@@ -40,6 +40,7 @@ is to use the output of the --env option with export:
 The following groups are generated from --list:
  - ID    (droplet ID)
  - NAME  (droplet NAME)
+ - digital_ocean
  - image_ID
  - image_NAME
  - distro_NAME  (distribution NAME from image)
@@ -73,14 +74,12 @@ For each host, the following variables are registered:
 
 -----
 ```
-usage: digital_ocean.py [-h] [--list] [--host HOST] [--all]
-                                 [--droplets] [--regions] [--images] [--sizes]
-                                 [--ssh-keys] [--domains] [--pretty]
-                                 [--cache-path CACHE_PATH]
-                                 [--cache-max_age CACHE_MAX_AGE]
-                                 [--force-cache]
-                                 [--refresh-cache]
-                                 [--api-token API_TOKEN]
+usage: digital_ocean.py [-h] [--list] [--host HOST] [--all] [--droplets]
+                        [--regions] [--images] [--sizes] [--ssh-keys]
+                        [--domains] [--tags] [--pretty]
+                        [--cache-path CACHE_PATH]
+                        [--cache-max_age CACHE_MAX_AGE] [--force-cache]
+                        [--refresh-cache] [--env] [--api-token API_TOKEN]
 
 Produce an Ansible Inventory file based on DigitalOcean credentials
 
@@ -91,65 +90,129 @@ optional arguments:
   --host HOST           Get all Ansible inventory variables about a specific
                         Droplet
   --all                 List all DigitalOcean information as JSON
-  --droplets            List Droplets as JSON
+  --droplets, -d        List Droplets as JSON
   --regions             List Regions as JSON
   --images              List Images as JSON
   --sizes               List Sizes as JSON
   --ssh-keys            List SSH keys as JSON
   --domains             List Domains as JSON
+  --tags                List Tags as JSON
   --pretty, -p          Pretty-print results
   --cache-path CACHE_PATH
                         Path to the cache files (default: .)
   --cache-max_age CACHE_MAX_AGE
                         Maximum age of the cached items (default: 0)
   --force-cache         Only use data from the cache
-  --refresh-cache       Force refresh of cache by making API requests to
+  --refresh-cache, -r   Force refresh of cache by making API requests to
                         DigitalOcean (default: False - use cache files)
+  --env, -e             Display DO_API_TOKEN
   --api-token API_TOKEN, -a API_TOKEN
                         DigitalOcean API Token
 ```
 
-'''
+"""
 
 # (c) 2013, Evan Wies <evan@neomantra.net>
+# (c) 2017, Ansible Project
+# (c) 2017, Abhijeet Kasurde <akasurde@redhat.com>
 #
 # Inspired by the EC2 inventory plugin:
 # https://github.com/ansible/ansible/blob/devel/contrib/inventory/ec2.py
 #
-# This file is part of Ansible,
-#
-# Ansible is free software: you can redistribute it and/or modify
-# it under the terms of the GNU General Public License as published by
-# the Free Software Foundation, either version 3 of the License, or
-# (at your option) any later version.
-#
-# Ansible is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-# GNU General Public License for more details.
-#
-# You should have received a copy of the GNU General Public License
-# along with Ansible.  If not, see <http://www.gnu.org/licenses/>.
+# GNU General Public License v3.0+ (see COPYING or https://www.gnu.org/licenses/gpl-3.0.txt)
+
+from __future__ import (absolute_import, division, print_function)
+__metaclass__ = type
 
 ######################################################################
 
-import os
-import sys
-import re
 import argparse
-from time import time
-import ConfigParser
 import ast
+import os
+import re
+import requests
+import sys
+from time import time
 
 try:
-    import json
+    import ConfigParser
 except ImportError:
-    import simplejson as json
+    import configparser as ConfigParser
 
-try:
-    from dopy.manager import DoManager
-except ImportError as e:
-    sys.exit("failed=True msg='`dopy` library required for this script'")
+import json
+
+
+class DoManager:
+    def __init__(self, api_token):
+        self.api_token = api_token
+        self.api_endpoint = 'https://api.digitalocean.com/v2'
+        self.headers = {'Authorization': 'Bearer {0}'.format(self.api_token),
+                        'Content-type': 'application/json'}
+        self.timeout = 60
+
+    def _url_builder(self, path):
+        if path[0] == '/':
+            path = path[1:]
+        return '%s/%s' % (self.api_endpoint, path)
+
+    def send(self, url, method='GET', data=None):
+        url = self._url_builder(url)
+        data = json.dumps(data)
+        try:
+            if method == 'GET':
+                resp_data = {}
+                incomplete = True
+                while incomplete:
+                    resp = requests.get(url, data=data, headers=self.headers, timeout=self.timeout)
+                    json_resp = resp.json()
+
+                    for key, value in json_resp.items():
+                        if isinstance(value, list) and key in resp_data:
+                            resp_data[key] += value
+                        else:
+                            resp_data[key] = value
+
+                    try:
+                        url = json_resp['links']['pages']['next']
+                    except KeyError:
+                        incomplete = False
+
+        except ValueError as e:
+            sys.exit("Unable to parse result from %s: %s" % (url, e))
+        return resp_data
+
+    def all_active_droplets(self):
+        resp = self.send('droplets/')
+        return resp['droplets']
+
+    def all_regions(self):
+        resp = self.send('regions/')
+        return resp['regions']
+
+    def all_images(self, filter_name='global'):
+        params = {'filter': filter_name}
+        resp = self.send('images/', data=params)
+        return resp['images']
+
+    def sizes(self):
+        resp = self.send('sizes/')
+        return resp['sizes']
+
+    def all_ssh_keys(self):
+        resp = self.send('account/keys')
+        return resp['ssh_keys']
+
+    def all_domains(self):
+        resp = self.send('domains/')
+        return resp['domains']
+
+    def show_droplet(self, droplet_id):
+        resp = self.send('droplets/%s' % droplet_id)
+        return resp['droplet']
+
+    def all_tags(self):
+        resp = self.send('tags')
+        return resp['tags']
 
 
 class DigitalOceanInventory(object):
@@ -159,7 +222,7 @@ class DigitalOceanInventory(object):
     ###########################################################################
 
     def __init__(self):
-        ''' Main execution path '''
+        """Main execution path """
 
         # DigitalOceanInventory data
         self.data = {}  # All DigitalOcean data
@@ -178,9 +241,9 @@ class DigitalOceanInventory(object):
 
         # Verify credentials were set
         if not hasattr(self, 'api_token'):
-            sys.stderr.write('''Could not find values for DigitalOcean api_token.
-They must be specified via either ini file, command line argument (--api-token),
-or environment variables (DO_API_TOKEN)\n''')
+            msg = 'Could not find values for DigitalOcean api_token. They must be specified via either ini file, ' \
+                  'command line argument (--api-token), or environment variables (DO_API_TOKEN)\n'
+            sys.stderr.write(msg)
             sys.exit(-1)
 
         # env command, show DigitalOcean credentials
@@ -196,10 +259,10 @@ or environment variables (DO_API_TOKEN)\n''')
             self.load_from_cache()
             if len(self.data) == 0:
                 if self.args.force_cache:
-                    sys.stderr.write('''Cache is empty and --force-cache was specified\n''')
+                    sys.stderr.write('Cache is empty and --force-cache was specified\n')
                     sys.exit(-1)
 
-        self.manager = DoManager(None, self.api_token, api_version=2)
+        self.manager = DoManager(self.api_token)
 
         # Pick the json_data to print based on the CLI command
         if self.args.droplets:
@@ -220,6 +283,9 @@ or environment variables (DO_API_TOKEN)\n''')
         elif self.args.domains:
             self.load_from_digital_ocean('domains')
             json_data = {'domains': self.data['domains']}
+        elif self.args.tags:
+            self.load_from_digital_ocean('tags')
+            json_data = {'tags': self.data['tags']}
         elif self.args.all:
             self.load_from_digital_ocean()
             json_data = self.data
@@ -234,19 +300,19 @@ or environment variables (DO_API_TOKEN)\n''')
             self.write_to_cache()
 
         if self.args.pretty:
-            print(json.dumps(json_data, sort_keys=True, indent=2))
+            print(json.dumps(json_data, indent=2))
         else:
             print(json.dumps(json_data))
-        # That's all she wrote...
 
     ###########################################################################
     # Script configuration
     ###########################################################################
 
     def read_settings(self):
-        ''' Reads the settings from the digital_ocean.ini file '''
-        config = ConfigParser.SafeConfigParser()
-        config.read(os.path.dirname(os.path.realpath(__file__)) + '/digital_ocean.ini')
+        """ Reads the settings from the digital_ocean.ini file """
+        config = ConfigParser.ConfigParser()
+        config_path = os.path.join(os.path.dirname(os.path.realpath(__file__)), 'digital_ocean.ini')
+        config.read(config_path)
 
         # Credentials
         if config.has_option('digital_ocean', 'api_token'):
@@ -267,7 +333,7 @@ or environment variables (DO_API_TOKEN)\n''')
             self.group_variables = ast.literal_eval(config.get('digital_ocean', 'group_variables'))
 
     def read_environment(self):
-        ''' Reads the settings from environment variables '''
+        """ Reads the settings from environment variables """
         # Setup credentials
         if os.getenv("DO_API_TOKEN"):
             self.api_token = os.getenv("DO_API_TOKEN")
@@ -275,7 +341,7 @@ or environment variables (DO_API_TOKEN)\n''')
             self.api_token = os.getenv("DO_API_KEY")
 
     def read_cli_args(self):
-        ''' Command line argument processing '''
+        """ Command line argument processing """
         parser = argparse.ArgumentParser(description='Produce an Ansible Inventory file based on DigitalOcean credentials')
 
         parser.add_argument('--list', action='store_true', help='List all active Droplets as Ansible inventory (default: True)')
@@ -288,6 +354,7 @@ or environment variables (DO_API_TOKEN)\n''')
         parser.add_argument('--sizes', action='store_true', help='List Sizes as JSON')
         parser.add_argument('--ssh-keys', action='store_true', help='List SSH keys as JSON')
         parser.add_argument('--domains', action='store_true', help='List Domains as JSON')
+        parser.add_argument('--tags', action='store_true', help='List Tags as JSON')
 
         parser.add_argument('--pretty', '-p', action='store_true', help='Pretty-print results')
 
@@ -309,6 +376,7 @@ or environment variables (DO_API_TOKEN)\n''')
         if (not self.args.droplets and not self.args.regions and
                 not self.args.images and not self.args.sizes and
                 not self.args.ssh_keys and not self.args.domains and
+                not self.args.tags and
                 not self.args.all and not self.args.host):
             self.args.list = True
 
@@ -317,7 +385,7 @@ or environment variables (DO_API_TOKEN)\n''')
     ###########################################################################
 
     def load_from_digital_ocean(self, resource=None):
-        '''Get JSON from DigitalOcean API'''
+        """Get JSON from DigitalOcean API """
         if self.args.force_cache and os.path.isfile(self.cache_filename):
             return
         # We always get fresh droplets
@@ -333,7 +401,7 @@ or environment variables (DO_API_TOKEN)\n''')
             self.data['regions'] = self.manager.all_regions()
             self.cache_refreshed = True
         if resource == 'images' or resource is None:
-            self.data['images'] = self.manager.all_images(filter=None)
+            self.data['images'] = self.manager.all_images()
             self.cache_refreshed = True
         if resource == 'sizes' or resource is None:
             self.data['sizes'] = self.manager.sizes()
@@ -344,9 +412,27 @@ or environment variables (DO_API_TOKEN)\n''')
         if resource == 'domains' or resource is None:
             self.data['domains'] = self.manager.all_domains()
             self.cache_refreshed = True
+        if resource == 'tags' or resource is None:
+            self.data['tags'] = self.manager.all_tags()
+            self.cache_refreshed = True
+
+    def add_inventory_group(self, key):
+        """ Method to create group dict """
+        host_dict = {'hosts': [], 'vars': {}}
+        self.inventory[key] = host_dict
+        return
+
+    def add_host(self, group, host):
+        """ Helper method to reduce host duplication """
+        if group not in self.inventory:
+            self.add_inventory_group(group)
+
+        if host not in self.inventory[group]['hosts']:
+            self.inventory[group]['hosts'].append(host)
+        return
 
     def build_inventory(self):
-        '''Build Ansible inventory of droplets'''
+        """ Build Ansible inventory of droplets """
         self.inventory = {
             'all': {
                 'hosts': [],
@@ -357,52 +443,44 @@ or environment variables (DO_API_TOKEN)\n''')
 
         # add all droplets by id and name
         for droplet in self.data['droplets']:
-            # when using private_networking, the API reports the private one in "ip_address".
-            if 'private_networking' in droplet['features'] and not self.use_private_network:
-                for net in droplet['networks']['v4']:
-                    if net['type'] == 'public':
-                        dest = net['ip_address']
-                    else:
-                        continue
-            else:
-                dest = droplet['ip_address']
+            for net in droplet['networks']['v4']:
+                if net['type'] == 'public':
+                    dest = net['ip_address']
+                else:
+                    continue
 
             self.inventory['all']['hosts'].append(dest)
 
-            self.inventory[droplet['id']] = [dest]
-            self.inventory[droplet['name']] = [dest]
+            self.add_host(droplet['id'], dest)
+
+            self.add_host(droplet['name'], dest)
 
             # groups that are always present
-            for group in ('region_' + droplet['region']['slug'],
+            for group in ('digital_ocean',
+                          'region_' + droplet['region']['slug'],
                           'image_' + str(droplet['image']['id']),
                           'size_' + droplet['size']['slug'],
-                          'distro_' + self.to_safe(droplet['image']['distribution']),
+                          'distro_' + DigitalOceanInventory.to_safe(droplet['image']['distribution']),
                           'status_' + droplet['status']):
-                if group not in self.inventory:
-                    self.inventory[group] = {'hosts': [], 'vars': {}}
-                self.inventory[group]['hosts'].append(dest)
+                self.add_host(group, dest)
 
             # groups that are not always present
             for group in (droplet['image']['slug'],
                           droplet['image']['name']):
                 if group:
-                    image = 'image_' + self.to_safe(group)
-                    if image not in self.inventory:
-                        self.inventory[image] = {'hosts': [], 'vars': {}}
-                    self.inventory[image]['hosts'].append(dest)
+                    image = 'image_' + DigitalOceanInventory.to_safe(group)
+                    self.add_host(image, dest)
 
             if droplet['tags']:
                 for tag in droplet['tags']:
-                    if tag not in self.inventory:
-                        self.inventory[tag] = {'hosts': [], 'vars': {}}
-                    self.inventory[tag]['hosts'].append(dest)
+                    self.add_host(tag, dest)
 
             # hostvars
             info = self.do_namespace(droplet)
             self.inventory['_meta']['hostvars'][dest] = info
 
     def load_droplet_variables_for_host(self):
-        '''Generate a JSON response to a --host call'''
+        """ Generate a JSON response to a --host call """
         host = int(self.args.host)
         droplet = self.manager.show_droplet(host)
         info = self.do_namespace(droplet)
@@ -413,7 +491,7 @@ or environment variables (DO_API_TOKEN)\n''')
     ###########################################################################
 
     def is_cache_valid(self):
-        ''' Determines if the cache files have expired, or if it is still valid '''
+        """ Determines if the cache files have expired, or if it is still valid """
         if os.path.isfile(self.cache_filename):
             mod_time = os.path.getmtime(self.cache_filename)
             current_time = time()
@@ -422,11 +500,10 @@ or environment variables (DO_API_TOKEN)\n''')
         return False
 
     def load_from_cache(self):
-        ''' Reads the data from the cache file and assigns it to member variables as Python Objects'''
+        """ Reads the data from the cache file and assigns it to member variables as Python Objects """
         try:
-            cache = open(self.cache_filename, 'r')
-            json_data = cache.read()
-            cache.close()
+            with open(self.cache_filename, 'r') as cache:
+                json_data = cache.read()
             data = json.loads(json_data)
         except IOError:
             data = {'data': {}, 'inventory': {}}
@@ -435,31 +512,24 @@ or environment variables (DO_API_TOKEN)\n''')
         self.inventory = data['inventory']
 
     def write_to_cache(self):
-        ''' Writes data in JSON format to a file '''
+        """ Writes data in JSON format to a file """
         data = {'data': self.data, 'inventory': self.inventory}
-        json_data = json.dumps(data, sort_keys=True, indent=2)
+        json_data = json.dumps(data, indent=2)
 
-        cache = open(self.cache_filename, 'w')
-        cache.write(json_data)
-        cache.close()
+        with open(self.cache_filename, 'w') as cache:
+            cache.write(json_data)
 
     ###########################################################################
     # Utilities
     ###########################################################################
+    @staticmethod
+    def to_safe(word):
+        """ Converts 'bad' characters in a string to underscores so they can be used as Ansible groups """
+        return re.sub(r"[^A-Za-z0-9\-.]", "_", word)
 
-    def push(self, my_dict, key, element):
-        ''' Pushed an element onto an array that may not have been defined in the dict '''
-        if key in my_dict:
-            my_dict[key].append(element)
-        else:
-            my_dict[key] = [element]
-
-    def to_safe(self, word):
-        ''' Converts 'bad' characters in a string to underscores so they can be used as Ansible groups '''
-        return re.sub("[^A-Za-z0-9\-\.]", "_", word)
-
-    def do_namespace(self, data):
-        ''' Returns a copy of the dictionary with all the keys put in a 'do_' namespace '''
+    @staticmethod
+    def do_namespace(data):
+        """ Returns a copy of the dictionary with all the keys put in a 'do_' namespace """
         info = {}
         for k, v in data.items():
             info['do_' + k] = v

networks/remote/terraform/cluster/main.tf

@@ -1,3 +1,12 @@
+terraform {
+  required_providers {
+    digitalocean = {
+      source = "digitalocean/digitalocean"
+      version = "~> 2.0"
+    }
+  }
+}
+
 resource "digitalocean_tag" "cluster" {
   name = "${var.name}"
 }

networks/remote/terraform/cluster/variables.tf

@@ -4,13 +4,13 @@ variable "name" {
 
 variable "regions" {
   description = "Regions to launch in"
-  type = "list"
+  type = list
   default = ["AMS3", "FRA1", "LON1", "NYC3", "SFO2", "SGP1", "TOR1"]
 }
 
 variable "ssh_key" {
   description = "SSH key filename to copy to the nodes"
-  type = "string"
+  type = string
 }
 
 variable "instance_size" {

networks/remote/terraform/main.tf

@@ -1,5 +1,14 @@
 #Terraform Configuration
 
+terraform {
+  required_providers {
+    digitalocean = {
+      source = "digitalocean/digitalocean"
+      version = "~> 2.0"
+    }
+  }
+}
+
 variable "DO_API_TOKEN" {
   description = "DigitalOcean Access Token"
 }
@@ -11,7 +20,7 @@ variable "TESTNET_NAME" {
 
 variable "SSH_KEY_FILE" {
   description = "SSH public key file to be used on the nodes"
-  type = "string"
+  type = string
 }
 
 variable "SERVERS" {

node/node.go

@@ -28,7 +28,6 @@ import (
 	"github.com/tendermint/tendermint/libs/service"
 	"github.com/tendermint/tendermint/libs/strings"
 	tmtime "github.com/tendermint/tendermint/libs/time"
-	"github.com/tendermint/tendermint/light"
 	"github.com/tendermint/tendermint/privval"
 	tmgrpc "github.com/tendermint/tendermint/privval/grpc"
 	"github.com/tendermint/tendermint/proxy"
@@ -221,7 +220,7 @@ func makeNode(config *cfg.Config,
 
 	// Determine whether we should do block sync. This must happen after the handshake, since the
 	// app may modify the validator set, specifying ourself as the only validator.
-	blockSync := config.FastSyncMode && !onlyValidatorIsUs(state, pubKey)
+	blockSync := config.BlockSync.Enable && !onlyValidatorIsUs(state, pubKey)
 
 	logNodeStartupInfo(state, pubKey, logger, consensusLogger, config.Mode)
 
@@ -319,15 +318,17 @@ func makeNode(config *cfg.Config,
 
 	stateSyncReactorShim = p2p.NewReactorShim(logger.With("module", "statesync"), "StateSyncShim", statesync.ChannelShims)
 
-	if config.P2P.DisableLegacy {
-		channels = makeChannelsFromShims(router, statesync.ChannelShims)
-		peerUpdates = peerManager.Subscribe()
-	} else {
+	if config.P2P.UseLegacy {
 		channels = getChannelsFromShim(stateSyncReactorShim)
 		peerUpdates = stateSyncReactorShim.PeerUpdates
+	} else {
+		channels = makeChannelsFromShims(router, statesync.ChannelShims)
+		peerUpdates = peerManager.Subscribe()
 	}
 
 	stateSyncReactor = statesync.NewReactor(
+		genDoc.ChainID,
+		genDoc.InitialHeight,
 		*config.StateSync,
 		stateSyncReactorShim.Logger,
 		proxyApp.Snapshot(),
@@ -335,6 +336,7 @@ func makeNode(config *cfg.Config,
 		channels[statesync.SnapshotChannel],
 		channels[statesync.ChunkChannel],
 		channels[statesync.LightBlockChannel],
+		channels[statesync.ParamsChannel],
 		peerUpdates,
 		stateStore,
 		blockStore,
@@ -373,13 +375,7 @@ func makeNode(config *cfg.Config,
 	pexCh := pex.ChannelDescriptor()
 	transport.AddChannelDescriptors([]*p2p.ChannelDescriptor{&pexCh})
 
-	if config.P2P.DisableLegacy {
-		addrBook = nil
-		pexReactor, err = createPEXReactorV2(config, logger, peerManager, router)
-		if err != nil {
-			return nil, err
-		}
-	} else {
+	if config.P2P.UseLegacy {
 		// setup Transport and Switch
 		sw = createSwitch(
 			config, transport, p2pMetrics, mpReactorShim, bcReactorForSwitch,
@@ -402,6 +398,12 @@ func makeNode(config *cfg.Config,
 		}
 
 		pexReactor = createPEXReactorAndAddToSwitch(addrBook, config, sw, logger)
+	} else {
+		addrBook = nil
+		pexReactor, err = createPEXReactorV2(config, logger, peerManager, router)
+		if err != nil {
+			return nil, err
+		}
 	}
 
 	if config.RPC.PprofListenAddress != "" {
@@ -441,25 +443,37 @@ func makeNode(config *cfg.Config,
 			ProxyAppQuery:   proxyApp.Query(),
 			ProxyAppMempool: proxyApp.Mempool(),
 
-			StateStore:       stateStore,
-			BlockStore:       blockStore,
-			EvidencePool:     evPool,
-			ConsensusState:   csState,
+			StateStore:     stateStore,
+			BlockStore:     blockStore,
+			EvidencePool:   evPool,
+			ConsensusState: csState,
+
+			ConsensusReactor: csReactor,
 			BlockSyncReactor: bcReactor.(cs.BlockSyncReactor),
 
 			P2PPeers:    sw,
 			PeerManager: peerManager,
 
-			GenDoc:           genDoc,
-			EventSinks:       eventSinks,
-			ConsensusReactor: csReactor,
-			EventBus:         eventBus,
-			Mempool:          mp,
-			Logger:           logger.With("module", "rpc"),
-			Config:           *config.RPC,
+			GenDoc:     genDoc,
+			EventSinks: eventSinks,
+			EventBus:   eventBus,
+			Mempool:    mp,
+			Logger:     logger.With("module", "rpc"),
+			Config:     *config.RPC,
 		},
 	}
 
+	// this is a terrible, because typed nil interfaces are not ==
+	// nil, so this is just cleanup to avoid having a non-nil
+	// value in the RPC environment that has the semantic
+	// properties of nil.
+	if sw == nil {
+		node.rpcEnv.P2PPeers = nil
+	} else if peerManager == nil {
+		node.rpcEnv.PeerManager = nil
+	}
+	// end hack
+
 	node.rpcEnv.P2PTransport = node
 
 	node.BaseService = *service.NewBaseService(logger, "Node", node)
@@ -518,12 +532,8 @@ func makeSeedNode(config *cfg.Config,
 	// p2p stack is removed.
 	pexCh := pex.ChannelDescriptor()
 	transport.AddChannelDescriptors([]*p2p.ChannelDescriptor{&pexCh})
-	if config.P2P.DisableLegacy {
-		pexReactor, err = createPEXReactorV2(config, logger, peerManager, router)
-		if err != nil {
-			return nil, err
-		}
-	} else {
+
+	if config.P2P.UseLegacy {
 		sw = createSwitch(
 			config, transport, p2pMetrics, nil, nil,
 			nil, nil, nil, nil, nodeInfo, nodeKey, p2pLogger,
@@ -545,6 +555,11 @@ func makeSeedNode(config *cfg.Config,
 		}
 
 		pexReactor = createPEXReactorAndAddToSwitch(addrBook, config, sw, logger)
+	} else {
+		pexReactor, err = createPEXReactorV2(config, logger, peerManager, router)
+		if err != nil {
+			return nil, err
+		}
 	}
 
 	if config.RPC.PprofListenAddress != "" {
@@ -607,18 +622,16 @@ func (n *nodeImpl) OnStart() error {
 	}
 
 	n.isListening = true
-	n.Logger.Info("p2p service", "legacy_enabled", !n.config.P2P.DisableLegacy)
+	n.Logger.Info("p2p service", "legacy_enabled", n.config.P2P.UseLegacy)
 
-	if n.config.P2P.DisableLegacy {
-		if err = n.router.Start(); err != nil {
-			return err
-		}
-	} else {
+	if n.config.P2P.UseLegacy {
 		// Add private IDs to addrbook to block those peers being added
 		n.addrBook.AddPrivateIDs(strings.SplitAndTrimEmpty(n.config.P2P.PrivatePeerIDs, ",", " "))
 		if err = n.sw.Start(); err != nil {
 			return err
 		}
+	} else if err = n.router.Start(); err != nil {
+		return err
 	}
 
 	if n.config.Mode != cfg.ModeSeed {
@@ -649,19 +662,19 @@ func (n *nodeImpl) OnStart() error {
 		}
 	}
 
-	if n.config.P2P.DisableLegacy {
-		if err := n.pexReactor.Start(); err != nil {
-			return err
-		}
-	} else {
+	if n.config.P2P.UseLegacy {
 		// Always connect to persistent peers
 		err = n.sw.DialPeersAsync(strings.SplitAndTrimEmpty(n.config.P2P.PersistentPeers, ",", " "))
 		if err != nil {
 			return fmt.Errorf("could not dial peers from persistent-peers field: %w", err)
 		}
+	} else if err := n.pexReactor.Start(); err != nil {
+		return err
 	}
 
 	// Run state sync
+	// TODO: We shouldn't run state sync if we already have state that has a
+	// LastBlockHeight that is not InitialHeight
 	if n.stateSync {
 		bcR, ok := n.bcReactor.(cs.BlockSyncReactor)
 		if !ok {
@@ -674,17 +687,52 @@ func (n *nodeImpl) OnStart() error {
 			return fmt.Errorf("unable to derive state: %w", err)
 		}
 
-		ssc := n.config.StateSync
-		sp, err := constructStateProvider(ssc, state, n.Logger.With("module", "light"))
-
-		if err != nil {
-			return fmt.Errorf("failed to set up light client state provider: %w", err)
+		// TODO: we may want to move these events within the respective
+		// reactors.
+		// At the beginning of the statesync start, we use the initialHeight as the event height
+		// because of the statesync doesn't have the concreate state height before fetched the snapshot.
+		d := types.EventDataStateSyncStatus{Complete: false, Height: state.InitialHeight}
+		if err := n.eventBus.PublishEventStateSyncStatus(d); err != nil {
+			n.eventBus.Logger.Error("failed to emit the statesync start event", "err", err)
 		}
 
-		if err := startStateSync(n.stateSyncReactor, bcR, n.consensusReactor, sp,
-			ssc, n.config.FastSyncMode, state.InitialHeight, n.eventBus); err != nil {
-			return fmt.Errorf("failed to start state sync: %w", err)
-		}
+		// FIXME: We shouldn't allow state sync to silently error out without
+		// bubbling up the error and gracefully shutting down the rest of the node
+		go func() {
+			n.Logger.Info("starting state sync")
+			state, err := n.stateSyncReactor.Sync(context.TODO())
+			if err != nil {
+				n.Logger.Error("state sync failed", "err", err)
+				return
+			}
+
+			n.consensusReactor.SetStateSyncingMetrics(0)
+
+			d := types.EventDataStateSyncStatus{Complete: true, Height: state.LastBlockHeight}
+			if err := n.eventBus.PublishEventStateSyncStatus(d); err != nil {
+				n.eventBus.Logger.Error("failed to emit the statesync start event", "err", err)
+			}
+
+			// TODO: Some form of orchestrator is needed here between the state
+			// advancing reactors to be able to control which one of the three
+			// is running
+			if n.config.BlockSync.Enable {
+				// FIXME Very ugly to have these metrics bleed through here.
+				n.consensusReactor.SetBlockSyncingMetrics(1)
+				if err := bcR.SwitchToBlockSync(state); err != nil {
+					n.Logger.Error("failed to switch to block sync", "err", err)
+					return
+				}
+
+				d := types.EventDataBlockSyncStatus{Complete: false, Height: state.LastBlockHeight}
+				if err := n.eventBus.PublishEventBlockSyncStatus(d); err != nil {
+					n.eventBus.Logger.Error("failed to emit the block sync starting event", "err", err)
+				}
+
+			} else {
+				n.consensusReactor.SwitchToConsensus(state, true)
+			}
+		}()
 	}
 
 	return nil
@@ -737,14 +785,14 @@ func (n *nodeImpl) OnStop() {
 		n.Logger.Error("failed to stop the PEX v2 reactor", "err", err)
 	}
 
-	if n.config.P2P.DisableLegacy {
-		if err := n.router.Stop(); err != nil {
-			n.Logger.Error("failed to stop router", "err", err)
-		}
-	} else {
+	if n.config.P2P.UseLegacy {
 		if err := n.sw.Stop(); err != nil {
 			n.Logger.Error("failed to stop switch", "err", err)
 		}
+	} else {
+		if err := n.router.Stop(); err != nil {
+			n.Logger.Error("failed to stop router", "err", err)
+		}
 	}
 
 	if err := n.transport.Close(); err != nil {
@@ -825,7 +873,7 @@ func (n *nodeImpl) startRPC() ([]net.Listener, error) {
 		rpcserver.RegisterRPCFuncs(mux, routes, rpcLogger)
 		listener, err := rpcserver.Listen(
 			listenAddr,
-			config,
+			config.MaxOpenConnections,
 		)
 		if err != nil {
 			return nil, err
@@ -883,7 +931,7 @@ func (n *nodeImpl) startRPC() ([]net.Listener, error) {
 		if config.WriteTimeout <= n.config.RPC.TimeoutBroadcastTxCommit {
 			config.WriteTimeout = n.config.RPC.TimeoutBroadcastTxCommit + 1*time.Second
 		}
-		listener, err := rpcserver.Listen(grpcListenAddr, config)
+		listener, err := rpcserver.Listen(grpcListenAddr, config.MaxOpenConnections)
 		if err != nil {
 			return nil, err
 		}
@@ -969,67 +1017,6 @@ func (n *nodeImpl) NodeInfo() types.NodeInfo {
 	return n.nodeInfo
 }
 
-// startStateSync starts an asynchronous state sync process, then switches to block sync mode.
-func startStateSync(
-	ssR statesync.SyncReactor,
-	bcR cs.BlockSyncReactor,
-	conR cs.ConsSyncReactor,
-	sp statesync.StateProvider,
-	config *cfg.StateSyncConfig,
-	blockSync bool,
-	stateInitHeight int64,
-	eb *types.EventBus,
-) error {
-	stateSyncLogger := eb.Logger.With("module", "statesync")
-
-	stateSyncLogger.Info("starting state sync...")
-
-	// at the beginning of the statesync start, we use the initialHeight as the event height
-	// because of the statesync doesn't have the concreate state height before fetched the snapshot.
-	d := types.EventDataStateSyncStatus{Complete: false, Height: stateInitHeight}
-	if err := eb.PublishEventStateSyncStatus(d); err != nil {
-		stateSyncLogger.Error("failed to emit the statesync start event", "err", err)
-	}
-
-	go func() {
-		state, err := ssR.Sync(context.TODO(), sp, config.DiscoveryTime)
-		if err != nil {
-			stateSyncLogger.Error("state sync failed", "err", err)
-			return
-		}
-
-		if err := ssR.Backfill(state); err != nil {
-			stateSyncLogger.Error("backfill failed; node has insufficient history to verify all evidence;"+
-				" proceeding optimistically...", "err", err)
-		}
-
-		conR.SetStateSyncingMetrics(0)
-
-		d := types.EventDataStateSyncStatus{Complete: true, Height: state.LastBlockHeight}
-		if err := eb.PublishEventStateSyncStatus(d); err != nil {
-			stateSyncLogger.Error("failed to emit the statesync start event", "err", err)
-		}
-
-		if blockSync {
-			// FIXME Very ugly to have these metrics bleed through here.
-			conR.SetBlockSyncingMetrics(1)
-			if err := bcR.SwitchToBlockSync(state); err != nil {
-				stateSyncLogger.Error("failed to switch to block sync", "err", err)
-				return
-			}
-
-			d := types.EventDataBlockSyncStatus{Complete: false, Height: state.LastBlockHeight}
-			if err := eb.PublishEventBlockSyncStatus(d); err != nil {
-				stateSyncLogger.Error("failed to emit the block sync starting event", "err", err)
-			}
-
-		} else {
-			conR.SwitchToConsensus(state, true)
-		}
-	}()
-	return nil
-}
-
 // genesisDocProvider returns a GenesisDoc.
 // It allows the GenesisDoc to be pulled from sources other than the
 // filesystem, for instance from a distributed key-value store cluster.
@@ -1212,24 +1199,3 @@ func getChannelsFromShim(reactorShim *p2p.ReactorShim) map[p2p.ChannelID]*p2p.Ch
 
 	return channels
 }
-
-func constructStateProvider(
-	ssc *cfg.StateSyncConfig,
-	state sm.State,
-	logger log.Logger,
-) (statesync.StateProvider, error) {
-	ctx, cancel := context.WithTimeout(context.TODO(), 10*time.Second)
-	defer cancel()
-
-	to := light.TrustOptions{
-		Period: ssc.TrustPeriod,
-		Height: ssc.TrustHeight,
-		Hash:   ssc.TrustHashBytes(),
-	}
-
-	return statesync.NewLightClientStateProvider(
-		ctx,
-		state.ChainID, state.Version, state.InitialHeight,
-		ssc.RPCServers, to, logger,
-	)
-}

node/node_test.go

@@ -21,16 +21,12 @@ import (
 	"github.com/tendermint/tendermint/crypto"
 	"github.com/tendermint/tendermint/crypto/ed25519"
 	"github.com/tendermint/tendermint/crypto/tmhash"
-	consmocks "github.com/tendermint/tendermint/internal/consensus/mocks"
-	ssmocks "github.com/tendermint/tendermint/internal/statesync/mocks"
 
 	"github.com/tendermint/tendermint/internal/evidence"
 	"github.com/tendermint/tendermint/internal/mempool"
 	mempoolv0 "github.com/tendermint/tendermint/internal/mempool/v0"
-	statesync "github.com/tendermint/tendermint/internal/statesync"
 	"github.com/tendermint/tendermint/internal/test/factory"
 	"github.com/tendermint/tendermint/libs/log"
-	tmpubsub "github.com/tendermint/tendermint/libs/pubsub"
 	tmrand "github.com/tendermint/tendermint/libs/rand"
 	tmtime "github.com/tendermint/tendermint/libs/time"
 	"github.com/tendermint/tendermint/privval"
@@ -43,6 +39,7 @@ import (
 
 func TestNodeStartStop(t *testing.T) {
 	config := cfg.ResetTestRoot("node_node_test")
+
 	defer os.RemoveAll(config.RootDir)
 
 	// create & start node
@@ -53,8 +50,6 @@ func TestNodeStartStop(t *testing.T) {
 	n, ok := ns.(*nodeImpl)
 	require.True(t, ok)
 
-	t.Logf("Started node %v", n.sw.NodeInfo())
-
 	// wait for the node to produce a block
 	blocksSub, err := n.EventBus().Subscribe(context.Background(), "node_test", types.EventQueryNewBlock)
 	require.NoError(t, err)
@@ -670,65 +665,3 @@ func loadStatefromGenesis(t *testing.T) sm.State {
 
 	return state
 }
-
-func TestNodeStartStateSync(t *testing.T) {
-	mockSSR := &statesync.MockSyncReactor{}
-	mockFSR := &consmocks.BlockSyncReactor{}
-	mockCSR := &consmocks.ConsSyncReactor{}
-	mockSP := &ssmocks.StateProvider{}
-	state := loadStatefromGenesis(t)
-	config := cfg.ResetTestRoot("load_state_from_genesis")
-
-	eventBus, err := createAndStartEventBus(log.TestingLogger())
-	defer func() {
-		err := eventBus.Stop()
-		require.NoError(t, err)
-	}()
-
-	require.NoError(t, err)
-	require.NotNil(t, eventBus)
-
-	sub, err := eventBus.Subscribe(context.Background(), "test-client", types.EventQueryStateSyncStatus, 10)
-	require.NoError(t, err)
-	require.NotNil(t, sub)
-
-	cfgSS := config.StateSync
-
-	mockSSR.On("Sync", context.TODO(), mockSP, cfgSS.DiscoveryTime).Return(state, nil).
-		On("Backfill", state).Return(nil)
-	mockCSR.On("SetStateSyncingMetrics", float64(0)).Return().
-		On("SwitchToConsensus", state, true).Return()
-
-	require.NoError(t,
-		startStateSync(mockSSR, mockFSR, mockCSR, mockSP, config.StateSync, false, state.InitialHeight, eventBus))
-
-	for cnt := 0; cnt < 2; {
-		select {
-		case <-time.After(3 * time.Second):
-			t.Errorf("StateSyncStatus timeout")
-		case msg := <-sub.Out():
-			if cnt == 0 {
-				ensureStateSyncStatus(t, msg, false, state.InitialHeight)
-				cnt++
-			} else {
-				// the state height = 0 because we are not actually update the state in this test
-				ensureStateSyncStatus(t, msg, true, 0)
-				cnt++
-			}
-		}
-	}
-
-	mockSSR.AssertNumberOfCalls(t, "Sync", 1)
-	mockSSR.AssertNumberOfCalls(t, "Backfill", 1)
-	mockCSR.AssertNumberOfCalls(t, "SetStateSyncingMetrics", 1)
-	mockCSR.AssertNumberOfCalls(t, "SwitchToConsensus", 1)
-}
-
-func ensureStateSyncStatus(t *testing.T, msg tmpubsub.Message, complete bool, height int64) {
-	t.Helper()
-	status, ok := msg.Data().(types.EventDataStateSyncStatus)
-
-	require.True(t, ok)
-	require.Equal(t, complete, status.Complete)
-	require.Equal(t, height, status.Height)
-}

node/setup.go

@@ -8,7 +8,6 @@ import (
 	"math"
 	"net"
 	_ "net/http/pprof" // nolint: gosec // securely exposed on separate, optional port
-	"strings"
 	"time"
 
 	dbm "github.com/tendermint/tm-db"
@@ -33,9 +32,7 @@ import (
 	"github.com/tendermint/tendermint/proxy"
 	sm "github.com/tendermint/tendermint/state"
 	"github.com/tendermint/tendermint/state/indexer"
-	kv "github.com/tendermint/tendermint/state/indexer/sink/kv"
-	null "github.com/tendermint/tendermint/state/indexer/sink/null"
-	psql "github.com/tendermint/tendermint/state/indexer/sink/psql"
+	"github.com/tendermint/tendermint/state/indexer/sink"
 	"github.com/tendermint/tendermint/store"
 	"github.com/tendermint/tendermint/types"
 	"github.com/tendermint/tendermint/version"
@@ -78,56 +75,9 @@ func createAndStartIndexerService(
 	logger log.Logger,
 	chainID string,
 ) (*indexer.Service, []indexer.EventSink, error) {
-
-	eventSinks := []indexer.EventSink{}
-
-	// check for duplicated sinks
-	sinks := map[string]bool{}
-	for _, s := range config.TxIndex.Indexer {
-		sl := strings.ToLower(s)
-		if sinks[sl] {
-			return nil, nil, errors.New("found duplicated sinks, please check the tx-index section in the config.toml")
-		}
-
-		sinks[sl] = true
-	}
-
-loop:
-	for k := range sinks {
-		switch k {
-		case string(indexer.NULL):
-			// When we see null in the config, the eventsinks will be reset with the
-			// nullEventSink.
-			eventSinks = []indexer.EventSink{null.NewEventSink()}
-			break loop
-
-		case string(indexer.KV):
-			store, err := dbProvider(&cfg.DBContext{ID: "tx_index", Config: config})
-			if err != nil {
-				return nil, nil, err
-			}
-
-			eventSinks = append(eventSinks, kv.NewEventSink(store))
-
-		case string(indexer.PSQL):
-			conn := config.TxIndex.PsqlConn
-			if conn == "" {
-				return nil, nil, errors.New("the psql connection settings cannot be empty")
-			}
-
-			es, _, err := psql.NewEventSink(conn, chainID)
-			if err != nil {
-				return nil, nil, err
-			}
-			eventSinks = append(eventSinks, es)
-
-		default:
-			return nil, nil, errors.New("unsupported event sink type")
-		}
-	}
-
-	if len(eventSinks) == 0 {
-		eventSinks = []indexer.EventSink{null.NewEventSink()}
+	eventSinks, err := sink.EventSinksFromConfig(config, dbProvider, chainID)
+	if err != nil {
+		return nil, nil, err
 	}
 
 	indexerService := indexer.NewIndexerService(eventSinks, eventBus)
@@ -216,12 +166,12 @@ func createMempoolReactor(
 		peerUpdates *p2p.PeerUpdates
 	)
 
-	if config.P2P.DisableLegacy {
-		channels = makeChannelsFromShims(router, channelShims)
-		peerUpdates = peerManager.Subscribe()
-	} else {
+	if config.P2P.UseLegacy {
 		channels = getChannelsFromShim(reactorShim)
 		peerUpdates = reactorShim.PeerUpdates
+	} else {
+		channels = makeChannelsFromShims(router, channelShims)
+		peerUpdates = peerManager.Subscribe()
 	}
 
 	switch config.Mempool.Version {
@@ -310,12 +260,12 @@ func createEvidenceReactor(
 		peerUpdates *p2p.PeerUpdates
 	)
 
-	if config.P2P.DisableLegacy {
-		channels = makeChannelsFromShims(router, evidence.ChannelShims)
-		peerUpdates = peerManager.Subscribe()
-	} else {
+	if config.P2P.UseLegacy {
 		channels = getChannelsFromShim(reactorShim)
 		peerUpdates = reactorShim.PeerUpdates
+	} else {
+		channels = makeChannelsFromShims(router, evidence.ChannelShims)
+		peerUpdates = peerManager.Subscribe()
 	}
 
 	evidenceReactor := evidence.NewReactor(
@@ -352,12 +302,12 @@ func createBlockchainReactor(
 			peerUpdates *p2p.PeerUpdates
 		)
 
-		if config.P2P.DisableLegacy {
-			channels = makeChannelsFromShims(router, bcv0.ChannelShims)
-			peerUpdates = peerManager.Subscribe()
-		} else {
+		if config.P2P.UseLegacy {
 			channels = getChannelsFromShim(reactorShim)
 			peerUpdates = reactorShim.PeerUpdates
+		} else {
+			channels = makeChannelsFromShims(router, bcv0.ChannelShims)
+			peerUpdates = peerManager.Subscribe()
 		}
 
 		reactor, err := bcv0.NewReactor(
@@ -416,12 +366,12 @@ func createConsensusReactor(
 		peerUpdates *p2p.PeerUpdates
 	)
 
-	if config.P2P.DisableLegacy {
-		channels = makeChannelsFromShims(router, cs.ChannelShims)
-		peerUpdates = peerManager.Subscribe()
-	} else {
+	if config.P2P.UseLegacy {
 		channels = getChannelsFromShim(reactorShim)
 		peerUpdates = reactorShim.PeerUpdates
+	} else {
+		channels = makeChannelsFromShims(router, cs.ChannelShims)
+		peerUpdates = peerManager.Subscribe()
 	}
 
 	reactor := cs.NewReactor(
@@ -756,6 +706,7 @@ func makeNodeInfo(
 			byte(statesync.SnapshotChannel),
 			byte(statesync.ChunkChannel),
 			byte(statesync.LightBlockChannel),
+			byte(statesync.ParamsChannel),
 		},
 		Moniker: config.Moniker,
 		Other: types.NodeInfoOther{

proto/tendermint/statesync/message.go

@@ -28,6 +28,12 @@ func (m *Message) Wrap(pb proto.Message) error {
 	case *LightBlockResponse:
 		m.Sum = &Message_LightBlockResponse{LightBlockResponse: msg}
 
+	case *ParamsRequest:
+		m.Sum = &Message_ParamsRequest{ParamsRequest: msg}
+
+	case *ParamsResponse:
+		m.Sum = &Message_ParamsResponse{ParamsResponse: msg}
+
 	default:
 		return fmt.Errorf("unknown message: %T", msg)
 	}
@@ -57,6 +63,12 @@ func (m *Message) Unwrap() (proto.Message, error) {
 	case *Message_LightBlockResponse:
 		return m.GetLightBlockResponse(), nil
 
+	case *Message_ParamsRequest:
+		return m.GetParamsRequest(), nil
+
+	case *Message_ParamsResponse:
+		return m.GetParamsResponse(), nil
+
 	default:
 		return nil, fmt.Errorf("unknown message: %T", msg)
 	}
@@ -106,6 +118,17 @@ func (m *Message) Validate() error {
 	// light block validation handled by the backfill process
 	case *Message_LightBlockResponse:
 
+	case *Message_ParamsRequest:
+		if m.GetParamsRequest().Height == 0 {
+			return errors.New("height cannot be 0")
+		}
+
+	case *Message_ParamsResponse:
+		resp := m.GetParamsResponse()
+		if resp.Height == 0 {
+			return errors.New("height cannot be 0")
+		}
+
 	default:
 		return fmt.Errorf("unknown message type: %T", msg)
 	}

proto/tendermint/statesync/message_test.go

@@ -9,6 +9,7 @@ import (
 
 	ssproto "github.com/tendermint/tendermint/proto/tendermint/statesync"
 	tmproto "github.com/tendermint/tendermint/proto/tendermint/types"
+	"github.com/tendermint/tendermint/types"
 )
 
 func TestValidateMsg(t *testing.T) {
@@ -161,6 +162,35 @@ func TestStateSyncVectors(t *testing.T) {
 			},
 			"2214080110021803220c697427732061206368756e6b",
 		},
+		{
+			"LightBlockRequest",
+			&ssproto.LightBlockRequest{
+				Height: 100,
+			},
+			"2a020864",
+		},
+		{
+			"LightBlockResponse",
+			&ssproto.LightBlockResponse{
+				LightBlock: nil,
+			},
+			"3200",
+		},
+		{
+			"ParamsRequest",
+			&ssproto.ParamsRequest{
+				Height: 9001,
+			},
+			"3a0308a946",
+		},
+		{
+			"ParamsResponse",
+			&ssproto.ParamsResponse{
+				Height:          9001,
+				ConsensusParams: types.DefaultConsensusParams().ToProto(),
+			},
+			"423408a946122f0a10088080c00a10ffffffffffffffffff01120e08a08d0612040880c60a188080401a090a07656432353531392200",
+		},
 	}
 
 	for _, tc := range testCases {

proto/tendermint/statesync/types.pb.go

@@ -5,6 +5,7 @@ package statesync
 
 import (
 	fmt "fmt"
+	_ "github.com/gogo/protobuf/gogoproto"
 	proto "github.com/gogo/protobuf/proto"
 	types "github.com/tendermint/tendermint/proto/tendermint/types"
 	io "io"
@@ -31,6 +32,8 @@ type Message struct {
 	//	*Message_ChunkResponse
 	//	*Message_LightBlockRequest
 	//	*Message_LightBlockResponse
+	//	*Message_ParamsRequest
+	//	*Message_ParamsResponse
 	Sum isMessage_Sum `protobuf_oneof:"sum"`
 }
 
@@ -91,6 +94,12 @@ type Message_LightBlockRequest struct {
 type Message_LightBlockResponse struct {
 	LightBlockResponse *LightBlockResponse `protobuf:"bytes,6,opt,name=light_block_response,json=lightBlockResponse,proto3,oneof" json:"light_block_response,omitempty"`
 }
+type Message_ParamsRequest struct {
+	ParamsRequest *ParamsRequest `protobuf:"bytes,7,opt,name=params_request,json=paramsRequest,proto3,oneof" json:"params_request,omitempty"`
+}
+type Message_ParamsResponse struct {
+	ParamsResponse *ParamsResponse `protobuf:"bytes,8,opt,name=params_response,json=paramsResponse,proto3,oneof" json:"params_response,omitempty"`
+}
 
 func (*Message_SnapshotsRequest) isMessage_Sum()   {}
 func (*Message_SnapshotsResponse) isMessage_Sum()  {}
@@ -98,6 +107,8 @@ func (*Message_ChunkRequest) isMessage_Sum()       {}
 func (*Message_ChunkResponse) isMessage_Sum()      {}
 func (*Message_LightBlockRequest) isMessage_Sum()  {}
 func (*Message_LightBlockResponse) isMessage_Sum() {}
+func (*Message_ParamsRequest) isMessage_Sum()      {}
+func (*Message_ParamsResponse) isMessage_Sum()     {}
 
 func (m *Message) GetSum() isMessage_Sum {
 	if m != nil {
@@ -148,6 +159,20 @@ func (m *Message) GetLightBlockResponse() *LightBlockResponse {
 	return nil
 }
 
+func (m *Message) GetParamsRequest() *ParamsRequest {
+	if x, ok := m.GetSum().(*Message_ParamsRequest); ok {
+		return x.ParamsRequest
+	}
+	return nil
+}
+
+func (m *Message) GetParamsResponse() *ParamsResponse {
+	if x, ok := m.GetSum().(*Message_ParamsResponse); ok {
+		return x.ParamsResponse
+	}
+	return nil
+}
+
 // XXX_OneofWrappers is for the internal use of the proto package.
 func (*Message) XXX_OneofWrappers() []interface{} {
 	return []interface{}{
@@ -157,6 +182,8 @@ func (*Message) XXX_OneofWrappers() []interface{} {
 		(*Message_ChunkResponse)(nil),
 		(*Message_LightBlockRequest)(nil),
 		(*Message_LightBlockResponse)(nil),
+		(*Message_ParamsRequest)(nil),
+		(*Message_ParamsResponse)(nil),
 	}
 }
 
@@ -496,6 +523,102 @@ func (m *LightBlockResponse) GetLightBlock() *types.LightBlock {
 	return nil
 }
 
+type ParamsRequest struct {
+	Height uint64 `protobuf:"varint,1,opt,name=height,proto3" json:"height,omitempty"`
+}
+
+func (m *ParamsRequest) Reset()         { *m = ParamsRequest{} }
+func (m *ParamsRequest) String() string { return proto.CompactTextString(m) }
+func (*ParamsRequest) ProtoMessage()    {}
+func (*ParamsRequest) Descriptor() ([]byte, []int) {
+	return fileDescriptor_a1c2869546ca7914, []int{7}
+}
+func (m *ParamsRequest) XXX_Unmarshal(b []byte) error {
+	return m.Unmarshal(b)
+}
+func (m *ParamsRequest) XXX_Marshal(b []byte, deterministic bool) ([]byte, error) {
+	if deterministic {
+		return xxx_messageInfo_ParamsRequest.Marshal(b, m, deterministic)
+	} else {
+		b = b[:cap(b)]
+		n, err := m.MarshalToSizedBuffer(b)
+		if err != nil {
+			return nil, err
+		}
+		return b[:n], nil
+	}
+}
+func (m *ParamsRequest) XXX_Merge(src proto.Message) {
+	xxx_messageInfo_ParamsRequest.Merge(m, src)
+}
+func (m *ParamsRequest) XXX_Size() int {
+	return m.Size()
+}
+func (m *ParamsRequest) XXX_DiscardUnknown() {
+	xxx_messageInfo_ParamsRequest.DiscardUnknown(m)
+}
+
+var xxx_messageInfo_ParamsRequest proto.InternalMessageInfo
+
+func (m *ParamsRequest) GetHeight() uint64 {
+	if m != nil {
+		return m.Height
+	}
+	return 0
+}
+
+type ParamsResponse struct {
+	Height          uint64                `protobuf:"varint,1,opt,name=height,proto3" json:"height,omitempty"`
+	ConsensusParams types.ConsensusParams `protobuf:"bytes,2,opt,name=consensus_params,json=consensusParams,proto3" json:"consensus_params"`
+}
+
+func (m *ParamsResponse) Reset()         { *m = ParamsResponse{} }
+func (m *ParamsResponse) String() string { return proto.CompactTextString(m) }
+func (*ParamsResponse) ProtoMessage()    {}
+func (*ParamsResponse) Descriptor() ([]byte, []int) {
+	return fileDescriptor_a1c2869546ca7914, []int{8}
+}
+func (m *ParamsResponse) XXX_Unmarshal(b []byte) error {
+	return m.Unmarshal(b)
+}
+func (m *ParamsResponse) XXX_Marshal(b []byte, deterministic bool) ([]byte, error) {
+	if deterministic {
+		return xxx_messageInfo_ParamsResponse.Marshal(b, m, deterministic)
+	} else {
+		b = b[:cap(b)]
+		n, err := m.MarshalToSizedBuffer(b)
+		if err != nil {
+			return nil, err
+		}
+		return b[:n], nil
+	}
+}
+func (m *ParamsResponse) XXX_Merge(src proto.Message) {
+	xxx_messageInfo_ParamsResponse.Merge(m, src)
+}
+func (m *ParamsResponse) XXX_Size() int {
+	return m.Size()
+}
+func (m *ParamsResponse) XXX_DiscardUnknown() {
+	xxx_messageInfo_ParamsResponse.DiscardUnknown(m)
+}
+
+var xxx_messageInfo_ParamsResponse proto.InternalMessageInfo
+
+func (m *ParamsResponse) GetHeight() uint64 {
+	if m != nil {
+		return m.Height
+	}
+	return 0
+}
+
+func (m *ParamsResponse) GetConsensusParams() types.ConsensusParams {
+	if m != nil {
+		return m.ConsensusParams
+	}
+	return types.ConsensusParams{}
+}
+
 func init() {
 	proto.RegisterType((*Message)(nil), "tendermint.statesync.Message")
 	proto.RegisterType((*SnapshotsRequest)(nil), "tendermint.statesync.SnapshotsRequest")
@@ -504,43 +627,51 @@ func init() {
 	proto.RegisterType((*ChunkResponse)(nil), "tendermint.statesync.ChunkResponse")
 	proto.RegisterType((*LightBlockRequest)(nil), "tendermint.statesync.LightBlockRequest")
 	proto.RegisterType((*LightBlockResponse)(nil), "tendermint.statesync.LightBlockResponse")
+	proto.RegisterType((*ParamsRequest)(nil), "tendermint.statesync.ParamsRequest")
+	proto.RegisterType((*ParamsResponse)(nil), "tendermint.statesync.ParamsResponse")
 }
 
 func init() { proto.RegisterFile("tendermint/statesync/types.proto", fileDescriptor_a1c2869546ca7914) }
 
 var fileDescriptor_a1c2869546ca7914 = []byte{
-	// 485 bytes of a gzipped FileDescriptorProto
-	0x1f, 0x8b, 0x08, 0x00, 0x00, 0x00, 0x00, 0x00, 0x02, 0xff, 0xac, 0x54, 0x51, 0x6b, 0xd3, 0x50,
-	0x14, 0x4e, 0x5c, 0xdb, 0x8d, 0xb3, 0x46, 0x96, 0x63, 0x91, 0x32, 0x46, 0x18, 0x11, 0x74, 0x20,
-	0xa4, 0xa0, 0x8f, 0xe2, 0x4b, 0x7d, 0x99, 0x30, 0x5f, 0xee, 0x1c, 0xa8, 0x08, 0x23, 0x4d, 0xaf,
-	0x4d, 0xb0, 0x49, 0x6a, 0xcf, 0x2d, 0xb8, 0x1f, 0xe0, 0x93, 0x2f, 0x82, 0x7f, 0xca, 0xc7, 0x3d,
-	0xfa, 0x28, 0xed, 0x1f, 0x91, 0x9c, 0xdc, 0x26, 0x77, 0x6d, 0x5d, 0x11, 0xf6, 0x96, 0xef, 0xeb,
-	0x77, 0x3e, 0xbe, 0x73, 0xcf, 0xe9, 0x81, 0x63, 0x25, 0xb3, 0xa1, 0x9c, 0xa6, 0x49, 0xa6, 0x7a,
-	0xa4, 0x42, 0x25, 0xe9, 0x2a, 0x8b, 0x7a, 0xea, 0x6a, 0x22, 0x29, 0x98, 0x4c, 0x73, 0x95, 0x63,
-	0xa7, 0x56, 0x04, 0x95, 0xe2, 0xf0, 0xc8, 0xa8, 0x63, 0xb5, 0x59, 0xe3, 0xff, 0x6c, 0xc0, 0xee,
-	0x1b, 0x49, 0x14, 0x8e, 0x24, 0x5e, 0x80, 0x4b, 0x59, 0x38, 0xa1, 0x38, 0x57, 0x74, 0x39, 0x95,
-	0x5f, 0x66, 0x92, 0x54, 0xd7, 0x3e, 0xb6, 0x4f, 0xf6, 0x9f, 0x3d, 0x0e, 0x36, 0x79, 0x07, 0xe7,
-	0x4b, 0xb9, 0x28, 0xd5, 0xa7, 0x96, 0x38, 0xa0, 0x15, 0x0e, 0xdf, 0x01, 0x9a, 0xb6, 0x34, 0xc9,
-	0x33, 0x92, 0xdd, 0x7b, 0xec, 0xfb, 0x64, 0xab, 0x6f, 0x29, 0x3f, 0xb5, 0x84, 0x4b, 0xab, 0x24,
-	0xbe, 0x06, 0x27, 0x8a, 0x67, 0xd9, 0xe7, 0x2a, 0xec, 0x0e, 0x9b, 0xfa, 0x9b, 0x4d, 0x5f, 0x15,
-	0xd2, 0x3a, 0x68, 0x3b, 0x32, 0x30, 0x9e, 0xc1, 0xfd, 0xa5, 0x95, 0x0e, 0xd8, 0x60, 0xaf, 0x47,
-	0xb7, 0x7a, 0x55, 0xe1, 0x9c, 0xc8, 0x24, 0xf0, 0x3d, 0x3c, 0x18, 0x27, 0xa3, 0x58, 0x5d, 0x0e,
-	0xc6, 0x79, 0x54, 0xc7, 0x6b, 0xde, 0xd6, 0xf3, 0x59, 0x51, 0xd0, 0x2f, 0xf4, 0x75, 0x46, 0x77,
-	0xbc, 0x4a, 0xe2, 0x47, 0xe8, 0xdc, 0xb4, 0xd6, 0x71, 0x5b, 0xec, 0x7d, 0xb2, 0xdd, 0xbb, 0xca,
-	0x8c, 0xe3, 0x35, 0xb6, 0xdf, 0x84, 0x1d, 0x9a, 0xa5, 0x3e, 0xc2, 0xc1, 0xea, 0x68, 0xfd, 0xef,
-	0x36, 0xb8, 0x6b, 0x73, 0xc1, 0x87, 0xd0, 0x8a, 0x65, 0xe1, 0xc3, 0x8b, 0xd2, 0x10, 0x1a, 0x15,
-	0xfc, 0xa7, 0x7c, 0x9a, 0x86, 0x8a, 0x07, 0xed, 0x08, 0x8d, 0x0a, 0x9e, 0x9f, 0x8a, 0x78, 0x56,
-	0x8e, 0xd0, 0x08, 0x11, 0x1a, 0x71, 0x48, 0x31, 0xbf, 0x7a, 0x5b, 0xf0, 0x37, 0x1e, 0xc2, 0x5e,
-	0x2a, 0x55, 0x38, 0x0c, 0x55, 0xc8, 0x4f, 0xd7, 0x16, 0x15, 0xf6, 0xdf, 0x42, 0xdb, 0x9c, 0xe7,
-	0x7f, 0xe7, 0xe8, 0x40, 0x33, 0xc9, 0x86, 0xf2, 0xab, 0x8e, 0x51, 0x02, 0xff, 0x9b, 0x0d, 0xce,
-	0x8d, 0xd1, 0xde, 0x8d, 0x6f, 0xc1, 0x72, 0x9f, 0xba, 0xbd, 0x12, 0x60, 0x17, 0x76, 0xd3, 0x84,
-	0x28, 0xc9, 0x46, 0xdc, 0xde, 0x9e, 0x58, 0x42, 0xff, 0x29, 0xb8, 0x6b, 0xeb, 0xf0, 0xaf, 0x28,
-	0xfe, 0x39, 0xe0, 0xfa, 0x7c, 0xf1, 0x25, 0xec, 0x1b, 0x7b, 0xa2, 0xff, 0xc6, 0x47, 0xe6, 0x7a,
-	0x94, 0x67, 0xc0, 0x28, 0x85, 0x7a, 0x21, 0xfa, 0x17, 0xbf, 0xe6, 0x9e, 0x7d, 0x3d, 0xf7, 0xec,
-	0x3f, 0x73, 0xcf, 0xfe, 0xb1, 0xf0, 0xac, 0xeb, 0x85, 0x67, 0xfd, 0x5e, 0x78, 0xd6, 0x87, 0x17,
-	0xa3, 0x44, 0xc5, 0xb3, 0x41, 0x10, 0xe5, 0x69, 0xcf, 0x3c, 0x2d, 0xf5, 0x27, 0x5f, 0x96, 0xde,
-	0xa6, 0x73, 0x35, 0x68, 0xf1, 0x6f, 0xcf, 0xff, 0x06, 0x00, 0x00, 0xff, 0xff, 0xc1, 0x45, 0x35,
-	0xee, 0xcd, 0x04, 0x00, 0x00,
+	// 589 bytes of a gzipped FileDescriptorProto
+	0x1f, 0x8b, 0x08, 0x00, 0x00, 0x00, 0x00, 0x00, 0x02, 0xff, 0xac, 0x95, 0x4f, 0x8b, 0xd3, 0x40,
+	0x18, 0xc6, 0x13, 0xb7, 0xdd, 0x96, 0x77, 0x9b, 0x6e, 0x3b, 0x16, 0x29, 0x65, 0x8d, 0x6b, 0x14,
+	0x77, 0x41, 0x68, 0x41, 0x8f, 0xe2, 0xa5, 0x7b, 0x59, 0x61, 0x45, 0x99, 0x75, 0x41, 0x45, 0x28,
+	0x69, 0x3a, 0x26, 0xc1, 0xe6, 0x8f, 0x7d, 0xa7, 0xe0, 0x82, 0x57, 0x4f, 0x5e, 0xfc, 0x2c, 0x7e,
+	0x8a, 0x3d, 0xee, 0xd1, 0x93, 0x48, 0xfb, 0x45, 0x24, 0x93, 0x69, 0x32, 0x6d, 0xda, 0x2e, 0x82,
+	0xb7, 0xbc, 0xcf, 0x3c, 0xf9, 0xf5, 0x99, 0xc9, 0xc3, 0x14, 0x0e, 0x39, 0x0b, 0x47, 0x6c, 0x12,
+	0xf8, 0x21, 0xef, 0x21, 0xb7, 0x39, 0xc3, 0xcb, 0xd0, 0xe9, 0xf1, 0xcb, 0x98, 0x61, 0x37, 0x9e,
+	0x44, 0x3c, 0x22, 0xad, 0xdc, 0xd1, 0xcd, 0x1c, 0x9d, 0x96, 0x1b, 0xb9, 0x91, 0x30, 0xf4, 0x92,
+	0xa7, 0xd4, 0xdb, 0x39, 0x50, 0x68, 0x82, 0xa1, 0x92, 0x3a, 0x77, 0x0b, 0xab, 0xb1, 0x3d, 0xb1,
+	0x03, 0xb9, 0x6c, 0xfd, 0x2c, 0x43, 0xe5, 0x25, 0x43, 0xb4, 0x5d, 0x46, 0x2e, 0xa0, 0x89, 0xa1,
+	0x1d, 0xa3, 0x17, 0x71, 0x1c, 0x4c, 0xd8, 0xe7, 0x29, 0x43, 0xde, 0xd6, 0x0f, 0xf5, 0xe3, 0xbd,
+	0x27, 0x8f, 0xba, 0xeb, 0x02, 0x75, 0xcf, 0x17, 0x76, 0x9a, 0xba, 0x4f, 0x35, 0xda, 0xc0, 0x15,
+	0x8d, 0xbc, 0x05, 0xa2, 0x62, 0x31, 0x8e, 0x42, 0x64, 0xed, 0x5b, 0x82, 0x7b, 0x74, 0x23, 0x37,
+	0xb5, 0x9f, 0x6a, 0xb4, 0x89, 0xab, 0x22, 0x79, 0x01, 0x86, 0xe3, 0x4d, 0xc3, 0x4f, 0x59, 0xd8,
+	0x1d, 0x01, 0xb5, 0xd6, 0x43, 0x4f, 0x12, 0x6b, 0x1e, 0xb4, 0xe6, 0x28, 0x33, 0x39, 0x83, 0xfa,
+	0x02, 0x25, 0x03, 0x96, 0x04, 0xeb, 0xc1, 0x56, 0x56, 0x16, 0xce, 0x70, 0x54, 0x81, 0xbc, 0x83,
+	0xdb, 0x63, 0xdf, 0xf5, 0xf8, 0x60, 0x38, 0x8e, 0x9c, 0x3c, 0x5e, 0x79, 0xdb, 0x9e, 0xcf, 0x92,
+	0x17, 0xfa, 0x89, 0x3f, 0xcf, 0xd8, 0x1c, 0xaf, 0x8a, 0xe4, 0x03, 0xb4, 0x96, 0xd1, 0x32, 0xee,
+	0xae, 0x60, 0x1f, 0xdf, 0xcc, 0xce, 0x32, 0x93, 0x71, 0x41, 0x4d, 0x8e, 0x21, 0xad, 0x47, 0x96,
+	0xb9, 0xb2, 0xed, 0x18, 0x5e, 0x0b, 0x6f, 0x9e, 0xd7, 0x88, 0x55, 0x81, 0xbc, 0x82, 0xfd, 0x8c,
+	0x26, 0x63, 0x56, 0x05, 0xee, 0xe1, 0x76, 0x5c, 0x16, 0xb1, 0x1e, 0x2f, 0x29, 0xfd, 0x32, 0xec,
+	0xe0, 0x34, 0xb0, 0x08, 0x34, 0x56, 0x9b, 0x67, 0x7d, 0xd7, 0xa1, 0x59, 0xa8, 0x0d, 0xb9, 0x03,
+	0xbb, 0x1e, 0x4b, 0xb6, 0x29, 0x7a, 0x5c, 0xa2, 0x72, 0x4a, 0xf4, 0x8f, 0xd1, 0x24, 0xb0, 0xb9,
+	0xe8, 0xa1, 0x41, 0xe5, 0x94, 0xe8, 0xe2, 0x4b, 0xa2, 0xa8, 0x92, 0x41, 0xe5, 0x44, 0x08, 0x94,
+	0x3c, 0x1b, 0x3d, 0x51, 0x8a, 0x1a, 0x15, 0xcf, 0xa4, 0x03, 0xd5, 0x80, 0x71, 0x7b, 0x64, 0x73,
+	0x5b, 0x7c, 0xd9, 0x1a, 0xcd, 0x66, 0xeb, 0x0d, 0xd4, 0xd4, 0xba, 0xfd, 0x73, 0x8e, 0x16, 0x94,
+	0xfd, 0x70, 0xc4, 0xbe, 0xc8, 0x18, 0xe9, 0x60, 0x7d, 0xd3, 0xc1, 0x58, 0x6a, 0xde, 0xff, 0xe1,
+	0x26, 0xaa, 0xd8, 0xa7, 0xdc, 0x5e, 0x3a, 0x90, 0x36, 0x54, 0x02, 0x1f, 0xd1, 0x0f, 0x5d, 0xb1,
+	0xbd, 0x2a, 0x5d, 0x8c, 0xd6, 0x63, 0x68, 0x16, 0xda, 0xba, 0x29, 0x8a, 0x75, 0x0e, 0xa4, 0x58,
+	0x3f, 0xf2, 0x1c, 0xf6, 0x94, 0x1a, 0xcb, 0x5b, 0xe6, 0x40, 0xad, 0x45, 0x7a, 0x89, 0x29, 0xaf,
+	0x42, 0xde, 0x57, 0xeb, 0x08, 0x8c, 0xa5, 0xee, 0x6d, 0xfc, 0xf5, 0xaf, 0x50, 0x5f, 0x6e, 0xd5,
+	0xc6, 0x23, 0xa3, 0xd0, 0x70, 0x12, 0x43, 0x88, 0x53, 0x1c, 0xa4, 0xbd, 0x93, 0x97, 0xd4, 0xfd,
+	0x62, 0xac, 0x93, 0x85, 0x33, 0x85, 0xf7, 0x4b, 0x57, 0xbf, 0xef, 0x69, 0x74, 0xdf, 0x59, 0x91,
+	0x2f, 0xae, 0x66, 0xa6, 0x7e, 0x3d, 0x33, 0xf5, 0x3f, 0x33, 0x53, 0xff, 0x31, 0x37, 0xb5, 0xeb,
+	0xb9, 0xa9, 0xfd, 0x9a, 0x9b, 0xda, 0xfb, 0x67, 0xae, 0xcf, 0xbd, 0xe9, 0xb0, 0xeb, 0x44, 0x41,
+	0x4f, 0xbd, 0xa1, 0xf3, 0xc7, 0xf4, 0x9e, 0x5f, 0xf7, 0x4f, 0x31, 0xdc, 0x15, 0x6b, 0x4f, 0xff,
+	0x06, 0x00, 0x00, 0xff, 0xff, 0xa1, 0xb2, 0xfd, 0x65, 0x48, 0x06, 0x00, 0x00,
 }
 
 func (m *Message) Marshal() (dAtA []byte, err error) {
@@ -701,6 +832,48 @@ func (m *Message_LightBlockResponse) MarshalToSizedBuffer(dAtA []byte) (int, err
 	}
 	return len(dAtA) - i, nil
 }
+func (m *Message_ParamsRequest) MarshalTo(dAtA []byte) (int, error) {
+	size := m.Size()
+	return m.MarshalToSizedBuffer(dAtA[:size])
+}
+
+func (m *Message_ParamsRequest) MarshalToSizedBuffer(dAtA []byte) (int, error) {
+	i := len(dAtA)
+	if m.ParamsRequest != nil {
+		{
+			size, err := m.ParamsRequest.MarshalToSizedBuffer(dAtA[:i])
+			if err != nil {
+				return 0, err
+			}
+			i -= size
+			i = encodeVarintTypes(dAtA, i, uint64(size))
+		}
+		i--
+		dAtA[i] = 0x3a
+	}
+	return len(dAtA) - i, nil
+}
+func (m *Message_ParamsResponse) MarshalTo(dAtA []byte) (int, error) {
+	size := m.Size()
+	return m.MarshalToSizedBuffer(dAtA[:size])
+}
+
+func (m *Message_ParamsResponse) MarshalToSizedBuffer(dAtA []byte) (int, error) {
+	i := len(dAtA)
+	if m.ParamsResponse != nil {
+		{
+			size, err := m.ParamsResponse.MarshalToSizedBuffer(dAtA[:i])
+			if err != nil {
+				return 0, err
+			}
+			i -= size
+			i = encodeVarintTypes(dAtA, i, uint64(size))
+		}
+		i--
+		dAtA[i] = 0x42
+	}
+	return len(dAtA) - i, nil
+}
 func (m *SnapshotsRequest) Marshal() (dAtA []byte, err error) {
 	size := m.Size()
 	dAtA = make([]byte, size)
@@ -932,6 +1105,72 @@ func (m *LightBlockResponse) MarshalToSizedBuffer(dAtA []byte) (int, error) {
 	return len(dAtA) - i, nil
 }
 
+func (m *ParamsRequest) Marshal() (dAtA []byte, err error) {
+	size := m.Size()
+	dAtA = make([]byte, size)
+	n, err := m.MarshalToSizedBuffer(dAtA[:size])
+	if err != nil {
+		return nil, err
+	}
+	return dAtA[:n], nil
+}
+
+func (m *ParamsRequest) MarshalTo(dAtA []byte) (int, error) {
+	size := m.Size()
+	return m.MarshalToSizedBuffer(dAtA[:size])
+}
+
+func (m *ParamsRequest) MarshalToSizedBuffer(dAtA []byte) (int, error) {
+	i := len(dAtA)
+	_ = i
+	var l int
+	_ = l
+	if m.Height != 0 {
+		i = encodeVarintTypes(dAtA, i, uint64(m.Height))
+		i--
+		dAtA[i] = 0x8
+	}
+	return len(dAtA) - i, nil
+}
+
+func (m *ParamsResponse) Marshal() (dAtA []byte, err error) {
+	size := m.Size()
+	dAtA = make([]byte, size)
+	n, err := m.MarshalToSizedBuffer(dAtA[:size])
+	if err != nil {
+		return nil, err
+	}
+	return dAtA[:n], nil
+}
+
+func (m *ParamsResponse) MarshalTo(dAtA []byte) (int, error) {
+	size := m.Size()
+	return m.MarshalToSizedBuffer(dAtA[:size])
+}
+
+func (m *ParamsResponse) MarshalToSizedBuffer(dAtA []byte) (int, error) {
+	i := len(dAtA)
+	_ = i
+	var l int
+	_ = l
+	{
+		size, err := m.ConsensusParams.MarshalToSizedBuffer(dAtA[:i])
+		if err != nil {
+			return 0, err
+		}
+		i -= size
+		i = encodeVarintTypes(dAtA, i, uint64(size))
+	}
+	i--
+	dAtA[i] = 0x12
+	if m.Height != 0 {
+		i = encodeVarintTypes(dAtA, i, uint64(m.Height))
+		i--
+		dAtA[i] = 0x8
+	}
+	return len(dAtA) - i, nil
+}
+
 func encodeVarintTypes(dAtA []byte, offset int, v uint64) int {
 	offset -= sovTypes(v)
 	base := offset
@@ -1027,6 +1266,30 @@ func (m *Message_LightBlockResponse) Size() (n int) {
 	}
 	return n
 }
+func (m *Message_ParamsRequest) Size() (n int) {
+	if m == nil {
+		return 0
+	}
+	var l int
+	_ = l
+	if m.ParamsRequest != nil {
+		l = m.ParamsRequest.Size()
+		n += 1 + l + sovTypes(uint64(l))
+	}
+	return n
+}
+func (m *Message_ParamsResponse) Size() (n int) {
+	if m == nil {
+		return 0
+	}
+	var l int
+	_ = l
+	if m.ParamsResponse != nil {
+		l = m.ParamsResponse.Size()
+		n += 1 + l + sovTypes(uint64(l))
+	}
+	return n
+}
 func (m *SnapshotsRequest) Size() (n int) {
 	if m == nil {
 		return 0
@@ -1130,6 +1393,32 @@ func (m *LightBlockResponse) Size() (n int) {
 	return n
 }
 
+func (m *ParamsRequest) Size() (n int) {
+	if m == nil {
+		return 0
+	}
+	var l int
+	_ = l
+	if m.Height != 0 {
+		n += 1 + sovTypes(uint64(m.Height))
+	}
+	return n
+}
+
+func (m *ParamsResponse) Size() (n int) {
+	if m == nil {
+		return 0
+	}
+	var l int
+	_ = l
+	if m.Height != 0 {
+		n += 1 + sovTypes(uint64(m.Height))
+	}
+	l = m.ConsensusParams.Size()
+	n += 1 + l + sovTypes(uint64(l))
+	return n
+}
+
 func sovTypes(x uint64) (n int) {
 	return (math_bits.Len64(x|1) + 6) / 7
 }
@@ -1375,6 +1664,76 @@ func (m *Message) Unmarshal(dAtA []byte) error {
 			}
 			m.Sum = &Message_LightBlockResponse{v}
 			iNdEx = postIndex
+		case 7:
+			if wireType != 2 {
+				return fmt.Errorf("proto: wrong wireType = %d for field ParamsRequest", wireType)
+			}
+			var msglen int
+			for shift := uint(0); ; shift += 7 {
+				if shift >= 64 {
+					return ErrIntOverflowTypes
+				}
+				if iNdEx >= l {
+					return io.ErrUnexpectedEOF
+				}
+				b := dAtA[iNdEx]
+				iNdEx++
+				msglen |= int(b&0x7F) << shift
+				if b < 0x80 {
+					break
+				}
+			}
+			if msglen < 0 {
+				return ErrInvalidLengthTypes
+			}
+			postIndex := iNdEx + msglen
+			if postIndex < 0 {
+				return ErrInvalidLengthTypes
+			}
+			if postIndex > l {
+				return io.ErrUnexpectedEOF
+			}
+			v := &ParamsRequest{}
+			if err := v.Unmarshal(dAtA[iNdEx:postIndex]); err != nil {
+				return err
+			}
+			m.Sum = &Message_ParamsRequest{v}
+			iNdEx = postIndex
+		case 8:
+			if wireType != 2 {
+				return fmt.Errorf("proto: wrong wireType = %d for field ParamsResponse", wireType)
+			}
+			var msglen int
+			for shift := uint(0); ; shift += 7 {
+				if shift >= 64 {
+					return ErrIntOverflowTypes
+				}
+				if iNdEx >= l {
+					return io.ErrUnexpectedEOF
+				}
+				b := dAtA[iNdEx]
+				iNdEx++
+				msglen |= int(b&0x7F) << shift
+				if b < 0x80 {
+					break
+				}
+			}
+			if msglen < 0 {
+				return ErrInvalidLengthTypes
+			}
+			postIndex := iNdEx + msglen
+			if postIndex < 0 {
+				return ErrInvalidLengthTypes
+			}
+			if postIndex > l {
+				return io.ErrUnexpectedEOF
+			}
+			v := &ParamsResponse{}
+			if err := v.Unmarshal(dAtA[iNdEx:postIndex]); err != nil {
+				return err
+			}
+			m.Sum = &Message_ParamsResponse{v}
+			iNdEx = postIndex
 		default:
 			iNdEx = preIndex
 			skippy, err := skipTypes(dAtA[iNdEx:])
@@ -2044,6 +2403,177 @@ func (m *LightBlockResponse) Unmarshal(dAtA []byte) error {
 	}
 	return nil
 }
+func (m *ParamsRequest) Unmarshal(dAtA []byte) error {
+	l := len(dAtA)
+	iNdEx := 0
+	for iNdEx < l {
+		preIndex := iNdEx
+		var wire uint64
+		for shift := uint(0); ; shift += 7 {
+			if shift >= 64 {
+				return ErrIntOverflowTypes
+			}
+			if iNdEx >= l {
+				return io.ErrUnexpectedEOF
+			}
+			b := dAtA[iNdEx]
+			iNdEx++
+			wire |= uint64(b&0x7F) << shift
+			if b < 0x80 {
+				break
+			}
+		}
+		fieldNum := int32(wire >> 3)
+		wireType := int(wire & 0x7)
+		if wireType == 4 {
+			return fmt.Errorf("proto: ParamsRequest: wiretype end group for non-group")
+		}
+		if fieldNum <= 0 {
+			return fmt.Errorf("proto: ParamsRequest: illegal tag %d (wire type %d)", fieldNum, wire)
+		}
+		switch fieldNum {
+		case 1:
+			if wireType != 0 {
+				return fmt.Errorf("proto: wrong wireType = %d for field Height", wireType)
+			}
+			m.Height = 0
+			for shift := uint(0); ; shift += 7 {
+				if shift >= 64 {
+					return ErrIntOverflowTypes
+				}
+				if iNdEx >= l {
+					return io.ErrUnexpectedEOF
+				}
+				b := dAtA[iNdEx]
+				iNdEx++
+				m.Height |= uint64(b&0x7F) << shift
+				if b < 0x80 {
+					break
+				}
+			}
+		default:
+			iNdEx = preIndex
+			skippy, err := skipTypes(dAtA[iNdEx:])
+			if err != nil {
+				return err
+			}
+			if (skippy < 0) || (iNdEx+skippy) < 0 {
+				return ErrInvalidLengthTypes
+			}
+			if (iNdEx + skippy) > l {
+				return io.ErrUnexpectedEOF
+			}
+			iNdEx += skippy
+		}
+	}
+
+	if iNdEx > l {
+		return io.ErrUnexpectedEOF
+	}
+	return nil
+}
+func (m *ParamsResponse) Unmarshal(dAtA []byte) error {
+	l := len(dAtA)
+	iNdEx := 0
+	for iNdEx < l {
+		preIndex := iNdEx
+		var wire uint64
+		for shift := uint(0); ; shift += 7 {
+			if shift >= 64 {
+				return ErrIntOverflowTypes
+			}
+			if iNdEx >= l {
+				return io.ErrUnexpectedEOF
+			}
+			b := dAtA[iNdEx]
+			iNdEx++
+			wire |= uint64(b&0x7F) << shift
+			if b < 0x80 {
+				break
+			}
+		}
+		fieldNum := int32(wire >> 3)
+		wireType := int(wire & 0x7)
+		if wireType == 4 {
+			return fmt.Errorf("proto: ParamsResponse: wiretype end group for non-group")
+		}
+		if fieldNum <= 0 {
+			return fmt.Errorf("proto: ParamsResponse: illegal tag %d (wire type %d)", fieldNum, wire)
+		}
+		switch fieldNum {
+		case 1:
+			if wireType != 0 {
+				return fmt.Errorf("proto: wrong wireType = %d for field Height", wireType)
+			}
+			m.Height = 0
+			for shift := uint(0); ; shift += 7 {
+				if shift >= 64 {
+					return ErrIntOverflowTypes
+				}
+				if iNdEx >= l {
+					return io.ErrUnexpectedEOF
+				}
+				b := dAtA[iNdEx]
+				iNdEx++
+				m.Height |= uint64(b&0x7F) << shift
+				if b < 0x80 {
+					break
+				}
+			}
+		case 2:
+			if wireType != 2 {
+				return fmt.Errorf("proto: wrong wireType = %d for field ConsensusParams", wireType)
+			}
+			var msglen int
+			for shift := uint(0); ; shift += 7 {
+				if shift >= 64 {
+					return ErrIntOverflowTypes
+				}
+				if iNdEx >= l {
+					return io.ErrUnexpectedEOF
+				}
+				b := dAtA[iNdEx]
+				iNdEx++
+				msglen |= int(b&0x7F) << shift
+				if b < 0x80 {
+					break
+				}
+			}
+			if msglen < 0 {
+				return ErrInvalidLengthTypes
+			}
+			postIndex := iNdEx + msglen
+			if postIndex < 0 {
+				return ErrInvalidLengthTypes
+			}
+			if postIndex > l {
+				return io.ErrUnexpectedEOF
+			}
+			if err := m.ConsensusParams.Unmarshal(dAtA[iNdEx:postIndex]); err != nil {
+				return err
+			}
+			iNdEx = postIndex
+		default:
+			iNdEx = preIndex
+			skippy, err := skipTypes(dAtA[iNdEx:])
+			if err != nil {
+				return err
+			}
+			if (skippy < 0) || (iNdEx+skippy) < 0 {
+				return ErrInvalidLengthTypes
+			}
+			if (iNdEx + skippy) > l {
+				return io.ErrUnexpectedEOF
+			}
+			iNdEx += skippy
+		}
+	}
+
+	if iNdEx > l {
+		return io.ErrUnexpectedEOF
+	}
+	return nil
+}
 func skipTypes(dAtA []byte) (n int, err error) {
 	l := len(dAtA)
 	iNdEx := 0

proto/tendermint/statesync/types.proto

@@ -1,7 +1,9 @@
 syntax = "proto3";
 package tendermint.statesync;
 
+import "gogoproto/gogo.proto";
 import "tendermint/types/types.proto";
+import "tendermint/types/params.proto";
 
 option go_package = "github.com/tendermint/tendermint/proto/tendermint/statesync";
 
@@ -13,6 +15,8 @@ message Message {
     ChunkResponse      chunk_response       = 4;
     LightBlockRequest  light_block_request  = 5;
     LightBlockResponse light_block_response = 6;
+    ParamsRequest      params_request       = 7;
+    ParamsResponse     params_response      = 8;
   }
 }
 
@@ -46,4 +50,13 @@ message LightBlockRequest {
 
 message LightBlockResponse {
   tendermint.types.LightBlock light_block = 1;
+}
+
+message ParamsRequest {
+  uint64 height = 1;
+}
+
+message ParamsResponse {
+  uint64                           height           = 1;
+  tendermint.types.ConsensusParams consensus_params = 2 [(gogoproto.nullable) = false];
 }

release_notes.md

@@ -1 +0,0 @@
-<!-- TODO: add in release notes prior to release. -->

rpc/client/evidence_test.go

@@ -29,7 +29,7 @@ var defaultTestTime = time.Date(2018, 10, 10, 8, 20, 13, 695936996, time.UTC)
 func newEvidence(t *testing.T, val *privval.FilePV,
 	vote *types.Vote, vote2 *types.Vote,
 	chainID string) *types.DuplicateVoteEvidence {
-
+	t.Helper()
 	var err error
 
 	v := vote.ToProto()
@@ -44,7 +44,9 @@ func newEvidence(t *testing.T, val *privval.FilePV,
 	validator := types.NewValidator(val.Key.PubKey, 10)
 	valSet := types.NewValidatorSet([]*types.Validator{validator})
 
-	return types.NewDuplicateVoteEvidence(vote, vote2, defaultTestTime, valSet)
+	ev, err := types.NewDuplicateVoteEvidence(vote, vote2, defaultTestTime, valSet)
+	require.NoError(t, err)
+	return ev
 }
 
 func makeEvidences(

rpc/core/env.go

@@ -58,6 +58,11 @@ type peers interface {
 	Peers() p2p.IPeerSet
 }
 
+type consensusReactor interface {
+	WaitSync() bool
+	GetPeerState(peerID types.NodeID) (*consensus.PeerState, bool)
+}
+
 type peerManager interface {
 	Peers() []types.NodeID
 	Addresses(types.NodeID) []p2p.NodeAddress
@@ -72,11 +77,12 @@ type Environment struct {
 	ProxyAppMempool proxy.AppConnMempool
 
 	// interfaces defined in types and above
-	StateStore     sm.Store
-	BlockStore     sm.BlockStore
-	EvidencePool   sm.EvidencePool
-	ConsensusState consensusState
-	P2PPeers       peers
+	StateStore       sm.Store
+	BlockStore       sm.BlockStore
+	EvidencePool     sm.EvidencePool
+	ConsensusState   consensusState
+	ConsensusReactor consensusReactor
+	P2PPeers         peers
 
 	// Legacy p2p stack
 	P2PTransport transport
@@ -88,7 +94,6 @@ type Environment struct {
 	PubKey           crypto.PubKey
 	GenDoc           *types.GenesisDoc // cache the genesis structure
 	EventSinks       []indexer.EventSink
-	ConsensusReactor *consensus.Reactor
 	EventBus         *types.EventBus // thread safe
 	Mempool          mempl.Mempool
 	BlockSyncReactor consensus.BlockSyncReactor

rpc/jsonrpc/jsonrpc_test.go

@@ -110,7 +110,7 @@ func setup() {
 	wm.SetLogger(tcpLogger)
 	mux.HandleFunc(websocketEndpoint, wm.WebsocketHandler)
 	config := server.DefaultConfig()
-	listener1, err := server.Listen(tcpAddr, config)
+	listener1, err := server.Listen(tcpAddr, config.MaxOpenConnections)
 	if err != nil {
 		panic(err)
 	}
@@ -126,7 +126,7 @@ func setup() {
 	wm = server.NewWebsocketManager(Routes)
 	wm.SetLogger(unixLogger)
 	mux2.HandleFunc(websocketEndpoint, wm.WebsocketHandler)
-	listener2, err := server.Listen(unixAddr, config)
+	listener2, err := server.Listen(unixAddr, config.MaxOpenConnections)
 	if err != nil {
 		panic(err)
 	}

rpc/jsonrpc/server/http_server.go

@@ -261,7 +261,7 @@ func (h maxBytesHandler) ServeHTTP(w http.ResponseWriter, r *http.Request) {
 
 // Listen starts a new net.Listener on the given address.
 // It returns an error if the address is invalid or the call to Listen() fails.
-func Listen(addr string, config *Config) (listener net.Listener, err error) {
+func Listen(addr string, maxOpenConnections int) (listener net.Listener, err error) {
 	parts := strings.SplitN(addr, "://", 2)
 	if len(parts) != 2 {
 		return nil, fmt.Errorf(
@@ -274,8 +274,8 @@ func Listen(addr string, config *Config) (listener net.Listener, err error) {
 	if err != nil {
 		return nil, fmt.Errorf("failed to listen on %v: %v", addr, err)
 	}
-	if config.MaxOpenConnections > 0 {
-		listener = netutil.LimitListener(listener, config.MaxOpenConnections)
+	if maxOpenConnections > 0 {
+		listener = netutil.LimitListener(listener, maxOpenConnections)
 	}
 
 	return listener, nil

rpc/jsonrpc/server/http_server_test.go

@@ -39,8 +39,7 @@ func TestMaxOpenConnections(t *testing.T) {
 		fmt.Fprint(w, "some body")
 	})
 	config := DefaultConfig()
-	config.MaxOpenConnections = max
-	l, err := Listen("tcp://127.0.0.1:0", config)
+	l, err := Listen("tcp://127.0.0.1:0", max)
 	require.NoError(t, err)
 	defer l.Close()
 	go Serve(l, mux, log.TestingLogger(), config) //nolint:errcheck // ignore for tests

rpc/jsonrpc/test/main.go

@@ -33,7 +33,7 @@ func main() {
 
 	rpcserver.RegisterRPCFuncs(mux, routes, logger)
 	config := rpcserver.DefaultConfig()
-	listener, err := rpcserver.Listen("tcp://127.0.0.1:8008", config)
+	listener, err := rpcserver.Listen("tcp://127.0.0.1:8008", config.MaxOpenConnections)
 	if err != nil {
 		tmos.Exit(err.Error())
 	}

state/indexer/indexer_service.go

@@ -51,43 +51,47 @@ func (is *Service) OnStart() error {
 
 	go func() {
 		for {
-			msg := <-blockHeadersSub.Out()
+			select {
+			case <-blockHeadersSub.Canceled():
+				return
+			case msg := <-blockHeadersSub.Out():
 
-			eventDataHeader := msg.Data().(types.EventDataNewBlockHeader)
-			height := eventDataHeader.Header.Height
-			batch := NewBatch(eventDataHeader.NumTxs)
+				eventDataHeader := msg.Data().(types.EventDataNewBlockHeader)
+				height := eventDataHeader.Header.Height
+				batch := NewBatch(eventDataHeader.NumTxs)
 
-			for i := int64(0); i < eventDataHeader.NumTxs; i++ {
-				msg2 := <-txsSub.Out()
-				txResult := msg2.Data().(types.EventDataTx).TxResult
+				for i := int64(0); i < eventDataHeader.NumTxs; i++ {
+					msg2 := <-txsSub.Out()
+					txResult := msg2.Data().(types.EventDataTx).TxResult
 
-				if err = batch.Add(&txResult); err != nil {
-					is.Logger.Error(
-						"failed to add tx to batch",
-						"height", height,
-						"index", txResult.Index,
-						"err", err,
-					)
-				}
-			}
-
-			if !IndexingEnabled(is.eventSinks) {
-				continue
-			}
-
-			for _, sink := range is.eventSinks {
-				if err := sink.IndexBlockEvents(eventDataHeader); err != nil {
-					is.Logger.Error("failed to index block", "height", height, "err", err)
-				} else {
-					is.Logger.Debug("indexed block", "height", height, "sink", sink.Type())
+					if err = batch.Add(&txResult); err != nil {
+						is.Logger.Error(
+							"failed to add tx to batch",
+							"height", height,
+							"index", txResult.Index,
+							"err", err,
+						)
+					}
 				}
 
-				if len(batch.Ops) > 0 {
-					err := sink.IndexTxEvents(batch.Ops)
-					if err != nil {
-						is.Logger.Error("failed to index block txs", "height", height, "err", err)
+				if !IndexingEnabled(is.eventSinks) {
+					continue
+				}
+
+				for _, sink := range is.eventSinks {
+					if err := sink.IndexBlockEvents(eventDataHeader); err != nil {
+						is.Logger.Error("failed to index block", "height", height, "err", err)
 					} else {
-						is.Logger.Debug("indexed txs", "height", height, "sink", sink.Type())
+						is.Logger.Debug("indexed block", "height", height, "sink", sink.Type())
+					}
+
+					if len(batch.Ops) > 0 {
+						err := sink.IndexTxEvents(batch.Ops)
+						if err != nil {
+							is.Logger.Error("failed to index block txs", "height", height, "err", err)
+						} else {
+							is.Logger.Debug("indexed txs", "height", height, "sink", sink.Type())
+						}
 					}
 				}
 			}

state/indexer/indexer_service_test.go

@@ -139,7 +139,7 @@ func setupDB(t *testing.T) (*dockertest.Pool, error) {
 	assert.Nil(t, err)
 
 	resource, err = pool.RunWithOptions(&dockertest.RunOptions{
-		Repository: psql.DriverName,
+		Repository: "postgres",
 		Tag:        "13",
 		Env: []string{
 			"POSTGRES_USER=" + user,
@@ -164,19 +164,16 @@ func setupDB(t *testing.T) (*dockertest.Pool, error) {
 
 	conn := fmt.Sprintf(dsn, user, password, resource.GetPort(port+"/tcp"), dbName)
 
-	if err = pool.Retry(func() error {
-		var err error
-
-		pSink, psqldb, err = psql.NewEventSink(conn, "test-chainID")
-
+	assert.NoError(t, pool.Retry(func() error {
+		sink, err := psql.NewEventSink(conn, "test-chainID")
 		if err != nil {
 			return err
 		}
 
+		pSink = sink
+		psqldb = sink.DB()
 		return psqldb.Ping()
-	}); err != nil {
-		assert.Error(t, err)
-	}
+	}))
 
 	resetDB(t)
 

state/indexer/sink/psql/psql.go

@@ -1,3 +1,4 @@
+// Package psql implements an event sink backed by a PostgreSQL database.
 package psql
 
 import (
@@ -5,207 +6,252 @@ import (
 	"database/sql"
 	"errors"
 	"fmt"
+	"strings"
 	"time"
 
-	sq "github.com/Masterminds/squirrel"
-	proto "github.com/gogo/protobuf/proto"
+	"github.com/gogo/protobuf/proto"
 	abci "github.com/tendermint/tendermint/abci/types"
 	"github.com/tendermint/tendermint/libs/pubsub/query"
 	"github.com/tendermint/tendermint/state/indexer"
 	"github.com/tendermint/tendermint/types"
 )
 
-var _ indexer.EventSink = (*EventSink)(nil)
-
 const (
-	TableEventBlock = "block_events"
-	TableEventTx    = "tx_events"
-	TableResultTx   = "tx_results"
-	DriverName      = "postgres"
+	tableBlocks     = "blocks"
+	tableTxResults  = "tx_results"
+	tableEvents     = "events"
+	tableAttributes = "attributes"
+	driverName      = "postgres"
 )
 
-// EventSink is an indexer backend providing the tx/block index services.
+// EventSink is an indexer backend providing the tx/block index services.  This
+// implementation stores records in a PostgreSQL database using the schema
+// defined in state/indexer/sink/psql/schema.sql.
 type EventSink struct {
 	store   *sql.DB
 	chainID string
 }
 
-func NewEventSink(connStr string, chainID string) (indexer.EventSink, *sql.DB, error) {
-	db, err := sql.Open(DriverName, connStr)
+// NewEventSink constructs an event sink associated with the PostgreSQL
+// database specified by connStr. Events written to the sink are attributed to
+// the specified chainID.
+func NewEventSink(connStr, chainID string) (*EventSink, error) {
+	db, err := sql.Open(driverName, connStr)
 	if err != nil {
-		return nil, nil, err
+		return nil, err
 	}
 
 	return &EventSink{
 		store:   db,
 		chainID: chainID,
-	}, db, nil
+	}, nil
 }
 
-func (es *EventSink) Type() indexer.EventSinkType {
-	return indexer.PSQL
-}
+// DB returns the underlying Postgres connection used by the sink.
+// This is exported to support testing.
+func (es *EventSink) DB() *sql.DB { return es.store }
 
-func (es *EventSink) IndexBlockEvents(h types.EventDataNewBlockHeader) error {
-	sqlStmt := sq.
-		Insert(TableEventBlock).
-		Columns("key", "value", "height", "type", "created_at", "chain_id").
-		PlaceholderFormat(sq.Dollar).
-		Suffix("ON CONFLICT (key,height)").
-		Suffix("DO NOTHING")
+// Type returns the structure type for this sink, which is Postgres.
+func (es *EventSink) Type() indexer.EventSinkType { return indexer.PSQL }
 
-	ts := time.Now()
-	// index the reserved block height index
-	sqlStmt = sqlStmt.
-		Values(types.BlockHeightKey, fmt.Sprint(h.Header.Height), h.Header.Height, "", ts, es.chainID)
-
-	// index begin_block events
-	sqlStmt, err := indexBlockEvents(
-		sqlStmt, h.ResultBeginBlock.Events, types.EventTypeBeginBlock, h.Header.Height, ts, es.chainID)
+// runInTransaction executes query in a fresh database transaction.
+// If query reports an error, the transaction is rolled back and the
+// error from query is reported to the caller.
+// Otherwise, the result of committing the transaction is returned.
+func runInTransaction(db *sql.DB, query func(*sql.Tx) error) error {
+	dbtx, err := db.Begin()
 	if err != nil {
 		return err
 	}
-
-	// index end_block events
-	sqlStmt, err = indexBlockEvents(
-		sqlStmt, h.ResultEndBlock.Events, types.EventTypeEndBlock, h.Header.Height, ts, es.chainID)
-	if err != nil {
+	if err := query(dbtx); err != nil {
+		_ = dbtx.Rollback() // report the initial error, not the rollback
 		return err
 	}
-
-	_, err = sqlStmt.RunWith(es.store).Exec()
-	return err
+	return dbtx.Commit()
 }
 
-func (es *EventSink) IndexTxEvents(txr []*abci.TxResult) error {
-	// index the tx result
-	var txid uint32
-	sqlStmtTxResult := sq.
-		Insert(TableResultTx).
-		Columns("tx_result", "created_at").
-		PlaceholderFormat(sq.Dollar).
-		RunWith(es.store).
-		Suffix("ON CONFLICT (tx_result)").
-		Suffix("DO NOTHING").
-		Suffix("RETURNING \"id\"")
+// queryWithID executes the specified SQL query with the given arguments,
+// expecting a single-row, single-column result containing an ID. If the query
+// succeeds, the ID from the result is returned.
+func queryWithID(tx *sql.Tx, query string, args ...interface{}) (uint32, error) {
+	var id uint32
+	if err := tx.QueryRow(query, args...).Scan(&id); err != nil {
+		return 0, err
+	}
+	return id, nil
+}
 
-	sqlStmtEvents := sq.
-		Insert(TableEventTx).
-		Columns("key", "value", "height", "hash", "tx_result_id", "created_at", "chain_id").
-		PlaceholderFormat(sq.Dollar).
-		Suffix("ON CONFLICT (key,hash)").
-		Suffix("DO NOTHING")
+// insertEvents inserts a slice of events and any indexed attributes of those
+// events into the database associated with dbtx.
+//
+// If txID > 0, the event is attributed to the Tendermint transaction with that
+// ID; otherwise it is recorded as a block event.
+func insertEvents(dbtx *sql.Tx, blockID, txID uint32, evts []abci.Event) error {
+	// Populate the transaction ID field iff one is defined (> 0).
+	var txIDArg interface{}
+	if txID > 0 {
+		txIDArg = txID
+	}
 
-	ts := time.Now()
-	for _, tx := range txr {
-		txBz, err := proto.Marshal(tx)
+	// Add each event to the events table, and retrieve its row ID to use when
+	// adding any attributes the event provides.
+	for _, evt := range evts {
+		// Skip events with an empty type.
+		if evt.Type == "" {
+			continue
+		}
+
+		eid, err := queryWithID(dbtx, `
+INSERT INTO `+tableEvents+` (block_id, tx_id, type) VALUES ($1, $2, $3)
+  RETURNING rowid;
+`, blockID, txIDArg, evt.Type)
 		if err != nil {
 			return err
 		}
 
-		sqlStmtTxResult = sqlStmtTxResult.Values(txBz, ts)
-
-		// execute sqlStmtTxResult db query and retrieve the txid
-		r, err := sqlStmtTxResult.Query()
-		if err != nil {
-			return err
-		}
-		defer r.Close()
-
-		if !r.Next() {
-			return nil
-		}
-
-		if err := r.Scan(&txid); err != nil {
-			return err
-		}
-
-		// index the reserved height and hash indices
-		hash := fmt.Sprintf("%X", types.Tx(tx.Tx).Hash())
-
-		sqlStmtEvents = sqlStmtEvents.Values(types.TxHashKey, hash, tx.Height, hash, txid, ts, es.chainID)
-		sqlStmtEvents = sqlStmtEvents.Values(types.TxHeightKey, fmt.Sprint(tx.Height), tx.Height, hash, txid, ts, es.chainID)
-		for _, event := range tx.Result.Events {
-			// only index events with a non-empty type
-			if len(event.Type) == 0 {
+		// Add any attributes flagged for indexing.
+		for _, attr := range evt.Attributes {
+			if !attr.Index {
 				continue
 			}
-
-			for _, attr := range event.Attributes {
-				if len(attr.Key) == 0 {
-					continue
-				}
-
-				// index if `index: true` is set
-				compositeTag := fmt.Sprintf("%s.%s", event.Type, attr.Key)
-
-				// ensure event does not conflict with a reserved prefix key
-				if compositeTag == types.TxHashKey || compositeTag == types.TxHeightKey {
-					return fmt.Errorf("event type and attribute key \"%s\" is reserved; please use a different key", compositeTag)
-				}
-
-				if attr.GetIndex() {
-					sqlStmtEvents = sqlStmtEvents.Values(compositeTag, attr.Value, tx.Height, hash, txid, ts, es.chainID)
-				}
+			compositeKey := evt.Type + "." + attr.Key
+			if _, err := dbtx.Exec(`
+INSERT INTO `+tableAttributes+` (event_id, key, composite_key, value)
+  VALUES ($1, $2, $3, $4);
+`, eid, attr.Key, compositeKey, attr.Value); err != nil {
+				return err
 			}
 		}
 	}
-
-	// execute sqlStmtEvents db query...
-	_, err := sqlStmtEvents.RunWith(es.store).Exec()
-	return err
+	return nil
 }
 
+// makeIndexedEvent constructs an event from the specified composite key and
+// value. If the key has the form "type.name", the event will have a single
+// attribute with that name and the value; otherwise the event will have only
+// a type and no attributes.
+func makeIndexedEvent(compositeKey, value string) abci.Event {
+	i := strings.Index(compositeKey, ".")
+	if i < 0 {
+		return abci.Event{Type: compositeKey}
+	}
+	return abci.Event{Type: compositeKey[:i], Attributes: []abci.EventAttribute{
+		{Key: compositeKey[i+1:], Value: value, Index: true},
+	}}
+}
+
+// IndexBlockEvents indexes the specified block header, part of the
+// indexer.EventSink interface.
+func (es *EventSink) IndexBlockEvents(h types.EventDataNewBlockHeader) error {
+	ts := time.Now().UTC()
+
+	return runInTransaction(es.store, func(dbtx *sql.Tx) error {
+		// Add the block to the blocks table and report back its row ID for use
+		// in indexing the events for the block.
+		blockID, err := queryWithID(dbtx, `
+INSERT INTO `+tableBlocks+` (height, chain_id, created_at)
+  VALUES ($1, $2, $3)
+  ON CONFLICT DO NOTHING
+  RETURNING rowid;
+`, h.Header.Height, es.chainID, ts)
+		if err == sql.ErrNoRows {
+			return nil // we already saw this block; quietly succeed
+		} else if err != nil {
+			return fmt.Errorf("indexing block header: %w", err)
+		}
+
+		// Insert the special block meta-event for height.
+		if err := insertEvents(dbtx, blockID, 0, []abci.Event{
+			makeIndexedEvent(types.BlockHeightKey, fmt.Sprint(h.Header.Height)),
+		}); err != nil {
+			return fmt.Errorf("block meta-events: %w", err)
+		}
+		// Insert all the block events. Order is important here,
+		if err := insertEvents(dbtx, blockID, 0, h.ResultBeginBlock.Events); err != nil {
+			return fmt.Errorf("begin-block events: %w", err)
+		}
+		if err := insertEvents(dbtx, blockID, 0, h.ResultEndBlock.Events); err != nil {
+			return fmt.Errorf("end-block events: %w", err)
+		}
+		return nil
+	})
+}
+
+func (es *EventSink) IndexTxEvents(txrs []*abci.TxResult) error {
+	ts := time.Now().UTC()
+
+	for _, txr := range txrs {
+		// Encode the result message in protobuf wire format for indexing.
+		resultData, err := proto.Marshal(txr)
+		if err != nil {
+			return fmt.Errorf("marshaling tx_result: %w", err)
+		}
+
+		// Index the hash of the underlying transaction as a hex string.
+		txHash := fmt.Sprintf("%X", types.Tx(txr.Tx).Hash())
+
+		if err := runInTransaction(es.store, func(dbtx *sql.Tx) error {
+			// Find the block associated with this transaction. The block header
+			// must have been indexed prior to the transactions belonging to it.
+			blockID, err := queryWithID(dbtx, `
+SELECT rowid FROM `+tableBlocks+` WHERE height = $1 AND chain_id = $2;
+`, txr.Height, es.chainID)
+			if err != nil {
+				return fmt.Errorf("finding block ID: %w", err)
+			}
+
+			// Insert a record for this tx_result and capture its ID for indexing events.
+			txID, err := queryWithID(dbtx, `
+INSERT INTO `+tableTxResults+` (block_id, index, created_at, tx_hash, tx_result)
+  VALUES ($1, $2, $3, $4, $5)
+  ON CONFLICT DO NOTHING
+  RETURNING rowid;
+`, blockID, txr.Index, ts, txHash, resultData)
+			if err == sql.ErrNoRows {
+				return nil // we already saw this transaction; quietly succeed
+			} else if err != nil {
+				return fmt.Errorf("indexing tx_result: %w", err)
+			}
+
+			// Insert the special transaction meta-events for hash and height.
+			if err := insertEvents(dbtx, blockID, txID, []abci.Event{
+				makeIndexedEvent(types.TxHashKey, txHash),
+				makeIndexedEvent(types.TxHeightKey, fmt.Sprint(txr.Height)),
+			}); err != nil {
+				return fmt.Errorf("indexing transaction meta-events: %w", err)
+			}
+			// Index any events packaged with the transaction.
+			if err := insertEvents(dbtx, blockID, txID, txr.Result.Events); err != nil {
+				return fmt.Errorf("indexing transaction events: %w", err)
+			}
+			return nil
+
+		}); err != nil {
+			return err
+		}
+	}
+	return nil
+}
+
+// SearchBlockEvents is not implemented by this sink, and reports an error for all queries.
 func (es *EventSink) SearchBlockEvents(ctx context.Context, q *query.Query) ([]int64, error) {
 	return nil, errors.New("block search is not supported via the postgres event sink")
 }
 
+// SearchTxEvents is not implemented by this sink, and reports an error for all queries.
 func (es *EventSink) SearchTxEvents(ctx context.Context, q *query.Query) ([]*abci.TxResult, error) {
 	return nil, errors.New("tx search is not supported via the postgres event sink")
 }
 
+// GetTxByHash is not implemented by this sink, and reports an error for all queries.
 func (es *EventSink) GetTxByHash(hash []byte) (*abci.TxResult, error) {
 	return nil, errors.New("getTxByHash is not supported via the postgres event sink")
 }
 
+// HasBlock is not implemented by this sink, and reports an error for all queries.
 func (es *EventSink) HasBlock(h int64) (bool, error) {
 	return false, errors.New("hasBlock is not supported via the postgres event sink")
 }
 
-func indexBlockEvents(
-	sqlStmt sq.InsertBuilder,
-	events []abci.Event,
-	ty string,
-	height int64,
-	ts time.Time,
-	chainID string,
-) (sq.InsertBuilder, error) {
-	for _, event := range events {
-		// only index events with a non-empty type
-		if len(event.Type) == 0 {
-			continue
-		}
-
-		for _, attr := range event.Attributes {
-			if len(attr.Key) == 0 {
-				continue
-			}
-
-			// index iff the event specified index:true and it's not a reserved event
-			compositeKey := fmt.Sprintf("%s.%s", event.Type, attr.Key)
-			if compositeKey == types.BlockHeightKey {
-				return sqlStmt, fmt.Errorf(
-					"event type and attribute key \"%s\" is reserved; please use a different key", compositeKey)
-			}
-
-			if attr.GetIndex() {
-				sqlStmt = sqlStmt.Values(compositeKey, attr.Value, height, ty, ts, chainID)
-			}
-		}
-	}
-	return sqlStmt, nil
-}
-
-func (es *EventSink) Stop() error {
-	return es.store.Close()
-}
+// Stop closes the underlying PostgreSQL database.
+func (es *EventSink) Stop() error { return es.store.Close() }

state/indexer/sink/psql/psql_test.go

@@ -3,320 +3,64 @@ package psql
 import (
 	"context"
 	"database/sql"
-	"errors"
+	"flag"
 	"fmt"
 	"io/ioutil"
+	"log"
 	"os"
+	"os/signal"
 	"testing"
 	"time"
 
-	sq "github.com/Masterminds/squirrel"
-	schema "github.com/adlio/schema"
-	proto "github.com/gogo/protobuf/proto"
-	_ "github.com/lib/pq"
-	dockertest "github.com/ory/dockertest"
+	"github.com/adlio/schema"
+	"github.com/gogo/protobuf/proto"
+	"github.com/ory/dockertest"
 	"github.com/ory/dockertest/docker"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 	abci "github.com/tendermint/tendermint/abci/types"
 	"github.com/tendermint/tendermint/state/indexer"
 	"github.com/tendermint/tendermint/types"
+
+	// Register the Postgres database driver.
+	_ "github.com/lib/pq"
 )
 
-var db *sql.DB
-var resource *dockertest.Resource
-var chainID = "test-chainID"
+// Verify that the type satisfies the EventSink interface.
+var _ indexer.EventSink = (*EventSink)(nil)
 
 var (
+	doPauseAtExit = flag.Bool("pause-at-exit", false,
+		"If true, pause the test until interrupted at shutdown, to allow debugging")
+
+	// A hook that test cases can call to obtain the shared database instance
+	// used for testing the sink. This is initialized in TestMain (see below).
+	testDB func() *sql.DB
+)
+
+const (
 	user     = "postgres"
 	password = "secret"
 	port     = "5432"
 	dsn      = "postgres://%s:%s@localhost:%s/%s?sslmode=disable"
 	dbName   = "postgres"
+	chainID  = "test-chainID"
+
+	viewBlockEvents = "block_events"
+	viewTxEvents    = "tx_events"
 )
 
-func TestType(t *testing.T) {
-	pool, err := setupDB(t)
-	require.NoError(t, err)
+func TestMain(m *testing.M) {
+	flag.Parse()
 
-	psqlSink := &EventSink{store: db, chainID: chainID}
-	assert.Equal(t, indexer.PSQL, psqlSink.Type())
-	require.NoError(t, teardown(t, pool))
-}
-
-func TestBlockFuncs(t *testing.T) {
-	pool, err := setupDB(t)
-	require.NoError(t, err)
-
-	indexer := &EventSink{store: db, chainID: chainID}
-	require.NoError(t, indexer.IndexBlockEvents(getTestBlockHeader()))
-
-	r, err := verifyBlock(1)
-	assert.True(t, r)
-	require.NoError(t, err)
-
-	r, err = verifyBlock(2)
-	assert.False(t, r)
-	require.NoError(t, err)
-
-	r, err = indexer.HasBlock(1)
-	assert.False(t, r)
-	assert.Equal(t, errors.New("hasBlock is not supported via the postgres event sink"), err)
-
-	r, err = indexer.HasBlock(2)
-	assert.False(t, r)
-	assert.Equal(t, errors.New("hasBlock is not supported via the postgres event sink"), err)
-
-	r2, err := indexer.SearchBlockEvents(context.TODO(), nil)
-	assert.Nil(t, r2)
-	assert.Equal(t, errors.New("block search is not supported via the postgres event sink"), err)
-
-	require.NoError(t, verifyTimeStamp(TableEventBlock))
-
-	// try to insert the duplicate block events.
-	err = indexer.IndexBlockEvents(getTestBlockHeader())
-	require.NoError(t, err)
-
-	require.NoError(t, teardown(t, pool))
-}
-
-func TestTxFuncs(t *testing.T) {
-	pool, err := setupDB(t)
-	assert.Nil(t, err)
-
-	indexer := &EventSink{store: db, chainID: chainID}
-
-	txResult := txResultWithEvents([]abci.Event{
-		{Type: "account", Attributes: []abci.EventAttribute{{Key: "number", Value: "1", Index: true}}},
-		{Type: "account", Attributes: []abci.EventAttribute{{Key: "owner", Value: "Ivan", Index: true}}},
-		{Type: "", Attributes: []abci.EventAttribute{{Key: "not_allowed", Value: "Vlad", Index: true}}},
-	})
-	err = indexer.IndexTxEvents([]*abci.TxResult{txResult})
-	require.NoError(t, err)
-
-	tx, err := verifyTx(types.Tx(txResult.Tx).Hash())
-	require.NoError(t, err)
-	assert.Equal(t, txResult, tx)
-
-	require.NoError(t, verifyTimeStamp(TableEventTx))
-	require.NoError(t, verifyTimeStamp(TableResultTx))
-
-	tx, err = indexer.GetTxByHash(types.Tx(txResult.Tx).Hash())
-	assert.Nil(t, tx)
-	assert.Equal(t, errors.New("getTxByHash is not supported via the postgres event sink"), err)
-
-	r2, err := indexer.SearchTxEvents(context.TODO(), nil)
-	assert.Nil(t, r2)
-	assert.Equal(t, errors.New("tx search is not supported via the postgres event sink"), err)
-
-	// try to insert the duplicate tx events.
-	err = indexer.IndexTxEvents([]*abci.TxResult{txResult})
-	require.NoError(t, err)
-
-	assert.Nil(t, teardown(t, pool))
-}
-
-func TestStop(t *testing.T) {
-	pool, err := setupDB(t)
-	require.NoError(t, err)
-
-	indexer := &EventSink{store: db}
-	require.NoError(t, indexer.Stop())
-
-	defer db.Close()
-	require.NoError(t, pool.Purge(resource))
-}
-
-func getTestBlockHeader() types.EventDataNewBlockHeader {
-	return types.EventDataNewBlockHeader{
-		Header: types.Header{Height: 1},
-		ResultBeginBlock: abci.ResponseBeginBlock{
-			Events: []abci.Event{
-				{
-					Type: "begin_event",
-					Attributes: []abci.EventAttribute{
-						{
-							Key:   "proposer",
-							Value: "FCAA001",
-							Index: true,
-						},
-					},
-				},
-			},
-		},
-		ResultEndBlock: abci.ResponseEndBlock{
-			Events: []abci.Event{
-				{
-					Type: "end_event",
-					Attributes: []abci.EventAttribute{
-						{
-							Key:   "foo",
-							Value: "100",
-							Index: true,
-						},
-					},
-				},
-			},
-		},
-	}
-}
-
-func readSchema() ([]*schema.Migration, error) {
-
-	filename := "schema.sql"
-	contents, err := ioutil.ReadFile(filename)
-	if err != nil {
-		return nil, fmt.Errorf("failed to read sql file from '%s': %w", filename, err)
-	}
-
-	mg := &schema.Migration{}
-	mg.ID = time.Now().Local().String() + " db schema"
-	mg.Script = string(contents)
-	return append([]*schema.Migration{}, mg), nil
-}
-
-func resetDB(t *testing.T) {
-	q := "DROP TABLE IF EXISTS block_events,tx_events,tx_results"
-	_, err := db.Exec(q)
-
-	require.NoError(t, err)
-
-	q = "DROP TYPE IF EXISTS block_event_type"
-	_, err = db.Exec(q)
-	require.NoError(t, err)
-}
-
-func txResultWithEvents(events []abci.Event) *abci.TxResult {
-	tx := types.Tx("HELLO WORLD")
-	return &abci.TxResult{
-		Height: 1,
-		Index:  0,
-		Tx:     tx,
-		Result: abci.ResponseDeliverTx{
-			Data:   []byte{0},
-			Code:   abci.CodeTypeOK,
-			Log:    "",
-			Events: events,
-		},
-	}
-}
-
-func verifyTx(hash []byte) (*abci.TxResult, error) {
-	join := fmt.Sprintf("%s ON %s.id = tx_result_id", TableEventTx, TableResultTx)
-	sqlStmt := sq.
-		Select("tx_result", fmt.Sprintf("%s.id", TableResultTx), "tx_result_id", "hash", "chain_id").
-		Distinct().From(TableResultTx).
-		InnerJoin(join).
-		Where(fmt.Sprintf("hash = $1 AND chain_id = '%s'", chainID), fmt.Sprintf("%X", hash))
-
-	rows, err := sqlStmt.RunWith(db).Query()
-	if err != nil {
-		return nil, err
-	}
-
-	defer rows.Close()
-
-	if rows.Next() {
-		var txResult []byte
-		var txResultID, txid int
-		var h, cid string
-		err = rows.Scan(&txResult, &txResultID, &txid, &h, &cid)
-		if err != nil {
-			return nil, nil
-		}
-
-		msg := new(abci.TxResult)
-		err = proto.Unmarshal(txResult, msg)
-		if err != nil {
-			return nil, err
-		}
-
-		return msg, err
-	}
-
-	// No result
-	return nil, nil
-}
-
-func verifyTimeStamp(tb string) error {
-
-	// We assume the tx indexing time would not exceed 2 second from now
-	sqlStmt := sq.
-		Select(fmt.Sprintf("%s.created_at", tb)).
-		Distinct().From(tb).
-		Where(fmt.Sprintf("%s.created_at >= $1", tb), time.Now().Add(-2*time.Second))
-
-	rows, err := sqlStmt.RunWith(db).Query()
-	if err != nil {
-		return err
-	}
-
-	defer rows.Close()
-
-	if rows.Next() {
-		var ts string
-		return rows.Scan(&ts)
-	}
-
-	return errors.New("no result")
-}
-
-func verifyBlock(h int64) (bool, error) {
-	sqlStmt := sq.
-		Select("height").
-		Distinct().
-		From(TableEventBlock).
-		Where(fmt.Sprintf("height = %d", h))
-	rows, err := sqlStmt.RunWith(db).Query()
-	if err != nil {
-		return false, err
-	}
-
-	defer rows.Close()
-
-	if !rows.Next() {
-		return false, nil
-	}
-
-	sqlStmt = sq.
-		Select("type, height", "chain_id").
-		Distinct().
-		From(TableEventBlock).
-		Where(fmt.Sprintf("height = %d AND type = '%s' AND chain_id = '%s'", h, types.EventTypeBeginBlock, chainID))
-
-	rows, err = sqlStmt.RunWith(db).Query()
-	if err != nil {
-		return false, err
-	}
-	defer rows.Close()
-
-	if !rows.Next() {
-		return false, nil
-	}
-
-	sqlStmt = sq.
-		Select("type, height").
-		Distinct().
-		From(TableEventBlock).
-		Where(fmt.Sprintf("height = %d AND type = '%s'", h, types.EventTypeEndBlock))
-	rows, err = sqlStmt.RunWith(db).Query()
-
-	if err != nil {
-		return false, err
-	}
-	defer rows.Close()
-
-	return rows.Next(), nil
-}
-
-func setupDB(t *testing.T) (*dockertest.Pool, error) {
-	t.Helper()
+	// Set up docker and start a container running PostgreSQL.
 	pool, err := dockertest.NewPool(os.Getenv("DOCKER_URL"))
+	if err != nil {
+		log.Fatalf("Creating docker pool: %v", err)
+	}
 
-	require.NoError(t, err)
-
-	resource, err = pool.RunWithOptions(&dockertest.RunOptions{
-		Repository: DriverName,
+	resource, err := pool.RunWithOptions(&dockertest.RunOptions{
+		Repository: "postgres",
 		Tag:        "13",
 		Env: []string{
 			"POSTGRES_USER=" + user,
@@ -332,40 +76,269 @@ func setupDB(t *testing.T) (*dockertest.Pool, error) {
 			Name: "no",
 		}
 	})
+	if err != nil {
+		log.Fatalf("Starting docker pool: %v", err)
+	}
 
-	require.NoError(t, err)
-
-	// Set the container to expire in a minute to avoid orphaned containers
-	// hanging around
-	_ = resource.Expire(60)
+	if *doPauseAtExit {
+		log.Print("Pause at exit is enabled, containers will not expire")
+	} else {
+		const expireSeconds = 60
+		_ = resource.Expire(expireSeconds)
+		log.Printf("Container expiration set to %d seconds", expireSeconds)
+	}
 
+	// Connect to the database, clear any leftover data, and install the
+	// indexing schema.
 	conn := fmt.Sprintf(dsn, user, password, resource.GetPort(port+"/tcp"), dbName)
+	var db *sql.DB
 
-	if err = pool.Retry(func() error {
-		var err error
-
-		_, db, err = NewEventSink(conn, chainID)
-
+	if err := pool.Retry(func() error {
+		sink, err := NewEventSink(conn, chainID)
 		if err != nil {
 			return err
 		}
-
+		db = sink.DB() // set global for test use
 		return db.Ping()
 	}); err != nil {
-		require.NoError(t, err)
+		log.Fatalf("Connecting to database: %v", err)
 	}
 
-	resetDB(t)
+	if err := resetDatabase(db); err != nil {
+		log.Fatalf("Flushing database: %v", err)
+	}
 
 	sm, err := readSchema()
-	assert.Nil(t, err)
-	assert.Nil(t, schema.NewMigrator().Apply(db, sm))
-	return pool, nil
+	if err != nil {
+		log.Fatalf("Reading schema: %v", err)
+	} else if err := schema.NewMigrator().Apply(db, sm); err != nil {
+		log.Fatalf("Applying schema: %v", err)
+	}
+
+	// Set up the hook for tests to get the shared database handle.
+	testDB = func() *sql.DB { return db }
+
+	// Run the selected test cases.
+	code := m.Run()
+
+	// Clean up and shut down the database container.
+	if *doPauseAtExit {
+		log.Print("Testing complete, pausing for inspection. Send SIGINT to resume teardown")
+		waitForInterrupt()
+		log.Print("(resuming)")
+	}
+	log.Print("Shutting down database")
+	if err := pool.Purge(resource); err != nil {
+		log.Printf("WARNING: Purging pool failed: %v", err)
+	}
+	if err := db.Close(); err != nil {
+		log.Printf("WARNING: Closing database failed: %v", err)
+	}
+
+	os.Exit(code)
 }
 
-func teardown(t *testing.T, pool *dockertest.Pool) error {
-	t.Helper()
-	// When you're done, kill and remove the container
-	assert.Nil(t, pool.Purge(resource))
-	return db.Close()
+func TestType(t *testing.T) {
+	psqlSink := &EventSink{store: testDB(), chainID: chainID}
+	assert.Equal(t, indexer.PSQL, psqlSink.Type())
+}
+
+func TestIndexing(t *testing.T) {
+	t.Run("IndexBlockEvents", func(t *testing.T) {
+		indexer := &EventSink{store: testDB(), chainID: chainID}
+		require.NoError(t, indexer.IndexBlockEvents(newTestBlockHeader()))
+
+		verifyBlock(t, 1)
+		verifyBlock(t, 2)
+
+		verifyNotImplemented(t, "hasBlock", func() (bool, error) { return indexer.HasBlock(1) })
+		verifyNotImplemented(t, "hasBlock", func() (bool, error) { return indexer.HasBlock(2) })
+
+		verifyNotImplemented(t, "block search", func() (bool, error) {
+			v, err := indexer.SearchBlockEvents(context.Background(), nil)
+			return v != nil, err
+		})
+
+		require.NoError(t, verifyTimeStamp(tableBlocks))
+
+		// Attempting to reindex the same events should gracefully succeed.
+		require.NoError(t, indexer.IndexBlockEvents(newTestBlockHeader()))
+	})
+
+	t.Run("IndexTxEvents", func(t *testing.T) {
+		indexer := &EventSink{store: testDB(), chainID: chainID}
+
+		txResult := txResultWithEvents([]abci.Event{
+			makeIndexedEvent("account.number", "1"),
+			makeIndexedEvent("account.owner", "Ivan"),
+			makeIndexedEvent("account.owner", "Yulieta"),
+
+			{Type: "", Attributes: []abci.EventAttribute{{Key: "not_allowed", Value: "Vlad", Index: true}}},
+		})
+		require.NoError(t, indexer.IndexTxEvents([]*abci.TxResult{txResult}))
+
+		txr, err := loadTxResult(types.Tx(txResult.Tx).Hash())
+		require.NoError(t, err)
+		assert.Equal(t, txResult, txr)
+
+		require.NoError(t, verifyTimeStamp(tableTxResults))
+		require.NoError(t, verifyTimeStamp(viewTxEvents))
+
+		verifyNotImplemented(t, "getTxByHash", func() (bool, error) {
+			txr, err := indexer.GetTxByHash(types.Tx(txResult.Tx).Hash())
+			return txr != nil, err
+		})
+		verifyNotImplemented(t, "tx search", func() (bool, error) {
+			txr, err := indexer.SearchTxEvents(context.Background(), nil)
+			return txr != nil, err
+		})
+
+		// try to insert the duplicate tx events.
+		err = indexer.IndexTxEvents([]*abci.TxResult{txResult})
+		require.NoError(t, err)
+	})
+}
+
+func TestStop(t *testing.T) {
+	indexer := &EventSink{store: testDB()}
+	require.NoError(t, indexer.Stop())
+}
+
+// newTestBlockHeader constructs a fresh copy of a block header containing
+// known test values to exercise the indexer.
+func newTestBlockHeader() types.EventDataNewBlockHeader {
+	return types.EventDataNewBlockHeader{
+		Header: types.Header{Height: 1},
+		ResultBeginBlock: abci.ResponseBeginBlock{
+			Events: []abci.Event{
+				makeIndexedEvent("begin_event.proposer", "FCAA001"),
+				makeIndexedEvent("thingy.whatzit", "O.O"),
+			},
+		},
+		ResultEndBlock: abci.ResponseEndBlock{
+			Events: []abci.Event{
+				makeIndexedEvent("end_event.foo", "100"),
+				makeIndexedEvent("thingy.whatzit", "-.O"),
+			},
+		},
+	}
+}
+
+// readSchema loads the indexing database schema file
+func readSchema() ([]*schema.Migration, error) {
+	const filename = "schema.sql"
+	contents, err := ioutil.ReadFile(filename)
+	if err != nil {
+		return nil, fmt.Errorf("failed to read sql file from '%s': %w", filename, err)
+	}
+
+	return []*schema.Migration{{
+		ID:     time.Now().Local().String() + " db schema",
+		Script: string(contents),
+	}}, nil
+}
+
+// resetDB drops all the data from the test database.
+func resetDatabase(db *sql.DB) error {
+	_, err := db.Exec(`DROP TABLE IF EXISTS blocks,tx_results,events,attributes CASCADE;`)
+	if err != nil {
+		return fmt.Errorf("dropping tables: %v", err)
+	}
+	_, err = db.Exec(`DROP VIEW IF EXISTS event_attributes,block_events,tx_events CASCADE;`)
+	if err != nil {
+		return fmt.Errorf("dropping views: %v", err)
+	}
+	return nil
+}
+
+// txResultWithEvents constructs a fresh transaction result with fixed values
+// for testing, that includes the specified events.
+func txResultWithEvents(events []abci.Event) *abci.TxResult {
+	return &abci.TxResult{
+		Height: 1,
+		Index:  0,
+		Tx:     types.Tx("HELLO WORLD"),
+		Result: abci.ResponseDeliverTx{
+			Data:   []byte{0},
+			Code:   abci.CodeTypeOK,
+			Log:    "",
+			Events: events,
+		},
+	}
+}
+
+func loadTxResult(hash []byte) (*abci.TxResult, error) {
+	hashString := fmt.Sprintf("%X", hash)
+	var resultData []byte
+	if err := testDB().QueryRow(`
+SELECT tx_result FROM `+tableTxResults+` WHERE tx_hash = $1;
+`, hashString).Scan(&resultData); err != nil {
+		return nil, fmt.Errorf("lookup transaction for hash %q failed: %v", hashString, err)
+	}
+
+	txr := new(abci.TxResult)
+	if err := proto.Unmarshal(resultData, txr); err != nil {
+		return nil, fmt.Errorf("unmarshaling txr: %v", err)
+	}
+
+	return txr, nil
+}
+
+func verifyTimeStamp(tableName string) error {
+	return testDB().QueryRow(fmt.Sprintf(`
+SELECT DISTINCT %[1]s.created_at
+  FROM %[1]s
+  WHERE %[1]s.created_at >= $1;
+`, tableName), time.Now().Add(-2*time.Second)).Err()
+}
+
+func verifyBlock(t *testing.T, height int64) {
+	// Check that the blocks table contains an entry for this height.
+	if err := testDB().QueryRow(`
+SELECT height FROM `+tableBlocks+` WHERE height = $1;
+`, height).Err(); err == sql.ErrNoRows {
+		t.Errorf("No block found for height=%d", height)
+	} else if err != nil {
+		t.Fatalf("Database query failed: %v", err)
+	}
+
+	// Verify the presence of begin_block and end_block events.
+	if err := testDB().QueryRow(`
+SELECT type, height, chain_id FROM `+viewBlockEvents+`
+  WHERE height = $1 AND type = $2 AND chain_id = $3;
+`, height, types.EventTypeBeginBlock, chainID).Err(); err == sql.ErrNoRows {
+		t.Errorf("No %q event found for height=%d", types.EventTypeBeginBlock, height)
+	} else if err != nil {
+		t.Fatalf("Database query failed: %v", err)
+	}
+
+	if err := testDB().QueryRow(`
+SELECT type, height, chain_id FROM `+viewBlockEvents+`
+  WHERE height = $1 AND type = $2 AND chain_id = $3;
+`, height, types.EventTypeEndBlock, chainID).Err(); err == sql.ErrNoRows {
+		t.Errorf("No %q event found for height=%d", types.EventTypeEndBlock, height)
+	} else if err != nil {
+		t.Fatalf("Database query failed: %v", err)
+	}
+}
+
+// verifyNotImplemented calls f and verifies that it returns both a
+// false-valued flag and a non-nil error whose string matching the expected
+// "not supported" message with label prefixed.
+func verifyNotImplemented(t *testing.T, label string, f func() (bool, error)) {
+	t.Helper()
+	t.Logf("Verifying that %q reports it is not implemented", label)
+
+	want := label + " is not supported via the postgres event sink"
+	ok, err := f()
+	assert.False(t, ok)
+	require.NotNil(t, err)
+	assert.Equal(t, want, err.Error())
+}
+
+// waitForInterrupt blocks until a SIGINT is received by the process.
+func waitForInterrupt() {
+	ch := make(chan os.Signal, 1)
+	signal.Notify(ch, os.Interrupt)
+	<-ch
 }

state/indexer/sink/psql/schema.sql

@@ -1,32 +1,85 @@
-CREATE TYPE block_event_type AS ENUM ('begin_block', 'end_block', '');
-CREATE TABLE block_events (
-    id SERIAL PRIMARY KEY,
-    key VARCHAR NOT NULL,
-    value VARCHAR NOT NULL,
-    height INTEGER NOT NULL,
-    type block_event_type,
-    created_at TIMESTAMPTZ NOT NULL,
-    chain_id VARCHAR NOT NULL,
-    UNIQUE (key, height)
+/*
+  This file defines the database schema for the PostgresQL ("psql") event sink
+  implementation in Tendermint. The operator must create a database and install
+  this schema before using the database to index events.
+ */
+
+-- The blocks table records metadata about each block.
+-- The block record does not include its events or transactions (see tx_results).
+CREATE TABLE blocks (
+  rowid      BIGSERIAL PRIMARY KEY,
+
+  height     BIGINT NOT NULL,
+  chain_id   VARCHAR NOT NULL,
+
+  -- When this block header was logged into the sink, in UTC.
+  created_at TIMESTAMPTZ NOT NULL,
+
+  UNIQUE (height, chain_id)
 );
+
+-- Index blocks by height and chain, since we need to resolve block IDs when
+-- indexing transaction records and transaction events.
+CREATE INDEX idx_blocks_height_chain ON blocks(height, chain_id);
+
+-- The tx_results table records metadata about transaction results.  Note that
+-- the events from a transaction are stored separately.
 CREATE TABLE tx_results (
-    id SERIAL PRIMARY KEY,
-    tx_result BYTEA NOT NULL,
-    created_at TIMESTAMPTZ NOT NULL,
-    UNIQUE (tx_result)
+  rowid BIGSERIAL PRIMARY KEY,
+
+  -- The block to which this transaction belongs.
+  block_id BIGINT NOT NULL REFERENCES blocks(rowid),
+  -- The sequential index of the transaction within the block.
+  index INTEGER NOT NULL,
+  -- When this result record was logged into the sink, in UTC.
+  created_at TIMESTAMPTZ NOT NULL,
+  -- The hex-encoded hash of the transaction.
+  tx_hash VARCHAR NOT NULL,
+  -- The protobuf wire encoding of the TxResult message.
+  tx_result BYTEA NOT NULL,
+
+  UNIQUE (block_id, index)
 );
-CREATE TABLE tx_events (
-    id SERIAL PRIMARY KEY,
-    key VARCHAR NOT NULL,
-    value VARCHAR NOT NULL,
-    height INTEGER NOT NULL,
-    hash VARCHAR NOT NULL,
-    tx_result_id SERIAL,
-    created_at TIMESTAMPTZ NOT NULL,
-    chain_id VARCHAR NOT NULL,
-    UNIQUE (hash, key),
-    FOREIGN KEY (tx_result_id) REFERENCES tx_results(id) ON DELETE CASCADE
+
+-- The events table records events. All events (both block and transaction) are
+-- associated with a block ID; transaction events also have a transaction ID.
+CREATE TABLE events (
+  rowid BIGSERIAL PRIMARY KEY,
+
+  -- The block and transaction this event belongs to.
+  -- If tx_id is NULL, this is a block event.
+  block_id BIGINT NOT NULL REFERENCES blocks(rowid),
+  tx_id    BIGINT NULL REFERENCES tx_results(rowid),
+
+  -- The application-defined type label for the event.
+  type VARCHAR NOT NULL
 );
-CREATE INDEX idx_block_events_key_value ON block_events(key, value);
-CREATE INDEX idx_tx_events_key_value ON tx_events(key, value);
-CREATE INDEX idx_tx_events_hash ON tx_events(hash);
+
+-- The attributes table records event attributes.
+CREATE TABLE attributes (
+   event_id      BIGINT NOT NULL REFERENCES events(rowid),
+   key           VARCHAR NOT NULL, -- bare key
+   composite_key VARCHAR NOT NULL, -- composed type.key
+   value         VARCHAR NULL,
+
+   UNIQUE (event_id, key)
+);
+
+-- A joined view of events and their attributes. Events that do not have any
+-- attributes are represented as a single row with empty key and value fields.
+CREATE VIEW event_attributes AS
+  SELECT block_id, tx_id, type, key, composite_key, value
+  FROM events LEFT JOIN attributes ON (events.rowid = attributes.event_id);
+
+-- A joined view of all block events (those having tx_id NULL).
+CREATE VIEW block_events AS
+  SELECT blocks.rowid as block_id, height, chain_id, type, key, composite_key, value
+  FROM blocks JOIN event_attributes ON (blocks.rowid = event_attributes.block_id)
+  WHERE event_attributes.tx_id IS NULL;
+
+-- A joined view of all transaction events.
+CREATE VIEW tx_events AS
+  SELECT height, index, chain_id, type, key, composite_key, value, tx_results.created_at
+  FROM blocks JOIN tx_results ON (blocks.rowid = tx_results.block_id)
+  JOIN event_attributes ON (tx_results.rowid = event_attributes.tx_id)
+  WHERE event_attributes.tx_id IS NOT NULL;

state/indexer/sink/sink.go

@@ -0,0 +1,65 @@
+package sink
+
+import (
+	"errors"
+	"strings"
+
+	"github.com/tendermint/tendermint/config"
+	"github.com/tendermint/tendermint/state/indexer"
+	"github.com/tendermint/tendermint/state/indexer/sink/kv"
+	"github.com/tendermint/tendermint/state/indexer/sink/null"
+	"github.com/tendermint/tendermint/state/indexer/sink/psql"
+)
+
+// EventSinksFromConfig constructs a slice of indexer.EventSink using the provided
+// configuration.
+//
+//nolint:lll
+func EventSinksFromConfig(cfg *config.Config, dbProvider config.DBProvider, chainID string) ([]indexer.EventSink, error) {
+	if len(cfg.TxIndex.Indexer) == 0 {
+		return []indexer.EventSink{null.NewEventSink()}, nil
+	}
+
+	// check for duplicated sinks
+	sinks := map[string]struct{}{}
+	for _, s := range cfg.TxIndex.Indexer {
+		sl := strings.ToLower(s)
+		if _, ok := sinks[sl]; ok {
+			return nil, errors.New("found duplicated sinks, please check the tx-index section in the config.toml")
+		}
+		sinks[sl] = struct{}{}
+	}
+	eventSinks := []indexer.EventSink{}
+	for k := range sinks {
+		switch indexer.EventSinkType(k) {
+		case indexer.NULL:
+			// When we see null in the config, the eventsinks will be reset with the
+			// nullEventSink.
+			return []indexer.EventSink{null.NewEventSink()}, nil
+
+		case indexer.KV:
+			store, err := dbProvider(&config.DBContext{ID: "tx_index", Config: cfg})
+			if err != nil {
+				return nil, err
+			}
+
+			eventSinks = append(eventSinks, kv.NewEventSink(store))
+
+		case indexer.PSQL:
+			conn := cfg.TxIndex.PsqlConn
+			if conn == "" {
+				return nil, errors.New("the psql connection settings cannot be empty")
+			}
+
+			es, err := psql.NewEventSink(conn, chainID)
+			if err != nil {
+				return nil, err
+			}
+			eventSinks = append(eventSinks, es)
+		default:
+			return nil, errors.New("unsupported event sink type")
+		}
+	}
+	return eventSinks, nil
+
+}

state/state.go

@@ -9,7 +9,6 @@ import (
 
 	"github.com/gogo/protobuf/proto"
 
-	tmtime "github.com/tendermint/tendermint/libs/time"
 	tmstate "github.com/tendermint/tendermint/proto/tendermint/state"
 	tmversion "github.com/tendermint/tendermint/proto/tendermint/version"
 	"github.com/tendermint/tendermint/types"
@@ -287,7 +286,7 @@ func (state State) MakeBlock(
 // the votes sent by honest processes, i.e., a faulty processes can not arbitrarily increase or decrease the
 // computed value.
 func MedianTime(commit *types.Commit, validators *types.ValidatorSet) time.Time {
-	weightedTimes := make([]*tmtime.WeightedTime, len(commit.Signatures))
+	weightedTimes := make([]*weightedTime, len(commit.Signatures))
 	totalVotingPower := int64(0)
 
 	for i, commitSig := range commit.Signatures {
@@ -298,11 +297,11 @@ func MedianTime(commit *types.Commit, validators *types.ValidatorSet) time.Time
 		// If there's no condition, TestValidateBlockCommit panics; not needed normally.
 		if validator != nil {
 			totalVotingPower += validator.VotingPower
-			weightedTimes[i] = tmtime.NewWeightedTime(commitSig.Timestamp, validator.VotingPower)
+			weightedTimes[i] = newWeightedTime(commitSig.Timestamp, validator.VotingPower)
 		}
 	}
 
-	return tmtime.WeightedMedian(weightedTimes, totalVotingPower)
+	return weightedMedian(weightedTimes, totalVotingPower)
 }
 
 //------------------------------------------------------------------------

state/time.go

@@ -0,0 +1,46 @@
+package state
+
+import (
+	"sort"
+	"time"
+)
+
+// weightedTime for computing a median.
+type weightedTime struct {
+	Time   time.Time
+	Weight int64
+}
+
+// newWeightedTime with time and weight.
+func newWeightedTime(time time.Time, weight int64) *weightedTime {
+	return &weightedTime{
+		Time:   time,
+		Weight: weight,
+	}
+}
+
+// weightedMedian computes weighted median time for a given array of WeightedTime and the total voting power.
+func weightedMedian(weightedTimes []*weightedTime, totalVotingPower int64) (res time.Time) {
+	median := totalVotingPower / 2
+
+	sort.Slice(weightedTimes, func(i, j int) bool {
+		if weightedTimes[i] == nil {
+			return false
+		}
+		if weightedTimes[j] == nil {
+			return true
+		}
+		return weightedTimes[i].Time.UnixNano() < weightedTimes[j].Time.UnixNano()
+	})
+
+	for _, weightedTime := range weightedTimes {
+		if weightedTime != nil {
+			if median <= weightedTime.Weight {
+				res = weightedTime.Time
+				break
+			}
+			median -= weightedTime.Weight
+		}
+	}
+	return
+}

state/time_test.go

@@ -1,54 +1,55 @@
-package time
+package state
 
 import (
 	"testing"
 	"time"
 
 	"github.com/stretchr/testify/assert"
+	tmtime "github.com/tendermint/tendermint/libs/time"
 )
 
 func TestWeightedMedian(t *testing.T) {
-	m := make([]*WeightedTime, 3)
+	m := make([]*weightedTime, 3)
 
-	t1 := Now()
+	t1 := tmtime.Now()
 	t2 := t1.Add(5 * time.Second)
 	t3 := t1.Add(10 * time.Second)
 
-	m[2] = NewWeightedTime(t1, 33) // faulty processes
-	m[0] = NewWeightedTime(t2, 40) // correct processes
-	m[1] = NewWeightedTime(t3, 27) // correct processes
+	m[2] = newWeightedTime(t1, 33) // faulty processes
+	m[0] = newWeightedTime(t2, 40) // correct processes
+	m[1] = newWeightedTime(t3, 27) // correct processes
 	totalVotingPower := int64(100)
 
-	median := WeightedMedian(m, totalVotingPower)
+	median := weightedMedian(m, totalVotingPower)
 	assert.Equal(t, t2, median)
 	// median always returns value between values of correct processes
 	assert.Equal(t, true, (median.After(t1) || median.Equal(t1)) &&
 		(median.Before(t3) || median.Equal(t3)))
 
-	m[1] = NewWeightedTime(t1, 40) // correct processes
-	m[2] = NewWeightedTime(t2, 27) // correct processes
-	m[0] = NewWeightedTime(t3, 33) // faulty processes
+	m[1] = newWeightedTime(t1, 40) // correct processes
+	m[2] = newWeightedTime(t2, 27) // correct processes
+	m[0] = newWeightedTime(t3, 33) // faulty processes
 	totalVotingPower = int64(100)
 
-	median = WeightedMedian(m, totalVotingPower)
+	median = weightedMedian(m, totalVotingPower)
 	assert.Equal(t, t2, median)
 	// median always returns value between values of correct processes
 	assert.Equal(t, true, (median.After(t1) || median.Equal(t1)) &&
 		(median.Before(t2) || median.Equal(t2)))
 
-	m = make([]*WeightedTime, 8)
+	m = make([]*weightedTime, 8)
 	t4 := t1.Add(15 * time.Second)
 	t5 := t1.Add(60 * time.Second)
 
-	m[3] = NewWeightedTime(t1, 10) // correct processes
-	m[1] = NewWeightedTime(t2, 10) // correct processes
-	m[5] = NewWeightedTime(t2, 10) // correct processes
-	m[4] = NewWeightedTime(t3, 23) // faulty processes
-	m[0] = NewWeightedTime(t4, 20) // correct processes
-	m[7] = NewWeightedTime(t5, 10) // faulty processes
+	m[3] = newWeightedTime(t1, 10) // correct processes
+	m[1] = newWeightedTime(t2, 10) // correct processes
+	m[5] = newWeightedTime(t2, 10) // correct processes
+	m[4] = newWeightedTime(t3, 23) // faulty processes
+	m[0] = newWeightedTime(t4, 20) // correct processes
+	m[7] = newWeightedTime(t5, 10) // faulty processes
 	totalVotingPower = int64(83)
 
-	median = WeightedMedian(m, totalVotingPower)
+	median = weightedMedian(m, totalVotingPower)
 	assert.Equal(t, t3, median)
 	// median always returns value between values of correct processes
 	assert.Equal(t, true, (median.After(t1) || median.Equal(t1)) &&

test/e2e/generator/generate.go

@@ -26,13 +26,29 @@ var (
 	}
 
 	// The following specify randomly chosen values for testnet nodes.
-	nodeDatabases        = uniformChoice{"goleveldb", "cleveldb", "rocksdb", "boltdb", "badgerdb"}
-	nodeABCIProtocols    = uniformChoice{"unix", "tcp", "builtin", "grpc"}
-	nodePrivvalProtocols = uniformChoice{"file", "unix", "tcp", "grpc"}
+	nodeDatabases = weightedChoice{
+		"goleveldb": 35,
+		"badgerdb":  35,
+		"boltdb":    15,
+		"rocksdb":   10,
+		"cleveldb":  5,
+	}
+	nodeABCIProtocols = weightedChoice{
+		"builtin": 50,
+		"tcp":     20,
+		"grpc":    20,
+		"unix":    10,
+	}
+	nodePrivvalProtocols = weightedChoice{
+		"file": 50,
+		"grpc": 20,
+		"tcp":  20,
+		"unix": 10,
+	}
 	// FIXME: v2 disabled due to flake
 	nodeBlockSyncs        = uniformChoice{"v0"} // "v2"
 	nodeMempools          = uniformChoice{"v0", "v1"}
-	nodeStateSyncs        = uniformChoice{false, true}
+	nodeStateSyncs        = uniformChoice{e2e.StateSyncDisabled, e2e.StateSyncP2P, e2e.StateSyncRPC}
 	nodePersistIntervals  = uniformChoice{0, 1, 5}
 	nodeSnapshotIntervals = uniformChoice{0, 3}
 	nodeRetainBlocks      = uniformChoice{0, int(e2e.EvidenceAgeHeight), int(e2e.EvidenceAgeHeight) + 5}
@@ -107,11 +123,11 @@ func generateTestnet(r *rand.Rand, opt map[string]interface{}) (e2e.Manifest, er
 
 	switch opt["p2p"].(P2PMode) {
 	case NewP2PMode:
-		manifest.DisableLegacyP2P = true
+		manifest.UseLegacyP2P = true
 	case LegacyP2PMode:
-		manifest.DisableLegacyP2P = false
+		manifest.UseLegacyP2P = false
 	case HybridP2PMode:
-		manifest.DisableLegacyP2P = false
+		manifest.UseLegacyP2P = true
 		p2pNodeFactor = 2
 	default:
 		return manifest, fmt.Errorf("unknown p2p mode %s", opt["p2p"])
@@ -138,9 +154,9 @@ func generateTestnet(r *rand.Rand, opt map[string]interface{}) (e2e.Manifest, er
 		node := generateNode(r, e2e.ModeSeed, 0, manifest.InitialHeight, false)
 
 		if p2pNodeFactor == 0 {
-			node.DisableLegacyP2P = manifest.DisableLegacyP2P
+			node.UseLegacyP2P = manifest.UseLegacyP2P
 		} else if p2pNodeFactor%i == 0 {
-			node.DisableLegacyP2P = !manifest.DisableLegacyP2P
+			node.UseLegacyP2P = !manifest.UseLegacyP2P
 		}
 
 		manifest.Nodes[fmt.Sprintf("seed%02d", i)] = node
@@ -162,9 +178,9 @@ func generateTestnet(r *rand.Rand, opt map[string]interface{}) (e2e.Manifest, er
 			r, e2e.ModeValidator, startAt, manifest.InitialHeight, i <= 2)
 
 		if p2pNodeFactor == 0 {
-			node.DisableLegacyP2P = manifest.DisableLegacyP2P
+			node.UseLegacyP2P = manifest.UseLegacyP2P
 		} else if p2pNodeFactor%i == 0 {
-			node.DisableLegacyP2P = !manifest.DisableLegacyP2P
+			node.UseLegacyP2P = !manifest.UseLegacyP2P
 		}
 
 		manifest.Nodes[name] = node
@@ -198,9 +214,9 @@ func generateTestnet(r *rand.Rand, opt map[string]interface{}) (e2e.Manifest, er
 		node := generateNode(r, e2e.ModeFull, startAt, manifest.InitialHeight, false)
 
 		if p2pNodeFactor == 0 {
-			node.DisableLegacyP2P = manifest.DisableLegacyP2P
+			node.UseLegacyP2P = manifest.UseLegacyP2P
 		} else if p2pNodeFactor%i == 0 {
-			node.DisableLegacyP2P = !manifest.DisableLegacyP2P
+			node.UseLegacyP2P = !manifest.UseLegacyP2P
 		}
 		manifest.Nodes[fmt.Sprintf("full%02d", i)] = node
 	}
@@ -270,18 +286,22 @@ func generateNode(
 	node := e2e.ManifestNode{
 		Mode:             string(mode),
 		StartAt:          startAt,
-		Database:         nodeDatabases.Choose(r).(string),
-		ABCIProtocol:     nodeABCIProtocols.Choose(r).(string),
-		PrivvalProtocol:  nodePrivvalProtocols.Choose(r).(string),
+		Database:         nodeDatabases.Choose(r),
+		ABCIProtocol:     nodeABCIProtocols.Choose(r),
+		PrivvalProtocol:  nodePrivvalProtocols.Choose(r),
 		BlockSync:        nodeBlockSyncs.Choose(r).(string),
 		Mempool:          nodeMempools.Choose(r).(string),
-		StateSync:        nodeStateSyncs.Choose(r).(bool) && startAt > 0,
+		StateSync:        e2e.StateSyncDisabled,
 		PersistInterval:  ptrUint64(uint64(nodePersistIntervals.Choose(r).(int))),
 		SnapshotInterval: uint64(nodeSnapshotIntervals.Choose(r).(int)),
 		RetainBlocks:     uint64(nodeRetainBlocks.Choose(r).(int)),
 		Perturb:          nodePerturbations.Choose(r),
 	}
 
+	if startAt > 0 {
+		node.StateSync = nodeStateSyncs.Choose(r).(string)
+	}
+
 	// If this node is forced to be an archive node, retain all blocks and
 	// enable state sync snapshotting.
 	if forceArchive {
@@ -310,7 +330,7 @@ func generateNode(
 		}
 	}
 
-	if node.StateSync {
+	if node.StateSync != e2e.StateSyncDisabled {
 		node.BlockSync = "v0"
 	}
 
@@ -321,7 +341,7 @@ func generateLightNode(r *rand.Rand, startAt int64, providers []string) *e2e.Man
 	return &e2e.ManifestNode{
 		Mode:            string(e2e.ModeLight),
 		StartAt:         startAt,
-		Database:        nodeDatabases.Choose(r).(string),
+		Database:        nodeDatabases.Choose(r),
 		ABCIProtocol:    "builtin",
 		PersistInterval: ptrUint64(0),
 		PersistentPeers: providers,