Compare commits

...
Author SHA1 Message Date
Chris LuandGitHub f6df7126b6 feat(admin): add profiling options for debugging high memory/CPU usage (#8923)
* feat(admin): add profiling options for debugging high memory/CPU usage

Add -debug, -debug.port, -cpuprofile, and -memprofile flags to the admin
command, matching the profiling support already available in master, volume,
and other server commands. This enables investigation of resource usage
issues like #8919.

* refactor(admin): move profiling flags into AdminOptions struct

Move cpuprofile and memprofile flags from global variables into the
AdminOptions struct and init() function for consistency with other flags.

* fix(debug): bind pprof server to localhost only and document profiling flags

StartDebugServer was binding to all interfaces (0.0.0.0), exposing
runtime profiling data to the network. Restrict to 127.0.0.1 since
this is a development/debugging tool.

Also add a "Debugging and Profiling" section to the admin command's
help text documenting the new flags.
2026-04-04 10:05:19 -07:00
Chris LuandGitHub 9add18e169 fix(volume-rust): fix volume balance between Go and Rust servers (#8915)
Two bugs prevented reliable volume balancing when a Rust volume server
is the copy target:

1. find_last_append_at_ns returned None for delete tombstones (Size==0
   in dat header), falling back to file mtime truncated to seconds.
   This caused the tail step to re-send needles from the last sub-second
   window. Fix: change `needle_size <= 0` to `< 0` since Size==0 delete
   needles still have a valid timestamp in their tail.

2. VolumeTailReceiver called read_body_v2 on delete needles, which have
   no DataSize/Data/flags — only checksum+timestamp+padding after the
   header. Fix: skip read_body_v2 when size == 0, reject negative sizes.

Also:
- Unify gRPC server bind: use TcpListener::bind before spawn for both
  TLS and non-TLS paths, propagating bind errors at startup.
- Add mixed Go+Rust cluster test harness and integration tests covering
  VolumeCopy in both directions, copy with deletes, and full balance
  move with tail tombstone propagation and source deletion.
- Make FindOrBuildRustBinary configurable for default vs no-default
  features (4-byte vs 5-byte offsets).
2026-04-04 09:13:23 -07:00
Chris LuandGitHub d1823d3784 fix(s3): include static identities in listing operations (#8903)
* fix(s3): include static identities in listing operations

Static identities loaded from -s3.config file were only stored in the
S3 API server's in-memory state. Listing operations (s3.configure shell
command, aws iam list-users) queried the credential manager which only
returned dynamic identities from the backend store.

Register static identities with the credential manager after loading
so they are included in LoadConfiguration and ListUsers results, and
filtered out before SaveConfiguration to avoid persisting them to the
dynamic store.

Fixes https://github.com/seaweedfs/seaweedfs/discussions/8896

* fix: avoid mutating caller's config and defensive copies

- SaveConfiguration: use shallow struct copy instead of mutating the
  caller's config.Identities field
- SetStaticIdentities: skip nil entries to avoid panics
- GetStaticIdentities: defensively copy PolicyNames slice to avoid
  aliasing the original

* fix: filter nil static identities and sync on config reload

- SetStaticIdentities: filter nil entries from the stored slice (not
  just from staticNames) to prevent panics in LoadConfiguration/ListUsers
- Extract updateCredentialManagerStaticIdentities helper and call it
  from both startup and the grace.OnReload handler so the credential
  manager's static snapshot stays current after config file reloads

* fix: add mutex for static identity fields and fix ListUsers for store callers

- Add sync.RWMutex to protect staticIdentities/staticNames against
  concurrent reads during config reload
- Revert CredentialManager.ListUsers to return only store users, since
  internal callers (e.g. DeletePolicy) look up each user in the store
  and fail on non-existent static entries
- Merge static usernames in the filer gRPC ListUsers handler instead,
  via the new GetStaticUsernames method
- Fix CI: TestIAMPolicyManagement/managed_policy_crud_lifecycle was
  failing because DeletePolicy iterated static users that don't exist
  in the store

* fix: show static identities in admin UI and weed shell

The admin UI and weed shell s3.configure command query the filer's
credential manager via gRPC, which is a separate instance from the S3
server's credential manager. Static identities were only registered
on the S3 server's credential manager, so they never appeared in the
filer's responses.

- Add CredentialManager.LoadS3ConfigFile to parse a static S3 config
  file and register its identities
- Add FilerOptions.s3ConfigFile so the filer can load the same static
  config that the S3 server uses
- Wire s3ConfigFile through in weed mini and weed server modes
- Merge static usernames in filer gRPC ListUsers handler
- Add CredentialManager.GetStaticUsernames helper
- Add sync.RWMutex to protect concurrent access to static identity
  fields
- Avoid importing weed/filer from weed/credential (which pulled in
  filer store init() registrations and broke test isolation)
- Add docker/compose/s3_static_users_example.json

* fix(admin): make static users read-only in admin UI

Static users loaded from the -s3.config file should not be editable
or deletable through the admin UI since they are managed via the
config file.

- Add IsStatic field to ObjectStoreUser, set from credential manager
- Hide edit, delete, and access key buttons for static users in the
  users table template
- Show a "static" badge next to static user names
- Return 403 Forbidden from UpdateUser and DeleteUser API handlers
  when the target user is a static identity

* fix(admin): show details for static users

GetObjectStoreUserDetails called credentialManager.GetUser which only
queries the dynamic store. For static users this returned
ErrUserNotFound. Fall back to GetStaticIdentity when the store lookup
fails.

* fix(admin): load static S3 identities in admin server

The admin server has its own credential manager (gRPC store) which is
a separate instance from the S3 server's and filer's. It had no static
identity data, so IsStaticIdentity returned false (edit/delete buttons
shown) and GetStaticIdentity returned nil (details page failed).

Pass the -s3.config file path through to the admin server and call
LoadS3ConfigFile on its credential manager, matching the approach
used for the filer.

* fix: use protobuf is_static field instead of passing config file path

The previous approach passed -s3.config file path to every component
(filer, admin). This is wrong because the admin server should not need
to know about S3 config files.

Instead, add an is_static field to the Identity protobuf message.
The field is set when static identities are serialized (in
GetStaticIdentities and LoadS3ConfigFile). Any gRPC client that loads
configuration via GetConfiguration automatically sees which identities
are static, without needing the config file.

- Add is_static field (tag 8) to iam_pb.Identity proto message
- Set IsStatic=true in GetStaticIdentities and LoadS3ConfigFile
- Admin GetObjectStoreUsers reads identity.IsStatic from proto
- Admin IsStaticUser helper loads config via gRPC to check the flag
- Filer GetUser gRPC handler falls back to GetStaticIdentity
- Remove s3ConfigFile from AdminOptions and NewAdminServer signature
2026-04-03 20:01:28 -07:00
Chris LuandGitHub 0798b274dd feat(s3): add concurrent chunk prefetch for large file downloads (#8917)
* feat(s3): add concurrent chunk prefetch for large file downloads

Add a pipe-based prefetch pipeline that overlaps chunk fetching with
response writing during S3 GetObject, SSE downloads, and filer proxy.

While chunk N streams to the HTTP response, fetch goroutines for the
next K chunks establish HTTP connections to volume servers ahead of
time, eliminating the RTT gap between sequential chunk fetches.

Uses io.Pipe for minimal memory overhead (~1MB per download regardless
of chunk size, vs buffering entire chunks). Also increases the
streaming read buffer from 64KB to 256KB to reduce syscall overhead.

Benchmark results (64KB chunks, prefetch=4):
- 0ms latency:  1058 → 2362 MB/s (2.2× faster)
- 5ms latency:  11.0 → 41.7 MB/s (3.8× faster)
- 10ms latency: 5.9  → 23.3 MB/s (4.0× faster)
- 20ms latency: 3.1  → 12.1 MB/s (3.9× faster)

* fix: address review feedback for prefetch pipeline

- Fix data race: use *chunkPipeResult (pointer) on channel to avoid
  copying struct while fetch goroutines write to it. Confirmed clean
  with -race detector.
- Remove concurrent map write: retryWithCacheInvalidation no longer
  updates fileId2Url map. Producer only reads it; consumer never writes.
- Use mem.Allocate/mem.Free for copy buffer to reduce GC pressure.
- Add local cancellable context so consumer errors (client disconnect)
  immediately stop the producer and all in-flight fetch goroutines.

* fix(test): remove dead code and add Range header support in test server

- Remove unused allData variable in makeChunksAndServer
- Add Range header handling to createTestServer for partial chunk
  read coverage (206 Partial Content, 416 Range Not Satisfiable)

* fix: correct retry condition and goroutine leak in prefetch pipeline

- Fix retry condition: use result.fetchErr/result.written instead of
  copied to decide cache-invalidation retry. The old condition wrongly
  triggered retry when the fetch succeeded but the response writer
  failed on the first write (copied==0 despite fetcher having data).
  Now matches the sequential path (stream.go:197) which checks whether
  the fetcher itself wrote zero bytes.

- Fix goroutine leak: when the producer's send to the results channel
  is interrupted by context cancellation, the fetch goroutine was
  already launched but the result was never sent to the channel. The
  drain loop couldn't handle it. Now waits on result.done before
  returning so every fetch goroutine is properly awaited.
2026-04-03 19:57:30 -07:00
Chris LuandGitHub 3efe88c718 feat(s3): store and return checksum headers for additional checksum algorithms (#8914)
* feat(s3): store and return checksum headers for additional checksum algorithms

When clients upload with --checksum-algorithm (SHA256, CRC32, etc.),
SeaweedFS validated the checksum but discarded it. The checksum was
never stored in metadata or returned in PUT/HEAD/GET responses.

Now the checksum is computed alongside MD5 during upload, stored in
entry extended attributes, and returned as the appropriate
x-amz-checksum-* header in all responses.

Fixes #8911

* fix(s3): address review feedback and CI failures for checksum support

- Gate GET/HEAD checksum response headers on x-amz-checksum-mode: ENABLED
  per AWS S3 spec, fixing FlexibleChecksumError on ranged GETs and
  multipart copies
- Verify computed checksum against client-provided header value for
  non-chunked uploads, returning BadDigest on mismatch
- Add nil check for getCheckSumWriter to prevent panic
- Handle comma-separated values in X-Amz-Trailer header
- Use ordered slice instead of map for deterministic checksum header
  selection; extract shared mappings into package-level vars

* fix(s3): skip checksum header for ranged GET responses

The stored checksum covers the full object. Returning it for ranged
(partial) responses causes SDK checksum validation failures because the
SDK validates the header value against the partial content received.

Skip emitting x-amz-checksum-* headers when a Range request header is
present, fixing PyArrow large file read failures.

* fix(s3): reject unsupported checksum algorithm with 400

detectRequestedChecksumAlgorithm now returns an error code when
x-amz-sdk-checksum-algorithm or x-amz-checksum-algorithm contains
an unsupported value, instead of silently ignoring it.

* feat(s3): compute composite checksum for multipart uploads

Store the checksum algorithm during CreateMultipartUpload, then during
CompleteMultipartUpload compute a composite checksum from per-part
checksums following the AWS S3 spec: concatenate raw per-part checksums,
hash with the same algorithm, format as "base64-N" where N is part count.

The composite checksum is persisted on the final object entry and
returned in HEAD/GET responses (gated on x-amz-checksum-mode: ENABLED).

Reuses existing per-part checksum storage from putToFiler and the
getCheckSumWriter/checksumHeaders infrastructure.

* fix(s3): validate checksum algorithm in CreateMultipartUpload, error on missing part checksums

- Move detectRequestedChecksumAlgorithm call before mkdir callback so
  an unsupported algorithm returns 400 before the upload is created
- Change computeCompositeChecksum to return an error when a part is
  missing its checksum (the upload was initiated with a checksum
  algorithm, so all parts must have checksums)
- Propagate the error as ErrInvalidPart in CompleteMultipartUpload

* fix(s3): return checksum header in CompleteMultipartUpload response, validate per-part algorithm

- Add ChecksumHeaderName/ChecksumValue fields to CompleteMultipartUploadResult
  and set the x-amz-checksum-* HTTP response header in the handler, matching
  the AWS S3 CompleteMultipartUpload response spec
- Validate that each part's stored checksum algorithm matches the upload's
  expected algorithm before assembling the composite checksum; return an
  error if a part was uploaded with a different algorithm
2026-04-03 18:37:54 -07:00
Chris LuandGitHub 36f37b9b6a fix(filer): remove cancellation guard from RollbackTransaction and clean up #8909 (#8916)
* fix(filer): remove cancellation guard from RollbackTransaction and clean up #8909

RollbackTransaction is a cleanup operation that must succeed even when
the context is cancelled — guarding it causes the exact orphaned state
that #8909 was trying to prevent.

Also:
- Use single-evaluation `if err := ctx.Err(); err != nil` pattern
  instead of double-calling ctx.Err()
- Remove spurious blank lines before guards
- Add context.DeadlineExceeded test coverage
- Simplify tests from ~230 lines to ~130 lines

* fix(filer): call cancel() in expiredCtx and test rollback with expired context

- Call cancel() instead of suppressing it to avoid leaking timer resources
- Test RollbackTransaction with both cancelled and expired contexts
2026-04-03 17:55:27 -07:00
os-pradipbabarandGitHub d5128f00f1 fix: Prevent orphaned metadata from cancelled S3 operations (Issue #8908) (#8909)
fix(filer): check if context was already cancelled before ignoring cancellation
2026-04-03 16:22:46 -07:00
d49c2a7364 chore(weed/admin/plugin): prune unused functions (#8912)
Co-authored-by: Chris Lu <chrislusf@users.noreply.github.com>
2026-04-03 16:05:42 -07:00
Chris LuandGitHub 995dfc4d5d chore: remove ~50k lines of unreachable dead code (#8913)
* chore: remove unreachable dead code across the codebase

Remove ~50,000 lines of unreachable code identified by static analysis.

Major removals:
- weed/filer/redis_lua: entire unused Redis Lua filer store implementation
- weed/wdclient/net2, resource_pool: unused connection/resource pool packages
- weed/plugin/worker/lifecycle: unused lifecycle plugin worker
- weed/s3api: unused S3 policy templates, presigned URL IAM, streaming copy,
  multipart IAM, key rotation, and various SSE helper functions
- weed/mq/kafka: unused partition mapping, compression, schema, and protocol functions
- weed/mq/offset: unused SQL storage and migration code
- weed/worker: unused registry, task, and monitoring functions
- weed/query: unused SQL engine, parquet scanner, and type functions
- weed/shell: unused EC proportional rebalance functions
- weed/storage/erasure_coding/distribution: unused distribution analysis functions
- Individual unreachable functions removed from 150+ files across admin,
  credential, filer, iam, kms, mount, mq, operation, pb, s3api, server,
  shell, storage, topology, and util packages

* fix(s3): reset shared memory store in IAM test to prevent flaky failure

TestLoadIAMManagerFromConfig_EmptyConfigWithFallbackKey was flaky because
the MemoryStore credential backend is a singleton registered via init().
Earlier tests that create anonymous identities pollute the shared store,
causing LookupAnonymous() to unexpectedly return true.

Fix by calling Reset() on the memory store before the test runs.

* style: run gofmt on changed files

* fix: restore KMS functions used by integration tests

* fix(plugin): prevent panic on send to closed worker session channel

The Plugin.sendToWorker method could panic with "send on closed channel"
when a worker disconnected while a message was being sent. The race was
between streamSession.close() closing the outgoing channel and sendToWorker
writing to it concurrently.

Add a done channel to streamSession that is closed before the outgoing
channel, and check it in sendToWorker's select to safely detect closed
sessions without panicking.
2026-04-03 16:04:27 -07:00
Chris LuandGitHub 8fad85aed7 feat(s3): support WEED_S3_SSE_KEY env var for SSE-S3 KEK (#8904)
* feat(s3): support WEED_S3_SSE_KEY env var for SSE-S3 KEK

Add support for providing the SSE-S3 Key Encryption Key (KEK) via the
WEED_S3_SSE_KEY environment variable (hex-encoded 256-bit key). This
avoids storing the master key in plaintext on the filer at /etc/s3/sse_kek.

Key source priority:
1. WEED_S3_SSE_KEY environment variable (recommended)
2. Existing filer KEK at /etc/s3/sse_kek (backward compatible)
3. Auto-generate and save to filer (deprecated for new deployments)

Existing deployments with a filer-stored KEK continue to work unchanged.
A deprecation warning is logged when auto-generating a new filer KEK.

* refactor(s3): derive KEK from any string via HKDF instead of requiring hex

Accept any secret string in WEED_S3_SSE_KEY and derive a 256-bit key
using HKDF-SHA256 instead of requiring a hex-encoded key. This is
simpler for users — no need to generate hex, just set a passphrase.

* feat(s3): add WEED_S3_SSE_KEK and WEED_S3_SSE_KEY env vars for KEK

Two env vars for providing the SSE-S3 Key Encryption Key:

- WEED_S3_SSE_KEK: hex-encoded, same format as /etc/s3/sse_kek.
  If the filer file also exists, they must match.
- WEED_S3_SSE_KEY: any string, 256-bit key derived via HKDF-SHA256.
  Refuses to start if /etc/s3/sse_kek exists (must delete first).

Only one may be set. Existing filer-stored KEKs continue to work.
Auto-generating and storing new KEKs on filer is deprecated.

* fix(s3): stop auto-generating KEK, fail only when SSE-S3 is used

Instead of auto-generating a KEK and storing it on the filer when no
key source is configured, simply leave SSE-S3 disabled. Encrypt and
decrypt operations return a clear error directing the user to set
WEED_S3_SSE_KEK or WEED_S3_SSE_KEY.

* refactor(s3): move SSE-S3 KEK config to security.toml

Move KEK configuration from standalone env vars to security.toml's new
[sse_s3] section, following the same pattern as JWT keys and TLS certs.

  [sse_s3]
  kek = ""   # hex-encoded 256-bit key (same format as /etc/s3/sse_kek)
  key = ""   # any string, HKDF-derived

Viper's WEED_ prefix auto-mapping provides env var support:
WEED_SSE_S3_KEK and WEED_SSE_S3_KEY.

All existing behavior is preserved: filer KEK fallback, mismatch
detection, and HKDF derivation.

* refactor(s3): rename SSE-S3 config keys to s3.sse.kek / s3.sse.key

Use [s3.sse] section in security.toml, matching the existing naming
convention (e.g. [s3.*]). Env vars: WEED_S3_SSE_KEK, WEED_S3_SSE_KEY.

* fix(s3): address code review findings for SSE-S3 KEK

- Don't hold mutex during filer retry loop (up to 20s of sleep).
  Lock only to write filerClient and superKey.
- Remove dead generateAndSaveSuperKeyToFiler and unused constants.
- Return error from deriveKeyFromSecret instead of ignoring it.
- Fix outdated doc comment on InitializeWithFiler.
- Use t.Setenv in tests instead of manual os.Setenv/Unsetenv.

* fix(s3): don't block startup on filer errors when KEK is configured

- When s3.sse.kek is set, a temporarily unreachable filer no longer
  prevents startup. The filer consistency check becomes best-effort
  with a warning.
- Same treatment for s3.sse.key: filer unreachable logs a warning
  instead of failing.
- Rewrite error messages to suggest migration instead of file deletion,
  avoiding the risk of orphaning encrypted data.

Finding 3 (restore auto-generation) intentionally skipped — auto-gen
was removed by design to avoid storing plaintext KEK on filer.

* fix(test): set WEED_S3_SSE_KEY in SSE integration test server startup

SSE-S3 no longer auto-generates a KEK, so integration tests must
provide one. Set WEED_S3_SSE_KEY=test-sse-s3-key in all weed mini
invocations in the test Makefile.
2026-04-03 13:01:21 -07:00
Chris LuandGitHub 2e98902f29 fix(s3): use URL-safe secret keys for dashboard users and service accounts (#8902)
* fix(s3): use URL-safe secret keys for admin dashboard users and service accounts

The dashboard's generateSecretKey() used base64.StdEncoding which produces
+, /, and = characters that break S3 signature authentication. Reuse the
IAM package's GenerateSecretAccessKey() which was already fixed in #7990.

Fixes #8898

* fix: handle error from GenerateSecretAccessKey instead of ignoring it
2026-04-03 11:20:28 -07:00
Jaehoon KimandGitHub d3cea714d0 fix(filer.backup): local sink readonly permission (#8907) 2026-04-03 05:36:56 -07:00
dependabot[bot]GitHubdependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
91087c0737 build(deps): bump github.com/go-jose/go-jose/v4 from 4.1.3 to 4.1.4 in /test/kafka (#8899)
build(deps): bump github.com/go-jose/go-jose/v4 in /test/kafka

Bumps [github.com/go-jose/go-jose/v4](https://github.com/go-jose/go-jose) from 4.1.3 to 4.1.4.
- [Release notes](https://github.com/go-jose/go-jose/releases)
- [Commits](https://github.com/go-jose/go-jose/compare/v4.1.3...v4.1.4)

---
updated-dependencies:
- dependency-name: github.com/go-jose/go-jose/v4
  dependency-version: 4.1.4
  dependency-type: indirect
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2026-04-03 00:13:07 -07:00
dependabot[bot]GitHubdependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
d2d21cd26b build(deps): bump github.com/go-jose/go-jose/v4 from 4.1.3 to 4.1.4 (#8900)
Bumps [github.com/go-jose/go-jose/v4](https://github.com/go-jose/go-jose) from 4.1.3 to 4.1.4.
- [Release notes](https://github.com/go-jose/go-jose/releases)
- [Commits](https://github.com/go-jose/go-jose/compare/v4.1.3...v4.1.4)

---
updated-dependencies:
- dependency-name: github.com/go-jose/go-jose/v4
  dependency-version: 4.1.4
  dependency-type: indirect
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
2026-04-03 00:13:01 -07:00
Chris Lu 0503311ded Merge branch 'master' of https://github.com/seaweedfs/seaweedfs 2026-04-02 18:39:28 -07:00
Chris Lu bb23939b36 fix(volume-rust): resolve gRPC bind address from hostname
SocketAddr::parse() only accepts numeric IPs, so binding the gRPC
server to "localhost:18833" panicked. Use tokio::net::lookup_host()
to resolve hostnames before passing to tonic's serve_with_shutdown.
2026-04-02 18:36:45 -07:00
Chris LuandGitHub 059bee683f feat(s3): add STS GetFederationToken support (#8891)
* feat(s3): add STS GetFederationToken support

Implement the AWS STS GetFederationToken API, which allows long-term IAM
users to obtain temporary credentials scoped down by an optional inline
session policy. This is useful for server-side applications that mint
per-user temporary credentials.

Key behaviors:
- Requires SigV4 authentication from a long-term IAM user
- Rejects calls from temporary credentials (session tokens)
- Name parameter (2-64 chars) identifies the federated user
- DurationSeconds supports 900-129600 (15 min to 36 hours, default 12h)
- Optional inline session policy for permission scoping
- Caller's attached policies are embedded in the JWT token
- Returns federated user ARN: arn:aws:sts::<account>:federated-user/<Name>

No performance impact on the S3 hot path — credential vending is a
separate control-plane operation, and all policy data is embedded in
the stateless JWT token.

* fix(s3): address GetFederationToken PR review feedback

- Fix Name validation: max 32 chars (not 64) per AWS spec, add regex
  validation for [\w+=,.@-]+ character whitelist
- Refactor parseDurationSeconds into parseDurationSecondsWithBounds to
  eliminate duplicated duration parsing logic
- Add sts:GetFederationToken permission check via VerifyActionPermission
  mirroring the AssumeRole authorization pattern
- Change GetPoliciesForUser to return ([]string, error) so callers fail
  closed on policy-resolution failures instead of silently returning nil
- Move temporary-credentials rejection before SigV4 verification for
  early rejection and proper test coverage
- Update tests: verify specific error message for temp cred rejection,
  add regex validation test cases (spaces, slashes rejected)

* refactor(s3): use sts.Action* constants instead of hard-coded strings

Replace hard-coded "sts:AssumeRole" and "sts:GetFederationToken" strings
in VerifyActionPermission calls with sts.ActionAssumeRole and
sts.ActionGetFederationToken package constants.

* fix(s3): pass through sts: prefix in action resolver and merge policies

Two fixes:

1. mapBaseActionToS3Format now passes through "sts:" prefix alongside
   "s3:" and "iam:", preventing sts:GetFederationToken from being
   rewritten to s3:sts:GetFederationToken in VerifyActionPermission.
   This also fixes the existing sts:AssumeRole permission checks.

2. GetFederationToken policy embedding now merges identity.PolicyNames
   (from SigV4 identity) with policies from the IAM manager (which may
   include group-attached policies), deduplicated via a map. Previously
   the IAM manager lookup was skipped when identity.PolicyNames was
   non-empty, causing group policies to be omitted from the token.

* test(s3): add integration tests for sts: action passthrough and policy merge

Action resolver tests:
- TestMapBaseActionToS3Format_ServicePrefixPassthrough: verifies s3:, iam:,
  and sts: prefixed actions pass through unchanged while coarse actions
  (Read, Write) are mapped to S3 format
- TestResolveS3Action_STSActionsPassthrough: verifies sts:AssumeRole,
  sts:GetFederationToken, sts:GetCallerIdentity pass through ResolveS3Action
  unchanged with both nil and real HTTP requests

Policy merge tests:
- TestGetFederationToken_GetPoliciesForUser: tests IAMManager.GetPoliciesForUser
  with no user store (error), missing user, user with policies, user without
- TestGetFederationToken_PolicyMergeAndDedup: tests that identity.PolicyNames
  and IAM-manager-resolved policies are merged and deduplicated (SharedPolicy
  appears in both sources, result has 3 unique policies)
- TestGetFederationToken_PolicyMergeNoManager: tests that when IAM manager is
  unavailable, identity.PolicyNames alone are embedded

* test(s3): add end-to-end integration tests for GetFederationToken

Add integration tests that call GetFederationToken using real AWS SigV4
signed HTTP requests against a running SeaweedFS instance, following the
existing pattern in test/s3/iam/s3_sts_assume_role_test.go.

Tests:
- TestSTSGetFederationTokenValidation: missing name, name too short/long,
  invalid characters, duration too short/long, malformed policy, anonymous
  rejection (7 subtests)
- TestSTSGetFederationTokenRejectTemporaryCredentials: obtains temp creds
  via AssumeRole then verifies GetFederationToken rejects them
- TestSTSGetFederationTokenSuccess: basic success, custom 1h duration,
  36h max duration with expiration time verification
- TestSTSGetFederationTokenWithSessionPolicy: creates a bucket, obtains
  federated creds with GetObject-only session policy, verifies GetObject
  succeeds and PutObject is denied using the AWS SDK S3 client
2026-04-02 17:37:05 -07:00
Chris Lu b8236a10d1 perf(docker): pre-build Rust binaries to avoid 5-hour QEMU emulation
Cross-compile Rust volume server natively for amd64/arm64 using musl
targets in a separate job, then inject pre-built binaries into the
Docker build. This replaces the ~5-hour QEMU-emulated cargo build
with ~15 minutes of native cross-compilation.

The Dockerfile falls back to building from source when no pre-built
binary is found, preserving local build compatibility.
2026-04-02 16:57:28 -07:00
Chris LuandGitHub a4b896a224 fix(s3): skip directories before marker in ListObjectVersions pagination (#8890)
* fix(s3): skip directories before marker in ListObjectVersions pagination

ListObjectVersions was re-traversing the entire directory tree from the
beginning on every paginated request, only skipping entries at the leaf
level. For buckets with millions of objects in deep hierarchies, this
caused exponentially slower responses as pagination progressed.

Two optimizations:
1. Use keyMarker to compute a startFrom position at each directory level,
   skipping directly to the relevant entry instead of scanning from the
   beginning (mirroring how ListObjects uses marker descent).
2. Skip recursing into subdirectories whose keys are entirely before the
   keyMarker.

Changes per-page cost from O(entries_before_marker) to O(tree_depth).

* test(s3): add integration test for deep-hierarchy version listing pagination

Adds TestVersioningPaginationDeepDirectoryHierarchy which creates objects
across 20 subdirectories at depth 6 (mimicking Veeam 365 backup layout)
and paginates through them with small maxKeys. Verifies correctness
(no duplicates, sorted order, all objects found) and checks that later
pages don't take dramatically longer than earlier ones — the symptom
of the pre-fix re-traversal bug. Also tests delimiter+pagination
interaction across subdirectories.

* test(s3): strengthen deep-hierarchy pagination assertions

- Replace timing warning (t.Logf) with a failing assertion (t.Errorf)
  so pagination regressions actually fail the test.
- Replace generic count/uniqueness/sort checks on CommonPrefixes with
  exact equality against the expected prefix slice, catching wrong-but-
  sorted results.

* test(s3): use allKeys for exact assertion in deep-hierarchy pagination test

Wire the allKeys slice (previously unused dead code) into the version
listing assertion, replacing generic count/uniqueness/sort checks with
an exact equality comparison against the keys that were created.
2026-04-02 15:59:52 -07:00
Chris LuandGitHub 7c59b639c9 STS: add GetCallerIdentity support (#8893)
* STS: add GetCallerIdentity support

Implement the AWS STS GetCallerIdentity action, which returns the
ARN, account ID, and user ID of the caller based on SigV4 authentication.
This is commonly used by AWS SDKs and CLI tools (e.g. `aws sts get-caller-identity`)
to verify credentials and determine the authenticated identity.

* test: remove trivial GetCallerIdentity tests

Remove the XML unmarshal test (we don't consume this response as input)
and the routing constant test (just asserts a literal equals itself).

* fix: route GetCallerIdentity through STS in UnifiedPostHandler and use stable UserId

- UnifiedPostHandler only dispatched actions starting with "AssumeRole" to STS,
  so GetCallerIdentity in a POST body would fall through to the IAM path and
  get AccessDenied for non-admin users. Add explicit check for GetCallerIdentity.
- Use identity.Name as UserId instead of credential.AccessKey, which is a
  transient value and incorrect for STS assumed-role callers.
2026-04-02 15:59:09 -07:00
Lars LehtonenandGitHub 772ad67f6b fix(weed/filer/redis): dropped error (#8895) 2026-04-02 15:39:04 -07:00
3a5016bcd7 fix(weed/worker/tasks/ec_balance): non-recursive reportProgress (#8892)
* fix(weed/worker/tasks/ec_balance): non-recursive reportProgress

* fix(ec_balance): call ReportProgressWithStage and include volumeID in log

The original fix replaced infinite recursion with a glog.Infof, but
skipped the framework progress callback. This adds the missing
ReportProgressWithStage call so the admin server receives EC balance
progress, and includes volumeID in the log for disambiguation.

---------

Co-authored-by: Chris Lu <chris.lu@gmail.com>
2026-04-02 15:32:57 -07:00
Chris Lu 0d8b024911 Merge branch 'codex/pr-8889'
# Conflicts:
#	test/fuse_integration/git_operations_test.go
2026-04-02 13:21:03 -07:00
Chris Lu 0f5e6b1f34 test: recover initial FUSE git clone on mount 2026-04-02 13:19:40 -07:00
Chris LuandGitHub 98714b9f70 fix(test): address flaky S3 distributed lock integration test (#8888)
* fix(test): address flaky S3 distributed lock integration test

Two root causes:

1. Lock ring convergence race: After waitForFilerCount(2) confirms the
   master sees both filers, there's a window where filer0's lock ring
   still only contains itself (master's LockRingUpdate broadcast is
   delayed by the 1s stabilization timer). During this window filer0
   considers itself primary for ALL keys, so both filers can
   independently grant the same lock.

   Fix: Add waitForLockRingConverged() that acquires the same lock
   through both filers and verifies mutual exclusion before proceeding.

2. Hash function mismatch: ownerForObjectLock used util.HashStringToLong
   (MD5 + modulo) to predict lock owners, but the production DLM uses
   CRC32 consistent hashing via HashRing. This meant the test could
   pick keys that route to the same filer, not exercising the
   cross-filer coordination it intended to test.

   Fix: Use lock_manager.NewHashRing + GetPrimary() to match production
   routing exactly.

* fix(test): verify lock denial reason in convergence check

Ensure the convergence check only returns true when the second lock
attempt is denied specifically because the lock is already owned,
avoiding false positives from transient errors.

* fix(test): check one key per primary filer in convergence wait

A single arbitrary key can false-pass: if its real primary is the filer
with the stale ring, mutual exclusion holds trivially because that filer
IS the correct primary. Generate one test key per distinct primary using
the same consistent-hash ring as production, so a stale ring on any
filer is caught deterministically.
2026-04-02 13:11:04 -07:00
Chris LuandGitHub 9552e80b58 filer.sync: show active chunk transfers when sync progress stalls (#8889)
* filer.sync: show active chunk transfers when sync progress stalls

When the sync watermark is not advancing, print each in-progress chunk
transfer with its file path, bytes received so far, and current status
(downloading, uploading, or waiting with backoff duration). This helps
diagnose which files are blocking progress during replication.

Closes #8542

* filer.sync: include last error in stall diagnostics

* filer.sync: fix data races in ChunkTransferStatus

Add sync.RWMutex to ChunkTransferStatus and lock around all field
mutations in fetchAndWrite. ActiveTransfers now returns value copies
under RLock so callers get immutable snapshots.
2026-04-02 13:08:24 -07:00
Chris Lu a974190cb1 Merge branch 'master' of https://github.com/seaweedfs/seaweedfs 2026-04-02 12:21:30 -07:00
Chris Lu e7fc243ee1 Fix flaky s3tables tests: allocate all ports atomically
The test port allocation had a TOCTOU race where GetFreePort() would
open a listener, grab the port number, then immediately close it.
When called repeatedly, the OS could recycle a just-released port,
causing two services (e.g. Filer and S3) to be assigned the same port.

Replace per-call GetFreePort() with batch AllocatePorts() that holds
all listeners open until every port is obtained, matching the pattern
already used in test/volume_server/framework/cluster.go.
2026-04-02 11:58:03 -07:00
Chris LuandGitHub ab4e52ae2f fix(s3): use recursive delete for .versions directory cleanup (#8887)
* fix(s3): use recursive delete for .versions directory cleanup

When only delete markers remain in a .versions directory,
updateLatestVersionAfterDeletion tried to delete it non-recursively,
which failed with "fail to delete non-empty folder" because the delete
marker entries were still present. Use recursive deletion so the
directory and its remaining delete marker entries are cleaned up together.

* fix(s3): guard .versions directory deletion against truncated listings

When the version listing is truncated (>1000 entries), content versions
may exist beyond the first page. Skip the recursive directory deletion
in this case to prevent data loss.

* fix(s3): preserve delete markers in .versions directory

Delete markers must be preserved per S3 semantics — they are only
removed by an explicit DELETE with versionId. The previous fix would
recursively delete the entire .versions directory (including delete
markers) when no content versions were found.

Now the logic distinguishes three cases:
1. Content versions exist → update latest version metadata
2. Only delete markers remain (or listing truncated) → keep directory
3. Truly empty → safe to delete directory (non-recursive)
2026-04-02 11:55:13 -07:00
Chris LuandGitHub 888c32cbde fix(admin): respect urlPrefix in S3 bucket and S3Tables navigation links (#8885)
* fix(admin): respect urlPrefix in S3 bucket and S3Tables navigation links (#8884)

Several admin UI templates used hardcoded URLs (templ.SafeURL) instead of
dash.PUrl(ctx, ...) for navigation links, causing 404 errors when the
admin is deployed with --urlPrefix.

Fixed in: s3_buckets.templ, s3tables_buckets.templ, s3tables_tables.templ

* fix(admin): URL-escape bucketName in S3Tables navigation links

Add url.PathEscape(bucketName) for consistency and correctness in
s3tables_tables.templ (back-to-namespaces link) and s3tables_buckets.templ
(namespace link), matching the escaping already used in the table details link.
2026-04-02 11:54:19 -07:00
Chris LuandGitHub efbed39e25 S3: map canned ACL to file permissions and add configurable default file mode (#8886)
* S3: map canned ACL to file permissions and add configurable default file mode

S3 uploads were hardcoded to 0660 regardless of ACL headers. Now the
X-Amz-Acl header maps to Unix file permissions per-object:
- public-read, authenticated-read, bucket-owner-read → 0644
- public-read-write → 0666
- private, bucket-owner-full-control → 0660

Also adds -defaultFileMode / -s3.defaultFileMode flag to set a
server-wide default when no ACL header is present.

Closes #8874

* Address review feedback for S3 file mode feature

- Extract hardcoded 0660 to defaultFileMode constant
- Change parseDefaultFileMode to return error instead of calling Fatalf
- Add -s3.defaultFileMode flag to filer.go and mini.go (was missing)
- Add doc comment to S3Options about updating all four flag sites
- Add TestResolveFileMode with 10 test cases covering ACL mapping,
  server default, and priority ordering
2026-04-02 11:51:54 -07:00
Chris Lu e93f4e3f39 Merge branch 'master' of https://github.com/seaweedfs/seaweedfs 2026-04-02 11:50:06 -07:00
Chris Lu 647b46bd8a Fix flaky FUSE integration tests
- concurrent_operations_test: Add retry loop for transient I/O errors
  on file close during ConcurrentDirectoryOperations
- git_operations_test: Wait for pushed objects to become visible through
  FUSE mount before cloning in Phase 3
2026-04-02 11:49:13 -07:00
Chris Lu 24805ff478 fix(docker): add libgcc to Alpine runtime for Rust volume server (#8883)
The Rust weed-volume binary requires libgcc_s.so.1 for stack unwinding
(_Unwind_* symbols). Without it, the binary fails to load in the Alpine
container with "Error loading shared library libgcc_s.so.1".
2026-04-02 11:33:54 -07:00
338 changed files with 5772 additions and 46255 deletions
+1
View File
@@ -0,0 +1 @@
{"sessionId":"d6574c47-eafc-4a94-9dce-f9ffea22b53c","pid":10111,"acquiredAt":1775248373916}
+104 -5
View File
@@ -39,7 +39,80 @@ concurrency:
cancel-in-progress: false
jobs:
# ── Pre-build Rust volume server binaries natively ──────────────────
# Cross-compiles for amd64 and arm64 without QEMU, turning a 5-hour
# emulated cargo build into ~15 minutes of native compilation.
build-rust-binaries:
runs-on: ubuntu-22.04
strategy:
matrix:
include:
- target: x86_64-unknown-linux-musl
arch: amd64
- target: aarch64-unknown-linux-musl
arch: arm64
cross: true
steps:
- name: Checkout
uses: actions/checkout@v6
- name: Install protobuf compiler
run: sudo apt-get update && sudo apt-get install -y protobuf-compiler
- name: Install Rust toolchain
uses: dtolnay/rust-toolchain@stable
with:
targets: ${{ matrix.target }}
- name: Install musl tools (amd64)
if: ${{ !matrix.cross }}
run: sudo apt-get install -y musl-tools
- name: Install cross-compilation tools (arm64)
if: matrix.cross
run: |
sudo apt-get install -y gcc-aarch64-linux-gnu
echo "CARGO_TARGET_AARCH64_UNKNOWN_LINUX_MUSL_LINKER=aarch64-linux-gnu-gcc" >> "$GITHUB_ENV"
- name: Cache cargo registry and target
uses: actions/cache@v5
with:
path: |
~/.cargo/registry
~/.cargo/git
seaweed-volume/target
key: rust-docker-${{ matrix.target }}-${{ hashFiles('seaweed-volume/Cargo.lock') }}
restore-keys: |
rust-docker-${{ matrix.target }}-
- name: Build large-disk variant
env:
SEAWEEDFS_COMMIT: ${{ github.sha }}
run: |
cd seaweed-volume
cargo build --release --target ${{ matrix.target }}
cp target/${{ matrix.target }}/release/weed-volume ../weed-volume-large-disk-${{ matrix.arch }}
- name: Build normal variant
env:
SEAWEEDFS_COMMIT: ${{ github.sha }}
run: |
cd seaweed-volume
cargo build --release --target ${{ matrix.target }} --no-default-features
cp target/${{ matrix.target }}/release/weed-volume ../weed-volume-normal-${{ matrix.arch }}
- name: Upload artifacts
uses: actions/upload-artifact@v4
with:
name: rust-volume-${{ matrix.arch }}
path: |
weed-volume-large-disk-${{ matrix.arch }}
weed-volume-normal-${{ matrix.arch }}
# ── Build Docker containers ─────────────────────────────────────────
build:
needs: [build-rust-binaries]
runs-on: ubuntu-latest
strategy:
# Build sequentially to avoid rate limits
@@ -52,20 +125,23 @@ jobs:
dockerfile: ./docker/Dockerfile.go_build
build_args: ""
tag_suffix: ""
# Large disk - multi-arch
rust_variant: normal
# Large disk - multi-arch
- variant: large_disk
platforms: linux/amd64,linux/arm64,linux/arm/v7,linux/386
dockerfile: ./docker/Dockerfile.go_build
build_args: TAGS=5BytesOffset
tag_suffix: _large_disk
rust_variant: large-disk
# Full tags - multi-arch
- variant: full
platforms: linux/amd64,linux/arm64
dockerfile: ./docker/Dockerfile.go_build
build_args: TAGS=elastic,gocdk,rclone,sqlite,tarantool,tikv,ydb
tag_suffix: _full
rust_variant: normal
# Large disk + full tags - multi-arch
- variant: large_disk_full
@@ -73,19 +149,42 @@ jobs:
dockerfile: ./docker/Dockerfile.go_build
build_args: TAGS=5BytesOffset,elastic,gocdk,rclone,sqlite,tarantool,tikv,ydb
tag_suffix: _large_disk_full
rust_variant: large-disk
# RocksDB large disk - amd64 only
- variant: rocksdb
platforms: linux/amd64
dockerfile: ./docker/Dockerfile.rocksdb_large
build_args: ""
tag_suffix: _large_disk_rocksdb
rust_variant: large-disk
steps:
- name: Checkout
if: github.event_name != 'workflow_dispatch' || github.event.inputs.variant == 'all' || github.event.inputs.variant == matrix.variant
uses: actions/checkout@v6
- name: Download pre-built Rust binaries
if: github.event_name != 'workflow_dispatch' || github.event.inputs.variant == 'all' || github.event.inputs.variant == matrix.variant
uses: actions/download-artifact@v4
with:
pattern: rust-volume-*
merge-multiple: true
path: ./rust-bins
- name: Place Rust binaries in Docker context
if: github.event_name != 'workflow_dispatch' || github.event.inputs.variant == 'all' || github.event.inputs.variant == matrix.variant
run: |
mkdir -p docker/weed-volume-prebuilt
for arch in amd64 arm64; do
src="./rust-bins/weed-volume-${{ matrix.rust_variant }}-${arch}"
if [ -f "$src" ]; then
cp "$src" "docker/weed-volume-prebuilt/weed-volume-${arch}"
echo "Placed pre-built Rust binary for ${arch}"
fi
done
ls -la docker/weed-volume-prebuilt/
- name: Free Disk Space
if: github.event_name != 'workflow_dispatch' || github.event.inputs.variant == 'all' || github.event.inputs.variant == matrix.variant
run: |
@@ -93,7 +93,7 @@ jobs:
- name: Build Go weed binary
run: |
cd weed
go build -o weed .
go build -tags 5BytesOffset -o weed .
chmod +x weed
./weed version
@@ -169,7 +169,7 @@ jobs:
- name: Build Go weed binary
run: |
cd weed
go build -o weed .
go build -tags 5BytesOffset -o weed .
chmod +x weed
./weed version
+5
View File
@@ -0,0 +1,5 @@
{
"setup": [],
"teardown": [],
"run": []
}
+11 -5
View File
@@ -16,15 +16,21 @@ RUN cd /go/src/github.com/seaweedfs/seaweedfs/weed \
&& export LDFLAGS="-X github.com/seaweedfs/seaweedfs/weed/util/version.COMMIT=$(git rev-parse --short HEAD)" \
&& CGO_ENABLED=0 go install -tags "$TAGS" -ldflags "-extldflags -static ${LDFLAGS}"
# Rust volume server builder. Alpine packages avoid depending on the
# upstream rust:alpine manifest list, which no longer includes linux/386.
# Rust volume server: use pre-built binary from CI when available (placed in
# weed-volume-prebuilt/ by the build-rust-binaries job), otherwise compile
# from source. Pre-building avoids a multi-hour QEMU-emulated cargo build
# for non-native architectures.
FROM alpine:3.23 as rust_builder
ARG TARGETARCH
ARG TAGS
COPY weed-volume-prebuilt/ /prebuilt/
COPY --from=builder /go/src/github.com/seaweedfs/seaweedfs/seaweed-volume /build/seaweed-volume
COPY --from=builder /go/src/github.com/seaweedfs/seaweedfs/weed /build/weed
WORKDIR /build/seaweed-volume
ARG TAGS
RUN if [ "$TARGETARCH" = "amd64" ] || [ "$TARGETARCH" = "arm64" ]; then \
RUN if [ -f "/prebuilt/weed-volume-${TARGETARCH}" ]; then \
echo "Using pre-built Rust binary for ${TARGETARCH}" && \
cp "/prebuilt/weed-volume-${TARGETARCH}" /weed-volume; \
elif [ "$TARGETARCH" = "amd64" ] || [ "$TARGETARCH" = "arm64" ]; then \
apk add --no-cache musl-dev openssl-dev protobuf-dev git rust cargo; \
if [ "$TAGS" = "5BytesOffset" ]; then \
cargo build --release; \
@@ -51,7 +57,7 @@ COPY --from=builder /go/src/github.com/seaweedfs/seaweedfs/docker/entrypoint.sh
# Install dependencies and create non-root user
RUN apk upgrade --no-cache zlib && \
apk add --no-cache fuse curl su-exec && \
apk add --no-cache fuse curl su-exec libgcc && \
addgroup -g 1000 seaweed && \
adduser -D -u 1000 -G seaweed seaweed
@@ -0,0 +1,53 @@
{
"identities": [
{
"name": "admin",
"credentials": [
{
"accessKey": "AKIAIOSFODNN7EXAMPLE",
"secretKey": "wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY"
}
],
"actions": [
"Admin",
"Read",
"List",
"Tagging",
"Write"
]
},
{
"name": "steward",
"credentials": [
{
"accessKey": "steward-key",
"secretKey": "steward-secret"
}
],
"actions": [
"Read",
"List",
"Write"
]
},
{
"name": "le001",
"credentials": [
{
"accessKey": "le001-key",
"secretKey": "le001-secret"
}
],
"actions": [
"Read",
"List"
]
},
{
"name": "anonymous",
"actions": [
"Read"
]
}
]
}
+1 -1
View File
@@ -371,7 +371,7 @@ require (
github.com/geoffgarside/ber v1.2.0 // indirect
github.com/go-chi/chi/v5 v5.2.5 // indirect
github.com/go-darwin/apfs v0.0.0-20211011131704-f84b94dbf348 // indirect
github.com/go-jose/go-jose/v4 v4.1.3 // indirect
github.com/go-jose/go-jose/v4 v4.1.4 // indirect
github.com/go-logr/logr v1.4.3 // indirect
github.com/go-logr/stdr v1.2.2 // indirect
github.com/go-ole/go-ole v1.3.0 // indirect
+2 -2
View File
@@ -1076,8 +1076,8 @@ github.com/go-git/go-billy/v5 v5.6.2/go.mod h1:rcFC2rAsp/erv7CMz9GczHcuD0D32fWzH
github.com/go-gl/glfw v0.0.0-20190409004039-e6da0acd62b1/go.mod h1:vR7hzQXu2zJy9AVAgeJqvqgH9Q5CA+iKCZ2gyEVpxRU=
github.com/go-gl/glfw/v3.3/glfw v0.0.0-20191125211704-12ad95a8df72/go.mod h1:tQ2UAYgL5IevRw8kRxooKSPJfGvJ9fJQFa0TUsXzTg8=
github.com/go-gl/glfw/v3.3/glfw v0.0.0-20200222043503-6f7a984d4dc4/go.mod h1:tQ2UAYgL5IevRw8kRxooKSPJfGvJ9fJQFa0TUsXzTg8=
github.com/go-jose/go-jose/v4 v4.1.3 h1:CVLmWDhDVRa6Mi/IgCgaopNosCaHz7zrMeF9MlZRkrs=
github.com/go-jose/go-jose/v4 v4.1.3/go.mod h1:x4oUasVrzR7071A4TnHLGSPpNOm2a21K9Kf04k1rs08=
github.com/go-jose/go-jose/v4 v4.1.4 h1:moDMcTHmvE6Groj34emNPLs/qtYXRVcd6S7NHbHz3kA=
github.com/go-jose/go-jose/v4 v4.1.4/go.mod h1:x4oUasVrzR7071A4TnHLGSPpNOm2a21K9Kf04k1rs08=
github.com/go-kit/kit v0.8.0/go.mod h1:xBxKIO96dXMWWy0MnWVtmwkA9/13aqxPnvrjFYMA2as=
github.com/go-kit/kit v0.9.0/go.mod h1:xBxKIO96dXMWWy0MnWVtmwkA9/13aqxPnvrjFYMA2as=
github.com/go-kit/log v0.1.0/go.mod h1:zbhenjAZHb184qTLMA9ZjW7ThYL0H2mk7Q6pNt4vbaY=
+11
View File
@@ -2561,6 +2561,15 @@ version = "0.2.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "7c87def4c32ab89d880effc9e097653c8da5d6ef28e6b539d313baaacfbafcbe"
[[package]]
name = "openssl-src"
version = "300.5.5+3.5.5"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "3f1787d533e03597a7934fd0a765f0d28e94ecc5fb7789f8053b1e699a56f709"
dependencies = [
"cc",
]
[[package]]
name = "openssl-sys"
version = "0.9.111"
@@ -2569,6 +2578,7 @@ checksum = "82cab2d520aa75e3c58898289429321eb788c3106963d0dc886ec7a5f4adc321"
dependencies = [
"cc",
"libc",
"openssl-src",
"pkg-config",
"vcpkg",
]
@@ -4654,6 +4664,7 @@ dependencies = [
"memmap2",
"mime_guess",
"multer",
"openssl",
"parking_lot 0.12.5",
"pprof",
"prometheus",
+1 -1
View File
@@ -20,7 +20,7 @@ default = ["5bytes"]
[dependencies]
# Async runtime
tokio = { version = "1", features = ["full"] }
tokio-stream = "0.1"
tokio-stream = { version = "0.1", features = ["net"] }
tokio-io-timeout = "1"
# gRPC + protobuf
+68 -36
View File
@@ -600,31 +600,35 @@ async fn run(
})
};
// Bind the gRPC listener before spawning to propagate bind errors at startup.
let grpc_listener = tokio::net::TcpListener::bind(&grpc_addr)
.await
.unwrap_or_else(|e| panic!("Failed to bind gRPC to {}: {}", grpc_addr, e));
let grpc_local_addr = grpc_listener
.local_addr()
.unwrap_or_else(|e| panic!("Failed to get gRPC local addr: {}", e));
let grpc_handle = {
let grpc_state = state.clone();
let grpc_addr = grpc_addr.clone();
let grpc_tls_acceptor = grpc_tls_acceptor.clone();
let mut shutdown_rx = shutdown_tx.subscribe();
let shutdown_tx_grpc = shutdown_tx.clone();
tokio::spawn(async move {
let addr = grpc_addr.parse().expect("Invalid gRPC address");
let grpc_service = VolumeGrpcService {
state: grpc_state.clone(),
};
if let Some(tls_acceptor) = grpc_tls_acceptor {
let listener = tokio::net::TcpListener::bind(&grpc_addr)
.await
.unwrap_or_else(|e| panic!("Failed to bind gRPC to {}: {}", grpc_addr, e));
let incoming = grpc_tls_incoming(listener, tls_acceptor);
let reflection_v1 = tonic_reflection::server::Builder::configure()
.register_encoded_file_descriptor_set(seaweed_volume::pb::FILE_DESCRIPTOR_SET)
.build_v1()
.expect("Failed to build gRPC reflection v1 service");
let reflection_v1alpha = tonic_reflection::server::Builder::configure()
.register_encoded_file_descriptor_set(seaweed_volume::pb::FILE_DESCRIPTOR_SET)
.build_v1alpha()
.expect("Failed to build gRPC reflection v1alpha service");
info!("gRPC server listening on {} (TLS enabled)", addr);
if let Err(e) = build_grpc_server_builder()
let reflection_v1 = tonic_reflection::server::Builder::configure()
.register_encoded_file_descriptor_set(seaweed_volume::pb::FILE_DESCRIPTOR_SET)
.build_v1()
.expect("Failed to build gRPC reflection v1 service");
let reflection_v1alpha = tonic_reflection::server::Builder::configure()
.register_encoded_file_descriptor_set(seaweed_volume::pb::FILE_DESCRIPTOR_SET)
.build_v1alpha()
.expect("Failed to build gRPC reflection v1alpha service");
let result = if let Some(tls_acceptor) = grpc_tls_acceptor {
let incoming = grpc_tls_incoming(grpc_listener, tls_acceptor);
info!("gRPC server listening on {} (TLS enabled)", grpc_local_addr);
build_grpc_server_builder()
.layer(GrpcRequestIdLayer)
.add_service(reflection_v1)
.add_service(reflection_v1alpha)
@@ -633,32 +637,25 @@ async fn run(
let _ = shutdown_rx.recv().await;
})
.await
{
error!("gRPC server error: {}", e);
}
} else {
let reflection_v1 = tonic_reflection::server::Builder::configure()
.register_encoded_file_descriptor_set(seaweed_volume::pb::FILE_DESCRIPTOR_SET)
.build_v1()
.expect("Failed to build gRPC reflection v1 service");
let reflection_v1alpha = tonic_reflection::server::Builder::configure()
.register_encoded_file_descriptor_set(seaweed_volume::pb::FILE_DESCRIPTOR_SET)
.build_v1alpha()
.expect("Failed to build gRPC reflection v1alpha service");
info!("gRPC server listening on {}", addr);
if let Err(e) = build_grpc_server_builder()
let incoming =
tokio_stream::wrappers::TcpListenerStream::new(grpc_listener);
info!("gRPC server listening on {}", grpc_local_addr);
build_grpc_server_builder()
.layer(GrpcRequestIdLayer)
.add_service(reflection_v1)
.add_service(reflection_v1alpha)
.add_service(build_volume_grpc_service(grpc_service))
.serve_with_shutdown(addr, async move {
.serve_with_incoming_shutdown(incoming, async move {
let _ = shutdown_rx.recv().await;
})
.await
{
error!("gRPC server error: {}", e);
}
};
if let Err(ref e) = result {
error!("gRPC server error: {}", e);
let _ = shutdown_tx_grpc.send(());
}
result
})
};
@@ -767,9 +764,40 @@ async fn run(
}))
};
// Wait for all servers
// Wait for servers. Use select! with &mut so the losing handle is not
// dropped, then await it explicitly afterward.
let mut server_err: Option<String> = None;
let mut http_handle = http_handle;
let mut grpc_handle = grpc_handle;
let grpc_finished_first = tokio::select! {
_ = &mut http_handle => false,
_ = &mut grpc_handle => true,
};
// Inspect the gRPC result (already resolved if it finished first,
// otherwise await it now).
let grpc_result = if grpc_finished_first {
grpc_handle.await
} else {
// HTTP finished first; gRPC is still running. Await it.
grpc_handle.await
};
match grpc_result {
Ok(Ok(())) => {}
Ok(Err(e)) => {
let msg = format!("gRPC server exited with error: {}", e);
error!("{}", msg);
server_err = Some(msg);
// serve error already sent shutdown inside the task
}
Err(e) => {
let msg = format!("gRPC task panicked: {}", e);
error!("{}", msg);
server_err = Some(msg);
let _ = shutdown_tx.send(());
}
}
// Ensure the HTTP handle completes too.
let _ = http_handle.await;
let _ = grpc_handle.await;
if let Some(h) = public_handle {
let _ = h.await;
}
@@ -794,6 +822,10 @@ async fn run(
cpu_profile.finish().map_err(std::io::Error::other)?;
}
if let Some(err_msg) = server_err {
return Err(std::io::Error::other(err_msg).into());
}
info!("Volume server stopped.");
Ok(())
}
+33 -3
View File
@@ -1867,6 +1867,7 @@ impl VolumeServer for VolumeGrpcService {
{
let needle_header = resp.needle_header;
let mut needle_body = resp.needle_body;
let resp_version = resp.version;
if needle_header.is_empty() {
continue;
@@ -1891,8 +1892,36 @@ impl VolumeServer for VolumeGrpcService {
// Parse needle from header + body
let mut n = Needle::default();
n.read_header(&needle_header);
n.read_body_v2(&needle_body)
.map_err(|e| Status::internal(format!("parse needle body: {}", e)))?;
if n.size.0 < 0 {
return Err(Status::invalid_argument(format!(
"unexpected negative needle size {} for needle {}",
n.size.0, n.id.0
)));
} else if n.size.0 > 0 {
// Normal needle: parse the body fields (DataSize, Data, flags, etc.)
n.read_body_v2(&needle_body)
.map_err(|e| Status::internal(format!("parse needle body: {}", e)))?;
} else {
// Delete tombstone (size == 0): body is checksum + timestamp
// (V3) or checksum only (V2) + padding. Validate minimum
// footer length for the protocol version.
use crate::storage::types::{
NEEDLE_CHECKSUM_SIZE, TIMESTAMP_SIZE, VERSION_3, Version,
};
let version = Version(resp_version as u8);
let min_footer = if version >= VERSION_3 {
NEEDLE_CHECKSUM_SIZE + TIMESTAMP_SIZE
} else {
NEEDLE_CHECKSUM_SIZE
};
if needle_body.len() < min_footer {
return Err(Status::invalid_argument(format!(
"tombstone needle {} body too short: got {} bytes, need >= {} for version {}",
n.id.0, needle_body.len(), min_footer, resp_version
)));
}
}
// Write needle to local volume
let mut store = state.store.write().unwrap();
@@ -4047,11 +4076,12 @@ fn find_last_append_at_ns(idx_path: &str, dat_path: &str, version: u32) -> Optio
let mut header = [0u8; 16];
dat_file.read_exact(&mut header).ok()?;
let needle_size = i32::from_be_bytes([header[12], header[13], header[14], header[15]]);
if needle_size <= 0 {
if needle_size < 0 {
return None;
}
// Seek to tail: offset + 16 (header) + size -> checksum (4) + timestamp (8)
// For delete needles (size == 0), the tail is right after the header.
let tail_offset = actual_offset as u64 + 16 + needle_size as u64;
dat_file.seek(SeekFrom::Start(tail_offset)).ok()?;
@@ -249,11 +249,18 @@ func testConcurrentDirectoryOperations(t *testing.T, framework *FuseTestFramewor
return
}
// Create file in subdirectory
// Create file in subdirectory with retry for transient FUSE errors
testFile := filepath.Join(subDir, "test.txt")
content := []byte(fmt.Sprintf("Worker %d, Subdir %d", workerID, i))
if err := os.WriteFile(testFile, content, 0644); err != nil {
addError(fmt.Errorf("worker %d file %d: %v", workerID, i, err))
var writeErr error
for attempt := 0; attempt < 3; attempt++ {
if writeErr = os.WriteFile(testFile, content, 0644); writeErr == nil {
break
}
time.Sleep(100 * time.Millisecond)
}
if writeErr != nil {
addError(fmt.Errorf("worker %d file %d: %v", workerID, i, writeErr))
return
}
}
+81 -3
View File
@@ -85,9 +85,19 @@ func testGitCloneAndPull(t *testing.T, mountPoint, localDir string) {
branch := gitOutput(t, localClone, "rev-parse", "--abbrev-ref", "HEAD")
gitRun(t, localClone, "push", "origin", branch)
// The bare repo lives on the FUSE mount and can briefly disappear after
// the push completes, and pushed pack objects may not be immediately
// consistent on the FUSE layer. Give the mount a chance to settle, then
// recover from the local clone if the remote is still missing.
if !waitForBareRepoEventually(t, bareRepo, 10*time.Second) {
t.Logf("bare repo %s did not stabilise after push; forcing recovery before clone", bareRepo)
}
refreshDirEntry(t, bareRepo)
time.Sleep(1 * time.Second)
// ---- Phase 3: Clone from mount bare repo into on-mount working dir ----
t.Log("Phase 3: clone from mount bare repo to on-mount working dir")
gitRun(t, "", "clone", bareRepo, mountClone)
ensureMountCloneFromBareWithRecovery(t, bareRepo, localClone, mountClone)
assertFileContains(t, filepath.Join(mountClone, "README.md"), "# Updated")
assertFileContains(t, filepath.Join(mountClone, "src/main.go"), "v2")
@@ -290,7 +300,7 @@ func waitForBareRepoEventually(t *testing.T, bareRepo string, timeout time.Durat
t.Helper()
deadline := time.Now().Add(timeout)
for time.Now().Before(deadline) {
if isBareRepo(bareRepo) {
if isBareRepoAccessible(bareRepo) {
return true
}
refreshDirEntry(t, bareRepo)
@@ -312,6 +322,14 @@ func isBareRepo(bareRepo string) bool {
return true
}
func isBareRepoAccessible(bareRepo string) bool {
if !isBareRepo(bareRepo) {
return false
}
out, err := tryGitCommand("", "--git-dir="+bareRepo, "rev-parse", "--is-bare-repository")
return err == nil && out == "true"
}
func ensureMountClone(t *testing.T, bareRepo, mountClone string) {
t.Helper()
require.NoError(t, tryEnsureMountClone(bareRepo, mountClone))
@@ -320,7 +338,7 @@ func ensureMountClone(t *testing.T, bareRepo, mountClone string) {
// tryEnsureBareRepo verifies the bare repo on the FUSE mount exists.
// If it has vanished, it re-creates it from the local clone.
func tryEnsureBareRepo(bareRepo, localClone string) error {
if _, err := os.Stat(filepath.Join(bareRepo, "HEAD")); err == nil {
if isBareRepoAccessible(bareRepo) {
return nil
}
branch, err := tryGitCommand(localClone, "rev-parse", "--abbrev-ref", "HEAD")
@@ -342,6 +360,36 @@ func tryEnsureBareRepo(bareRepo, localClone string) error {
return nil
}
func ensureMountCloneFromBareWithRecovery(t *testing.T, bareRepo, localClone, mountClone string) {
t.Helper()
const maxAttempts = 3
var lastErr error
for attempt := 1; attempt <= maxAttempts; attempt++ {
if lastErr = tryEnsureMountCloneFromBare(bareRepo, localClone, mountClone); lastErr == nil {
return
}
if attempt == maxAttempts {
require.NoError(t, lastErr, "git clone %s %s failed after %d recovery attempts", bareRepo, mountClone, maxAttempts)
}
t.Logf("clone recovery attempt %d: %v — removing clone for re-create", attempt, lastErr)
os.RemoveAll(mountClone)
time.Sleep(2 * time.Second)
}
}
func tryEnsureMountCloneFromBare(bareRepo, localClone, mountClone string) error {
if err := tryEnsureBareRepo(bareRepo, localClone); err != nil {
return fmt.Errorf("ensure bare repo: %w", err)
}
if err := tryEnsureMountClone(bareRepo, mountClone); err != nil {
return fmt.Errorf("ensure mount clone: %w", err)
}
if _, err := tryGitCommand(mountClone, "rev-parse", "HEAD"); err != nil {
return fmt.Errorf("verify mount clone: %w", err)
}
return nil
}
// tryEnsureMountClone is like ensureMountClone but returns an error instead
// of failing the test, for use in recovery loops.
func tryEnsureMountClone(bareRepo, mountClone string) error {
@@ -550,3 +598,33 @@ func TestTryEnsureBareRepoPreservesCurrentBranch(t *testing.T) {
assert.Equal(t, branch, restoredHead, "clone from recovered bare repo should check out the current branch")
assertFileContains(t, filepath.Join(restoredClone, "README.md"), "hello recovery")
}
func TestEnsureMountCloneFromBareWithRecoveryRecreatesMissingBareRepo(t *testing.T) {
tempDir, err := os.MkdirTemp("", "git_mount_clone_recovery_")
require.NoError(t, err)
defer os.RemoveAll(tempDir)
bareRepo := filepath.Join(tempDir, "repo.git")
localClone := filepath.Join(tempDir, "clone")
mountClone := filepath.Join(tempDir, "mount-clone")
gitRun(t, "", "init", "--bare", bareRepo)
gitRun(t, "", "clone", bareRepo, localClone)
gitRun(t, localClone, "config", "user.email", "test@seaweedfs.test")
gitRun(t, localClone, "config", "user.name", "Test")
writeFile(t, localClone, "README.md", "hello clone recovery\n")
gitRun(t, localClone, "add", "README.md")
gitRun(t, localClone, "commit", "-m", "initial commit")
branch := gitOutput(t, localClone, "rev-parse", "--abbrev-ref", "HEAD")
gitRun(t, localClone, "push", "origin", branch)
require.NoError(t, os.RemoveAll(bareRepo))
ensureMountCloneFromBareWithRecovery(t, bareRepo, localClone, mountClone)
head := gitOutput(t, mountClone, "rev-parse", "--abbrev-ref", "HEAD")
assert.Equal(t, branch, head, "recovered clone should stay on the pushed branch")
assertFileContains(t, filepath.Join(mountClone, "README.md"), "hello clone recovery")
}
+13 -13
View File
@@ -43,25 +43,25 @@ require (
github.com/andybalholm/cascadia v1.3.3 // indirect
github.com/appscode/go-querystring v0.0.0-20170504095604-0126cfb3f1dc // indirect
github.com/aws/aws-sdk-go v1.55.8 // indirect
github.com/aws/aws-sdk-go-v2 v1.41.4 // indirect
github.com/aws/aws-sdk-go-v2 v1.41.5 // indirect
github.com/aws/aws-sdk-go-v2/aws/protocol/eventstream v1.7.4 // indirect
github.com/aws/aws-sdk-go-v2/config v1.32.9 // indirect
github.com/aws/aws-sdk-go-v2/credentials v1.19.12 // indirect
github.com/aws/aws-sdk-go-v2/feature/ec2/imds v1.18.20 // indirect
github.com/aws/aws-sdk-go-v2/config v1.32.13 // indirect
github.com/aws/aws-sdk-go-v2/credentials v1.19.13 // indirect
github.com/aws/aws-sdk-go-v2/feature/ec2/imds v1.18.21 // indirect
github.com/aws/aws-sdk-go-v2/feature/s3/manager v1.20.12 // indirect
github.com/aws/aws-sdk-go-v2/internal/configsources v1.4.20 // indirect
github.com/aws/aws-sdk-go-v2/internal/endpoints/v2 v2.7.20 // indirect
github.com/aws/aws-sdk-go-v2/internal/ini v1.8.4 // indirect
github.com/aws/aws-sdk-go-v2/internal/configsources v1.4.21 // indirect
github.com/aws/aws-sdk-go-v2/internal/endpoints/v2 v2.7.21 // indirect
github.com/aws/aws-sdk-go-v2/internal/ini v1.8.6 // indirect
github.com/aws/aws-sdk-go-v2/internal/v4a v1.4.17 // indirect
github.com/aws/aws-sdk-go-v2/service/internal/accept-encoding v1.13.7 // indirect
github.com/aws/aws-sdk-go-v2/service/internal/checksum v1.9.8 // indirect
github.com/aws/aws-sdk-go-v2/service/internal/presigned-url v1.13.20 // indirect
github.com/aws/aws-sdk-go-v2/service/internal/presigned-url v1.13.21 // indirect
github.com/aws/aws-sdk-go-v2/service/internal/s3shared v1.19.17 // indirect
github.com/aws/aws-sdk-go-v2/service/s3 v1.96.0 // indirect
github.com/aws/aws-sdk-go-v2/service/signin v1.0.8 // indirect
github.com/aws/aws-sdk-go-v2/service/sso v1.30.13 // indirect
github.com/aws/aws-sdk-go-v2/service/ssooidc v1.35.17 // indirect
github.com/aws/aws-sdk-go-v2/service/sts v1.41.9 // indirect
github.com/aws/aws-sdk-go-v2/service/signin v1.0.9 // indirect
github.com/aws/aws-sdk-go-v2/service/sso v1.30.14 // indirect
github.com/aws/aws-sdk-go-v2/service/ssooidc v1.35.18 // indirect
github.com/aws/aws-sdk-go-v2/service/sts v1.41.10 // indirect
github.com/aws/smithy-go v1.24.2 // indirect
github.com/bahlo/generic-list-go v0.2.0 // indirect
github.com/beorn7/perks v1.0.1 // indirect
@@ -104,7 +104,7 @@ require (
github.com/go-chi/chi/v5 v5.2.5 // indirect
github.com/go-darwin/apfs v0.0.0-20211011131704-f84b94dbf348 // indirect
github.com/go-git/go-billy/v5 v5.6.2 // indirect
github.com/go-jose/go-jose/v4 v4.1.3 // indirect
github.com/go-jose/go-jose/v4 v4.1.4 // indirect
github.com/go-logr/logr v1.4.3 // indirect
github.com/go-logr/stdr v1.2.2 // indirect
github.com/go-ole/go-ole v1.3.0 // indirect
+28 -28
View File
@@ -112,44 +112,44 @@ github.com/appscode/go-querystring v0.0.0-20170504095604-0126cfb3f1dc h1:LoL75er
github.com/appscode/go-querystring v0.0.0-20170504095604-0126cfb3f1dc/go.mod h1:w648aMHEgFYS6xb0KVMMtZ2uMeemhiKCuD2vj6gY52A=
github.com/aws/aws-sdk-go v1.55.8 h1:JRmEUbU52aJQZ2AjX4q4Wu7t4uZjOu71uyNmaWlUkJQ=
github.com/aws/aws-sdk-go v1.55.8/go.mod h1:ZkViS9AqA6otK+JBBNH2++sx1sgxrPKcSzPPvQkUtXk=
github.com/aws/aws-sdk-go-v2 v1.41.4 h1:10f50G7WyU02T56ox1wWXq+zTX9I1zxG46HYuG1hH/k=
github.com/aws/aws-sdk-go-v2 v1.41.4/go.mod h1:mwsPRE8ceUUpiTgF7QmQIJ7lgsKUPQOUl3o72QBrE1o=
github.com/aws/aws-sdk-go-v2 v1.41.5 h1:dj5kopbwUsVUVFgO4Fi5BIT3t4WyqIDjGKCangnV/yY=
github.com/aws/aws-sdk-go-v2 v1.41.5/go.mod h1:mwsPRE8ceUUpiTgF7QmQIJ7lgsKUPQOUl3o72QBrE1o=
github.com/aws/aws-sdk-go-v2/aws/protocol/eventstream v1.7.4 h1:489krEF9xIGkOaaX3CE/Be2uWjiXrkCH6gUX+bZA/BU=
github.com/aws/aws-sdk-go-v2/aws/protocol/eventstream v1.7.4/go.mod h1:IOAPF6oT9KCsceNTvvYMNHy0+kMF8akOjeDvPENWxp4=
github.com/aws/aws-sdk-go-v2/config v1.32.9 h1:ktda/mtAydeObvJXlHzyGpK1xcsLaP16zfUPDGoW90A=
github.com/aws/aws-sdk-go-v2/config v1.32.9/go.mod h1:U+fCQ+9QKsLW786BCfEjYRj34VVTbPdsLP3CHSYXMOI=
github.com/aws/aws-sdk-go-v2/credentials v1.19.12 h1:oqtA6v+y5fZg//tcTWahyN9PEn5eDU/Wpvc2+kJ4aY8=
github.com/aws/aws-sdk-go-v2/credentials v1.19.12/go.mod h1:U3R1RtSHx6NB0DvEQFGyf/0sbrpJrluENHdPy1j/3TE=
github.com/aws/aws-sdk-go-v2/feature/ec2/imds v1.18.20 h1:zOgq3uezl5nznfoK3ODuqbhVg1JzAGDUhXOsU0IDCAo=
github.com/aws/aws-sdk-go-v2/feature/ec2/imds v1.18.20/go.mod h1:z/MVwUARehy6GAg/yQ1GO2IMl0k++cu1ohP9zo887wE=
github.com/aws/aws-sdk-go-v2/config v1.32.13 h1:5KgbxMaS2coSWRrx9TX/QtWbqzgQkOdEa3sZPhBhCSg=
github.com/aws/aws-sdk-go-v2/config v1.32.13/go.mod h1:8zz7wedqtCbw5e9Mi2doEwDyEgHcEE9YOJp6a8jdSMY=
github.com/aws/aws-sdk-go-v2/credentials v1.19.13 h1:mA59E3fokBvyEGHKFdnpNNrvaR351cqiHgRg+JzOSRI=
github.com/aws/aws-sdk-go-v2/credentials v1.19.13/go.mod h1:yoTXOQKea18nrM69wGF9jBdG4WocSZA1h38A+t/MAsk=
github.com/aws/aws-sdk-go-v2/feature/ec2/imds v1.18.21 h1:NUS3K4BTDArQqNu2ih7yeDLaS3bmHD0YndtA6UP884g=
github.com/aws/aws-sdk-go-v2/feature/ec2/imds v1.18.21/go.mod h1:YWNWJQNjKigKY1RHVJCuupeWDrrHjRqHm0N9rdrWzYI=
github.com/aws/aws-sdk-go-v2/feature/s3/manager v1.20.12 h1:Zy6Tme1AA13kX8x3CnkHx5cqdGWGaj/anwOiWGnA0Xo=
github.com/aws/aws-sdk-go-v2/feature/s3/manager v1.20.12/go.mod h1:ql4uXYKoTM9WUAUSmthY4AtPVrlTBZOvnBJTiCUdPxI=
github.com/aws/aws-sdk-go-v2/internal/configsources v1.4.20 h1:CNXO7mvgThFGqOFgbNAP2nol2qAWBOGfqR/7tQlvLmc=
github.com/aws/aws-sdk-go-v2/internal/configsources v1.4.20/go.mod h1:oydPDJKcfMhgfcgBUZaG+toBbwy8yPWubJXBVERtI4o=
github.com/aws/aws-sdk-go-v2/internal/endpoints/v2 v2.7.20 h1:tN6W/hg+pkM+tf9XDkWUbDEjGLb+raoBMFsTodcoYKw=
github.com/aws/aws-sdk-go-v2/internal/endpoints/v2 v2.7.20/go.mod h1:YJ898MhD067hSHA6xYCx5ts/jEd8BSOLtQDL3iZsvbc=
github.com/aws/aws-sdk-go-v2/internal/ini v1.8.4 h1:WKuaxf++XKWlHWu9ECbMlha8WOEGm0OUEZqm4K/Gcfk=
github.com/aws/aws-sdk-go-v2/internal/ini v1.8.4/go.mod h1:ZWy7j6v1vWGmPReu0iSGvRiise4YI5SkR3OHKTZ6Wuc=
github.com/aws/aws-sdk-go-v2/internal/configsources v1.4.21 h1:Rgg6wvjjtX8bNHcvi9OnXWwcE0a2vGpbwmtICOsvcf4=
github.com/aws/aws-sdk-go-v2/internal/configsources v1.4.21/go.mod h1:A/kJFst/nm//cyqonihbdpQZwiUhhzpqTsdbhDdRF9c=
github.com/aws/aws-sdk-go-v2/internal/endpoints/v2 v2.7.21 h1:PEgGVtPoB6NTpPrBgqSE5hE/o47Ij9qk/SEZFbUOe9A=
github.com/aws/aws-sdk-go-v2/internal/endpoints/v2 v2.7.21/go.mod h1:p+hz+PRAYlY3zcpJhPwXlLC4C+kqn70WIHwnzAfs6ps=
github.com/aws/aws-sdk-go-v2/internal/ini v1.8.6 h1:qYQ4pzQ2Oz6WpQ8T3HvGHnZydA72MnLuFK9tJwmrbHw=
github.com/aws/aws-sdk-go-v2/internal/ini v1.8.6/go.mod h1:O3h0IK87yXci+kg6flUKzJnWeziQUKciKrLjcatSNcY=
github.com/aws/aws-sdk-go-v2/internal/v4a v1.4.17 h1:JqcdRG//czea7Ppjb+g/n4o8i/R50aTBHkA7vu0lK+k=
github.com/aws/aws-sdk-go-v2/internal/v4a v1.4.17/go.mod h1:CO+WeGmIdj/MlPel2KwID9Gt7CNq4M65HUfBW97liM0=
github.com/aws/aws-sdk-go-v2/service/internal/accept-encoding v1.13.7 h1:5EniKhLZe4xzL7a+fU3C2tfUN4nWIqlLesfrjkuPFTY=
github.com/aws/aws-sdk-go-v2/service/internal/accept-encoding v1.13.7/go.mod h1:x0nZssQ3qZSnIcePWLvcoFisRXJzcTVvYpAAdYX8+GI=
github.com/aws/aws-sdk-go-v2/service/internal/checksum v1.9.8 h1:Z5EiPIzXKewUQK0QTMkutjiaPVeVYXX7KIqhXu/0fXs=
github.com/aws/aws-sdk-go-v2/service/internal/checksum v1.9.8/go.mod h1:FsTpJtvC4U1fyDXk7c71XoDv3HlRm8V3NiYLeYLh5YE=
github.com/aws/aws-sdk-go-v2/service/internal/presigned-url v1.13.20 h1:2HvVAIq+YqgGotK6EkMf+KIEqTISmTYh5zLpYyeTo1Y=
github.com/aws/aws-sdk-go-v2/service/internal/presigned-url v1.13.20/go.mod h1:V4X406Y666khGa8ghKmphma/7C0DAtEQYhkq9z4vpbk=
github.com/aws/aws-sdk-go-v2/service/internal/presigned-url v1.13.21 h1:c31//R3xgIJMSC8S6hEVq+38DcvUlgFY0FM6mSI5oto=
github.com/aws/aws-sdk-go-v2/service/internal/presigned-url v1.13.21/go.mod h1:r6+pf23ouCB718FUxaqzZdbpYFyDtehyZcmP5KL9FkA=
github.com/aws/aws-sdk-go-v2/service/internal/s3shared v1.19.17 h1:bGeHBsGZx0Dvu/eJC0Lh9adJa3M1xREcndxLNZlve2U=
github.com/aws/aws-sdk-go-v2/service/internal/s3shared v1.19.17/go.mod h1:dcW24lbU0CzHusTE8LLHhRLI42ejmINN8Lcr22bwh/g=
github.com/aws/aws-sdk-go-v2/service/s3 v1.96.0 h1:oeu8VPlOre74lBA/PMhxa5vewaMIMmILM+RraSyB8KA=
github.com/aws/aws-sdk-go-v2/service/s3 v1.96.0/go.mod h1:5jggDlZ2CLQhwJBiZJb4vfk4f0GxWdEDruWKEJ1xOdo=
github.com/aws/aws-sdk-go-v2/service/signin v1.0.8 h1:0GFOLzEbOyZABS3PhYfBIx2rNBACYcKty+XGkTgw1ow=
github.com/aws/aws-sdk-go-v2/service/signin v1.0.8/go.mod h1:LXypKvk85AROkKhOG6/YEcHFPoX+prKTowKnVdcaIxE=
github.com/aws/aws-sdk-go-v2/service/sso v1.30.13 h1:kiIDLZ005EcKomYYITtfsjn7dtOwHDOFy7IbPXKek2o=
github.com/aws/aws-sdk-go-v2/service/sso v1.30.13/go.mod h1:2h/xGEowcW/g38g06g3KpRWDlT+OTfxxI0o1KqayAB8=
github.com/aws/aws-sdk-go-v2/service/ssooidc v1.35.17 h1:jzKAXIlhZhJbnYwHbvUQZEB8KfgAEuG0dc08Bkda7NU=
github.com/aws/aws-sdk-go-v2/service/ssooidc v1.35.17/go.mod h1:Al9fFsXjv4KfbzQHGe6V4NZSZQXecFcvaIF4e70FoRA=
github.com/aws/aws-sdk-go-v2/service/sts v1.41.9 h1:Cng+OOwCHmFljXIxpEVXAGMnBia8MSU6Ch5i9PgBkcU=
github.com/aws/aws-sdk-go-v2/service/sts v1.41.9/go.mod h1:LrlIndBDdjA/EeXeyNBle+gyCwTlizzW5ycgWnvIxkk=
github.com/aws/aws-sdk-go-v2/service/signin v1.0.9 h1:QKZH0S178gCmFEgst8hN0mCX1KxLgHBKKY/CLqwP8lg=
github.com/aws/aws-sdk-go-v2/service/signin v1.0.9/go.mod h1:7yuQJoT+OoH8aqIxw9vwF+8KpvLZ8AWmvmUWHsGQZvI=
github.com/aws/aws-sdk-go-v2/service/sso v1.30.14 h1:GcLE9ba5ehAQma6wlopUesYg/hbcOhFNWTjELkiWkh4=
github.com/aws/aws-sdk-go-v2/service/sso v1.30.14/go.mod h1:WSvS1NLr7JaPunCXqpJnWk1Bjo7IxzZXrZi1QQCkuqM=
github.com/aws/aws-sdk-go-v2/service/ssooidc v1.35.18 h1:mP49nTpfKtpXLt5SLn8Uv8z6W+03jYVoOSAl/c02nog=
github.com/aws/aws-sdk-go-v2/service/ssooidc v1.35.18/go.mod h1:YO8TrYtFdl5w/4vmjL8zaBSsiNp3w0L1FfKVKenZT7w=
github.com/aws/aws-sdk-go-v2/service/sts v1.41.10 h1:p8ogvvLugcR/zLBXTXrTkj0RYBUdErbMnAFFp12Lm/U=
github.com/aws/aws-sdk-go-v2/service/sts v1.41.10/go.mod h1:60dv0eZJfeVXfbT1tFJinbHrDfSJ2GZl4Q//OSSNAVw=
github.com/aws/smithy-go v1.24.2 h1:FzA3bu/nt/vDvmnkg+R8Xl46gmzEDam6mZ1hzmwXFng=
github.com/aws/smithy-go v1.24.2/go.mod h1:YE2RhdIuDbA5E5bTdciG9KrW3+TiEONeUWCqxX9i1Fc=
github.com/bahlo/generic-list-go v0.2.0 h1:5sz/EEAK+ls5wF+NeqDpk5+iNdMDXrh3z3nPnH1Wvgk=
@@ -284,8 +284,8 @@ github.com/go-git/go-billy/v5 v5.6.2/go.mod h1:rcFC2rAsp/erv7CMz9GczHcuD0D32fWzH
github.com/go-gl/glfw v0.0.0-20190409004039-e6da0acd62b1/go.mod h1:vR7hzQXu2zJy9AVAgeJqvqgH9Q5CA+iKCZ2gyEVpxRU=
github.com/go-gl/glfw/v3.3/glfw v0.0.0-20191125211704-12ad95a8df72/go.mod h1:tQ2UAYgL5IevRw8kRxooKSPJfGvJ9fJQFa0TUsXzTg8=
github.com/go-gl/glfw/v3.3/glfw v0.0.0-20200222043503-6f7a984d4dc4/go.mod h1:tQ2UAYgL5IevRw8kRxooKSPJfGvJ9fJQFa0TUsXzTg8=
github.com/go-jose/go-jose/v4 v4.1.3 h1:CVLmWDhDVRa6Mi/IgCgaopNosCaHz7zrMeF9MlZRkrs=
github.com/go-jose/go-jose/v4 v4.1.3/go.mod h1:x4oUasVrzR7071A4TnHLGSPpNOm2a21K9Kf04k1rs08=
github.com/go-jose/go-jose/v4 v4.1.4 h1:moDMcTHmvE6Groj34emNPLs/qtYXRVcd6S7NHbHz3kA=
github.com/go-jose/go-jose/v4 v4.1.4/go.mod h1:x4oUasVrzR7071A4TnHLGSPpNOm2a21K9Kf04k1rs08=
github.com/go-logr/logr v1.2.2/go.mod h1:jdQByPbusPIv2/zmleS9BjJVeZ6kBagPoEUsqbVz/1A=
github.com/go-logr/logr v1.4.3 h1:CjnDlHq8ikf6E492q6eKboGOC0T8CDaOvkHCIg8idEI=
github.com/go-logr/logr v1.4.3/go.mod h1:9T104GzyrTigFIr8wt5mBrctHMim0Nb2HLGrmQ40KvY=
@@ -707,8 +707,8 @@ github.com/xanzy/ssh-agent v0.3.3 h1:+/15pJfg/RsTxqYcX6fHqOXZwwMP+2VyYWJeWM2qQFM
github.com/xanzy/ssh-agent v0.3.3/go.mod h1:6dzNDKs0J9rVPHPhaGCukekBHKqfl+L3KghI1Bc68Uw=
github.com/xdg-go/pbkdf2 v1.0.0 h1:Su7DPu48wXMwC3bs7MCNG+z4FhcyEuz5dlvchbq0B0c=
github.com/xdg-go/pbkdf2 v1.0.0/go.mod h1:jrpuAogTd400dnrH08LKmI/xc1MbPOebTwRqcT5RDeI=
github.com/xdg-go/scram v1.1.2 h1:FHX5I5B4i4hKRVRBCFRxq1iQRej7WO3hhBuJf+UUySY=
github.com/xdg-go/scram v1.1.2/go.mod h1:RT/sEzTbU5y00aCK8UOx6R7YryM0iF1N2MOmC3kKLN4=
github.com/xdg-go/scram v1.2.0 h1:bYKF2AEwG5rqd1BumT4gAnvwU/M9nBp2pTSxeZw7Wvs=
github.com/xdg-go/scram v1.2.0/go.mod h1:3dlrS0iBaWKYVt2ZfA4cj48umJZ+cAEbR6/SjLA88I8=
github.com/xdg-go/stringprep v1.0.4 h1:XLI/Ng3O1Atzq0oBs3TWm+5ZVgkq2aqdlvP9JtoZ6c8=
github.com/xdg-go/stringprep v1.0.4/go.mod h1:mPGuuIYwz7CmR2bT9j4GbQqutWS1zV24gijq1dTyGkM=
github.com/xeipuuv/gojsonpointer v0.0.0-20180127040702-4e3ac2762d5f/go.mod h1:N2zxlSyiKSe5eX1tZViRH5QA0qijqEDrYZiPEAiq3wU=
@@ -21,7 +21,9 @@ import (
"github.com/aws/aws-sdk-go-v2/credentials"
"github.com/aws/aws-sdk-go-v2/service/s3"
"github.com/seaweedfs/seaweedfs/test/volume_server/framework"
"github.com/seaweedfs/seaweedfs/weed/cluster/lock_manager"
"github.com/seaweedfs/seaweedfs/weed/pb"
"github.com/seaweedfs/seaweedfs/weed/pb/filer_pb"
"github.com/seaweedfs/seaweedfs/weed/pb/master_pb"
"github.com/stretchr/testify/require"
"google.golang.org/grpc"
@@ -139,6 +141,7 @@ func startDistributedLockCluster(t *testing.T) *distributedLockCluster {
require.NoError(t, cluster.waitForTCP(cluster.filerGRPCAddress(i), 30*time.Second), "wait for filer %d grpc\n%s", i, cluster.tailLog(fmt.Sprintf("filer%d.log", i)))
}
require.NoError(t, cluster.waitForFilerCount(2, 30*time.Second), "wait for filer group registration")
require.NoError(t, cluster.waitForLockRingConverged(30*time.Second), "wait for lock ring convergence")
for i := 0; i < 2; i++ {
require.NoError(t, cluster.startS3(i), "start s3 %d", i)
@@ -411,6 +414,130 @@ func (c *distributedLockCluster) waitForFilerCount(expected int, timeout time.Du
return fmt.Errorf("timed out waiting for %d filers in group %q", expected, c.filerGroup)
}
// waitForLockRingConverged verifies that both filers have a consistent view of the
// lock ring by acquiring the same lock through each filer and checking mutual exclusion.
// This guards against the window between master seeing both filers and the filers
// actually receiving the LockRingUpdate broadcast (delayed by the stabilization timer).
//
// A single arbitrary key could false-pass: if the key's real primary is filer0 and
// filer0 has a stale ring (only sees itself), it still grants correctly because it IS
// the primary. To catch the stale ring we must also test a key whose real primary is
// filer1. So we generate one test key per primary filer using the same consistent-hash
// ring as production, and require mutual exclusion for all of them.
func (c *distributedLockCluster) waitForLockRingConverged(timeout time.Duration) error {
deadline := time.Now().Add(timeout)
owners := make([]pb.ServerAddress, 0, len(c.filerPorts))
for i := range c.filerPorts {
owners = append(owners, c.filerServerAddress(i))
}
ring := lock_manager.NewHashRing(lock_manager.DefaultVnodeCount)
ring.SetServers(owners)
attempt := 0
for time.Now().Before(deadline) {
// Generate one unique test key per primary filer
testKeys := c.convergenceKeysPerPrimary(ring, owners, attempt)
attempt++
allConverged := true
for _, key := range testKeys {
converged, _ := c.checkLockMutualExclusion(key)
if !converged {
allConverged = false
break
}
}
if allConverged {
return nil
}
time.Sleep(500 * time.Millisecond)
}
return fmt.Errorf("lock ring did not converge: both filers independently grant the same lock")
}
// convergenceKeysPerPrimary returns one lock key per distinct primary filer,
// using the same consistent-hash ring as production routing.
func (c *distributedLockCluster) convergenceKeysPerPrimary(ring *lock_manager.HashRing, owners []pb.ServerAddress, attempt int) []string {
keysByOwner := make(map[pb.ServerAddress]string, len(owners))
for i := 0; len(keysByOwner) < len(owners) && i < 1024; i++ {
candidate := fmt.Sprintf("convergence-%d-%d", attempt, i)
primary := ring.GetPrimary(candidate)
if _, exists := keysByOwner[primary]; !exists {
keysByOwner[primary] = candidate
}
}
keys := make([]string, 0, len(keysByOwner))
for _, k := range keysByOwner {
keys = append(keys, k)
}
return keys
}
// checkLockMutualExclusion acquires a lock via filer0, then tries the same lock via filer1.
// Returns true if mutual exclusion holds (second attempt is denied).
func (c *distributedLockCluster) checkLockMutualExclusion(testKey string) (bool, error) {
conn0, err := grpc.NewClient(c.filerGRPCAddress(0), grpc.WithTransportCredentials(insecure.NewCredentials()))
if err != nil {
return false, err
}
defer conn0.Close()
conn1, err := grpc.NewClient(c.filerGRPCAddress(1), grpc.WithTransportCredentials(insecure.NewCredentials()))
if err != nil {
return false, err
}
defer conn1.Close()
client0 := filer_pb.NewSeaweedFilerClient(conn0)
client1 := filer_pb.NewSeaweedFilerClient(conn1)
// Acquire lock via filer0
ctx0, cancel0 := context.WithTimeout(context.Background(), 2*time.Second)
defer cancel0()
resp0, err := client0.DistributedLock(ctx0, &filer_pb.LockRequest{
Name: testKey,
SecondsToLock: 5,
Owner: "convergence-filer0",
})
if err != nil || resp0.RenewToken == "" {
return false, fmt.Errorf("filer0 lock failed: err=%v resp=%v", err, resp0)
}
defer func() {
// Always release the lock we acquired
client0.DistributedUnlock(context.Background(), &filer_pb.UnlockRequest{
Name: testKey,
RenewToken: resp0.RenewToken,
})
}()
// Try the same lock via filer1 - should be denied
ctx1, cancel1 := context.WithTimeout(context.Background(), 2*time.Second)
defer cancel1()
resp1, err := client1.DistributedLock(ctx1, &filer_pb.LockRequest{
Name: testKey,
SecondsToLock: 5,
Owner: "convergence-filer1",
})
if err != nil {
return false, err
}
if resp1.RenewToken != "" {
// Both filers granted the lock - ring not converged. Release the second lock.
client1.DistributedUnlock(context.Background(), &filer_pb.UnlockRequest{
Name: testKey,
RenewToken: resp1.RenewToken,
})
return false, nil
}
// Verify the denial is specifically because the lock is already held,
// not due to a transient error that might give a false positive.
if !strings.Contains(resp1.Error, "lock already owned") {
return false, nil
}
return true, nil
}
func (c *distributedLockCluster) waitForHTTP(url string, timeout time.Duration) error {
client := &net.Dialer{Timeout: time.Second}
httpClient := &httpClientWithDialer{dialer: client}
@@ -6,7 +6,6 @@ import (
"errors"
"fmt"
"io"
"sort"
"strings"
"sync"
"testing"
@@ -15,9 +14,9 @@ import (
"github.com/aws/aws-sdk-go-v2/aws"
"github.com/aws/aws-sdk-go-v2/service/s3"
"github.com/aws/smithy-go"
"github.com/seaweedfs/seaweedfs/weed/cluster/lock_manager"
"github.com/seaweedfs/seaweedfs/weed/pb"
"github.com/seaweedfs/seaweedfs/weed/s3api/s3_constants"
"github.com/seaweedfs/seaweedfs/weed/util"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
@@ -149,14 +148,15 @@ func (c *distributedLockCluster) findLockOwnerKeys(bucket, prefix string) map[pb
for i := range c.filerPorts {
owners = append(owners, c.filerServerAddress(i))
}
sort.Slice(owners, func(i, j int) bool {
return owners[i] < owners[j]
})
ring := lock_manager.NewHashRing(lock_manager.DefaultVnodeCount)
ring.SetServers(owners)
keysByOwner := make(map[pb.ServerAddress]string, len(owners))
for i := 0; i < 1024 && len(keysByOwner) < len(owners); i++ {
key := fmt.Sprintf("%s-%03d.txt", prefix, i)
lockOwner := ownerForObjectLock(bucket, key, owners)
lockKey := fmt.Sprintf("s3.object.write:/buckets/%s/%s", bucket, s3_constants.NormalizeObjectKey(key))
lockOwner := ring.GetPrimary(lockKey)
if _, exists := keysByOwner[lockOwner]; !exists {
keysByOwner[lockOwner] = key
}
@@ -164,15 +164,6 @@ func (c *distributedLockCluster) findLockOwnerKeys(bucket, prefix string) map[pb
return keysByOwner
}
func ownerForObjectLock(bucket, object string, owners []pb.ServerAddress) pb.ServerAddress {
lockKey := fmt.Sprintf("s3.object.write:/buckets/%s/%s", bucket, s3_constants.NormalizeObjectKey(object))
hash := util.HashStringToLong(lockKey)
if hash < 0 {
hash = -hash
}
return owners[hash%int64(len(owners))]
}
func lockOwnerLabel(owner pb.ServerAddress) string {
replacer := strings.NewReplacer(":", "_", ".", "_")
return "owner_" + replacer.Replace(string(owner))
@@ -0,0 +1,511 @@
package iam
import (
"encoding/xml"
"fmt"
"io"
"net/http"
"net/url"
"os"
"strings"
"testing"
"time"
"github.com/aws/aws-sdk-go/aws"
"github.com/aws/aws-sdk-go/aws/credentials"
"github.com/aws/aws-sdk-go/aws/session"
"github.com/aws/aws-sdk-go/service/s3"
v4 "github.com/aws/aws-sdk-go/aws/signer/v4"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
// GetFederationTokenTestResponse represents the STS GetFederationToken response
type GetFederationTokenTestResponse struct {
XMLName xml.Name `xml:"GetFederationTokenResponse"`
Result struct {
Credentials struct {
AccessKeyId string `xml:"AccessKeyId"`
SecretAccessKey string `xml:"SecretAccessKey"`
SessionToken string `xml:"SessionToken"`
Expiration string `xml:"Expiration"`
} `xml:"Credentials"`
FederatedUser struct {
FederatedUserId string `xml:"FederatedUserId"`
Arn string `xml:"Arn"`
} `xml:"FederatedUser"`
} `xml:"GetFederationTokenResult"`
}
func getTestCredentials() (string, string) {
accessKey := os.Getenv("STS_TEST_ACCESS_KEY")
if accessKey == "" {
accessKey = "admin"
}
secretKey := os.Getenv("STS_TEST_SECRET_KEY")
if secretKey == "" {
secretKey = "admin"
}
return accessKey, secretKey
}
// isGetFederationTokenImplemented checks if the running server supports GetFederationToken
func isGetFederationTokenImplemented(t *testing.T) bool {
accessKey, secretKey := getTestCredentials()
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"GetFederationToken"},
"Version": {"2011-06-15"},
"Name": {"probe"},
}, accessKey, secretKey)
if err != nil {
return false
}
defer resp.Body.Close()
body, _ := io.ReadAll(resp.Body)
var errResp STSErrorTestResponse
if xml.Unmarshal(body, &errResp) == nil {
if errResp.Error.Code == "InvalidAction" || errResp.Error.Code == "NotImplemented" {
return false
}
}
return true
}
// TestSTSGetFederationTokenValidation tests input validation for the GetFederationToken endpoint
func TestSTSGetFederationTokenValidation(t *testing.T) {
if testing.Short() {
t.Skip("Skipping integration test in short mode")
}
if !isSTSEndpointRunning(t) {
t.Fatal("SeaweedFS STS endpoint is not running at", TestSTSEndpoint, "- please run 'make setup-all-tests' first")
}
if !isGetFederationTokenImplemented(t) {
t.Fatal("GetFederationToken action is not implemented in the running server")
}
accessKey, secretKey := getTestCredentials()
t.Run("missing_name", func(t *testing.T) {
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"GetFederationToken"},
"Version": {"2011-06-15"},
// Name is missing
}, accessKey, secretKey)
require.NoError(t, err)
defer resp.Body.Close()
body, _ := io.ReadAll(resp.Body)
var errResp STSErrorTestResponse
require.NoError(t, xml.Unmarshal(body, &errResp), "Failed to parse: %s", string(body))
assert.Equal(t, "MissingParameter", errResp.Error.Code)
})
t.Run("name_too_short", func(t *testing.T) {
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"GetFederationToken"},
"Version": {"2011-06-15"},
"Name": {"A"},
}, accessKey, secretKey)
require.NoError(t, err)
defer resp.Body.Close()
body, _ := io.ReadAll(resp.Body)
var errResp STSErrorTestResponse
require.NoError(t, xml.Unmarshal(body, &errResp), "Failed to parse: %s", string(body))
assert.Equal(t, "InvalidParameterValue", errResp.Error.Code)
})
t.Run("name_too_long", func(t *testing.T) {
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"GetFederationToken"},
"Version": {"2011-06-15"},
"Name": {strings.Repeat("A", 33)},
}, accessKey, secretKey)
require.NoError(t, err)
defer resp.Body.Close()
body, _ := io.ReadAll(resp.Body)
var errResp STSErrorTestResponse
require.NoError(t, xml.Unmarshal(body, &errResp), "Failed to parse: %s", string(body))
assert.Equal(t, "InvalidParameterValue", errResp.Error.Code)
})
t.Run("name_invalid_characters", func(t *testing.T) {
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"GetFederationToken"},
"Version": {"2011-06-15"},
"Name": {"bad name"},
}, accessKey, secretKey)
require.NoError(t, err)
defer resp.Body.Close()
body, _ := io.ReadAll(resp.Body)
var errResp STSErrorTestResponse
require.NoError(t, xml.Unmarshal(body, &errResp), "Failed to parse: %s", string(body))
assert.Equal(t, "InvalidParameterValue", errResp.Error.Code)
})
t.Run("duration_too_short", func(t *testing.T) {
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"GetFederationToken"},
"Version": {"2011-06-15"},
"Name": {"TestApp"},
"DurationSeconds": {"100"},
}, accessKey, secretKey)
require.NoError(t, err)
defer resp.Body.Close()
body, _ := io.ReadAll(resp.Body)
var errResp STSErrorTestResponse
require.NoError(t, xml.Unmarshal(body, &errResp), "Failed to parse: %s", string(body))
assert.Equal(t, "InvalidParameterValue", errResp.Error.Code)
})
t.Run("duration_too_long", func(t *testing.T) {
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"GetFederationToken"},
"Version": {"2011-06-15"},
"Name": {"TestApp"},
"DurationSeconds": {"200000"},
}, accessKey, secretKey)
require.NoError(t, err)
defer resp.Body.Close()
body, _ := io.ReadAll(resp.Body)
var errResp STSErrorTestResponse
require.NoError(t, xml.Unmarshal(body, &errResp), "Failed to parse: %s", string(body))
assert.Equal(t, "InvalidParameterValue", errResp.Error.Code)
})
t.Run("malformed_policy", func(t *testing.T) {
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"GetFederationToken"},
"Version": {"2011-06-15"},
"Name": {"TestApp"},
"Policy": {"not-valid-json"},
}, accessKey, secretKey)
require.NoError(t, err)
defer resp.Body.Close()
body, _ := io.ReadAll(resp.Body)
var errResp STSErrorTestResponse
require.NoError(t, xml.Unmarshal(body, &errResp), "Failed to parse: %s", string(body))
assert.Equal(t, "MalformedPolicyDocument", errResp.Error.Code)
})
t.Run("anonymous_rejected", func(t *testing.T) {
// GetFederationToken requires SigV4, anonymous should fail
resp, err := callSTSAPI(t, url.Values{
"Action": {"GetFederationToken"},
"Version": {"2011-06-15"},
"Name": {"TestApp"},
})
require.NoError(t, err)
defer resp.Body.Close()
assert.NotEqual(t, http.StatusOK, resp.StatusCode)
})
}
// TestSTSGetFederationTokenRejectTemporaryCredentials tests that temporary
// credentials (session tokens) are rejected by GetFederationToken
func TestSTSGetFederationTokenRejectTemporaryCredentials(t *testing.T) {
if testing.Short() {
t.Skip("Skipping integration test in short mode")
}
if !isSTSEndpointRunning(t) {
t.Skip("SeaweedFS STS endpoint is not running at", TestSTSEndpoint)
}
if !isGetFederationTokenImplemented(t) {
t.Skip("GetFederationToken not implemented")
}
accessKey, secretKey := getTestCredentials()
// First, obtain temporary credentials via AssumeRole
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"AssumeRole"},
"Version": {"2011-06-15"},
"RoleArn": {"arn:aws:iam::role/admin"},
"RoleSessionName": {"temp-session"},
}, accessKey, secretKey)
require.NoError(t, err)
defer resp.Body.Close()
body, err := io.ReadAll(resp.Body)
require.NoError(t, err)
if resp.StatusCode != http.StatusOK {
t.Skipf("AssumeRole failed (may not be configured): status=%d body=%s", resp.StatusCode, string(body))
}
var assumeResp AssumeRoleTestResponse
require.NoError(t, xml.Unmarshal(body, &assumeResp), "Parse AssumeRole response: %s", string(body))
tempAccessKey := assumeResp.Result.Credentials.AccessKeyId
tempSecretKey := assumeResp.Result.Credentials.SecretAccessKey
tempSessionToken := assumeResp.Result.Credentials.SessionToken
require.NotEmpty(t, tempAccessKey)
require.NotEmpty(t, tempSessionToken)
// Now try GetFederationToken with the temporary credentials
// Include X-Amz-Security-Token header which marks this as a temp credential call
params := url.Values{
"Action": {"GetFederationToken"},
"Version": {"2011-06-15"},
"Name": {"ShouldFail"},
}
reqBody := params.Encode()
req, err := http.NewRequest(http.MethodPost, TestSTSEndpoint+"/", strings.NewReader(reqBody))
require.NoError(t, err)
req.Header.Set("Content-Type", "application/x-www-form-urlencoded")
req.Header.Set("X-Amz-Security-Token", tempSessionToken)
creds := credentials.NewStaticCredentials(tempAccessKey, tempSecretKey, tempSessionToken)
signer := v4.NewSigner(creds)
_, err = signer.Sign(req, strings.NewReader(reqBody), "sts", "us-east-1", time.Now())
require.NoError(t, err)
client := &http.Client{Timeout: 30 * time.Second}
resp2, err := client.Do(req)
require.NoError(t, err)
defer resp2.Body.Close()
body2, _ := io.ReadAll(resp2.Body)
assert.Equal(t, http.StatusForbidden, resp2.StatusCode,
"GetFederationToken should reject temporary credentials: %s", string(body2))
assert.Contains(t, string(body2), "temporary credentials",
"Error should mention temporary credentials")
}
// TestSTSGetFederationTokenSuccess tests a successful GetFederationToken call
// and verifies the returned credentials can be used to access S3
func TestSTSGetFederationTokenSuccess(t *testing.T) {
if testing.Short() {
t.Skip("Skipping integration test in short mode")
}
if !isSTSEndpointRunning(t) {
t.Skip("SeaweedFS STS endpoint is not running at", TestSTSEndpoint)
}
if !isGetFederationTokenImplemented(t) {
t.Skip("GetFederationToken not implemented")
}
accessKey, secretKey := getTestCredentials()
t.Run("basic_success", func(t *testing.T) {
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"GetFederationToken"},
"Version": {"2011-06-15"},
"Name": {"AppClient"},
}, accessKey, secretKey)
require.NoError(t, err)
defer resp.Body.Close()
body, err := io.ReadAll(resp.Body)
require.NoError(t, err)
t.Logf("Response status: %d, body: %s", resp.StatusCode, string(body))
if resp.StatusCode != http.StatusOK {
var errResp STSErrorTestResponse
_ = xml.Unmarshal(body, &errResp)
t.Fatalf("GetFederationToken failed: code=%s message=%s", errResp.Error.Code, errResp.Error.Message)
}
var stsResp GetFederationTokenTestResponse
require.NoError(t, xml.Unmarshal(body, &stsResp), "Parse response: %s", string(body))
creds := stsResp.Result.Credentials
assert.NotEmpty(t, creds.AccessKeyId)
assert.NotEmpty(t, creds.SecretAccessKey)
assert.NotEmpty(t, creds.SessionToken)
assert.NotEmpty(t, creds.Expiration)
fedUser := stsResp.Result.FederatedUser
assert.Contains(t, fedUser.Arn, "federated-user/AppClient")
assert.Contains(t, fedUser.FederatedUserId, "AppClient")
})
t.Run("with_custom_duration", func(t *testing.T) {
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"GetFederationToken"},
"Version": {"2011-06-15"},
"Name": {"DurationTest"},
"DurationSeconds": {"3600"},
}, accessKey, secretKey)
require.NoError(t, err)
defer resp.Body.Close()
body, _ := io.ReadAll(resp.Body)
t.Logf("Response status: %d, body: %s", resp.StatusCode, string(body))
if resp.StatusCode == http.StatusOK {
var stsResp GetFederationTokenTestResponse
require.NoError(t, xml.Unmarshal(body, &stsResp))
assert.NotEmpty(t, stsResp.Result.Credentials.AccessKeyId)
// Verify expiration is roughly 1 hour from now
expTime, err := time.Parse(time.RFC3339, stsResp.Result.Credentials.Expiration)
require.NoError(t, err)
diff := time.Until(expTime)
assert.InDelta(t, 3600, diff.Seconds(), 60,
"Expiration should be ~1 hour from now")
}
})
t.Run("with_36_hour_duration", func(t *testing.T) {
// GetFederationToken allows up to 36 hours (unlike AssumeRole's 12h max)
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"GetFederationToken"},
"Version": {"2011-06-15"},
"Name": {"LongDuration"},
"DurationSeconds": {"129600"}, // 36 hours
}, accessKey, secretKey)
require.NoError(t, err)
defer resp.Body.Close()
body, _ := io.ReadAll(resp.Body)
if resp.StatusCode == http.StatusOK {
var stsResp GetFederationTokenTestResponse
require.NoError(t, xml.Unmarshal(body, &stsResp))
expTime, err := time.Parse(time.RFC3339, stsResp.Result.Credentials.Expiration)
require.NoError(t, err)
diff := time.Until(expTime)
assert.InDelta(t, 129600, diff.Seconds(), 60,
"Expiration should be ~36 hours from now")
} else {
// Duration should not cause a rejection
var errResp STSErrorTestResponse
_ = xml.Unmarshal(body, &errResp)
assert.NotContains(t, errResp.Error.Message, "DurationSeconds",
"36-hour duration should be accepted by GetFederationToken")
}
})
}
// TestSTSGetFederationTokenWithSessionPolicy tests that vended credentials
// are scoped down by an inline session policy
func TestSTSGetFederationTokenWithSessionPolicy(t *testing.T) {
if testing.Short() {
t.Skip("Skipping integration test in short mode")
}
if !isSTSEndpointRunning(t) {
t.Skip("SeaweedFS STS endpoint is not running at", TestSTSEndpoint)
}
if !isGetFederationTokenImplemented(t) {
t.Skip("GetFederationToken not implemented")
}
accessKey, secretKey := getTestCredentials()
// Create a test bucket using admin credentials
adminSess, err := session.NewSession(&aws.Config{
Region: aws.String("us-east-1"),
Endpoint: aws.String(TestSTSEndpoint),
DisableSSL: aws.Bool(true),
S3ForcePathStyle: aws.Bool(true),
Credentials: credentials.NewStaticCredentials(accessKey, secretKey, ""),
})
require.NoError(t, err)
adminS3 := s3.New(adminSess)
bucket := fmt.Sprintf("fed-token-test-%d", time.Now().UnixNano())
_, err = adminS3.CreateBucket(&s3.CreateBucketInput{Bucket: aws.String(bucket)})
require.NoError(t, err)
defer adminS3.DeleteBucket(&s3.DeleteBucketInput{Bucket: aws.String(bucket)})
_, err = adminS3.PutObject(&s3.PutObjectInput{
Bucket: aws.String(bucket),
Key: aws.String("test.txt"),
Body: strings.NewReader("hello"),
})
require.NoError(t, err)
defer adminS3.DeleteObject(&s3.DeleteObjectInput{Bucket: aws.String(bucket), Key: aws.String("test.txt")})
// Get federated credentials with a session policy that only allows GetObject
sessionPolicy := fmt.Sprintf(`{
"Version": "2012-10-17",
"Statement": [{
"Effect": "Allow",
"Action": ["s3:GetObject"],
"Resource": ["arn:aws:s3:::%s/*"]
}]
}`, bucket)
resp, err := callSTSAPIWithSigV4(t, url.Values{
"Action": {"GetFederationToken"},
"Version": {"2011-06-15"},
"Name": {"ScopedClient"},
"Policy": {sessionPolicy},
}, accessKey, secretKey)
require.NoError(t, err)
defer resp.Body.Close()
body, _ := io.ReadAll(resp.Body)
t.Logf("GetFederationToken response: status=%d body=%s", resp.StatusCode, string(body))
if resp.StatusCode != http.StatusOK {
t.Skipf("GetFederationToken failed (may need IAM policy config): %s", string(body))
}
var stsResp GetFederationTokenTestResponse
require.NoError(t, xml.Unmarshal(body, &stsResp))
fedCreds := stsResp.Result.Credentials
require.NotEmpty(t, fedCreds.AccessKeyId)
require.NotEmpty(t, fedCreds.SessionToken)
// Create S3 client with the federated credentials
fedSess, err := session.NewSession(&aws.Config{
Region: aws.String("us-east-1"),
Endpoint: aws.String(TestSTSEndpoint),
DisableSSL: aws.Bool(true),
S3ForcePathStyle: aws.Bool(true),
Credentials: credentials.NewStaticCredentials(
fedCreds.AccessKeyId, fedCreds.SecretAccessKey, fedCreds.SessionToken),
})
require.NoError(t, err)
fedS3 := s3.New(fedSess)
// GetObject should succeed (allowed by session policy)
getResp, err := fedS3.GetObject(&s3.GetObjectInput{
Bucket: aws.String(bucket),
Key: aws.String("test.txt"),
})
if err == nil {
defer getResp.Body.Close()
t.Log("GetObject with federated credentials succeeded (as expected)")
} else {
t.Logf("GetObject with federated credentials: %v (session policy enforcement may vary)", err)
}
// PutObject should be denied (not allowed by session policy)
_, err = fedS3.PutObject(&s3.PutObjectInput{
Bucket: aws.String(bucket),
Key: aws.String("denied.txt"),
Body: strings.NewReader("should fail"),
})
if err != nil {
t.Log("PutObject correctly denied with federated credentials")
assert.Contains(t, err.Error(), "AccessDenied",
"PutObject should be denied by session policy")
} else {
// Clean up if unexpectedly succeeded
adminS3.DeleteObject(&s3.DeleteObjectInput{Bucket: aws.String(bucket), Key: aws.String("denied.txt")})
t.Log("PutObject unexpectedly succeeded — session policy enforcement may not be active")
}
}
+5 -4
View File
@@ -98,8 +98,8 @@ start-seaweedfs: check-binary
# Create S3 configuration with SSE-KMS support
@printf '{"identities":[{"name":"%s","credentials":[{"accessKey":"%s","secretKey":"%s"}],"actions":["Admin","Read","Write"]}],"kms":{"type":"%s","configs":{"keyId":"%s","encryptionContext":{},"bucketKey":false}}}' "$(ACCESS_KEY)" "$(ACCESS_KEY)" "$(SECRET_KEY)" "$(KMS_TYPE)" "$(KMS_KEY_ID)" > /tmp/seaweedfs-sse-s3.json
# Start weed mini
@AWS_ACCESS_KEY_ID=$(ACCESS_KEY) AWS_SECRET_ACCESS_KEY=$(SECRET_KEY) $(SEAWEEDFS_BINARY) mini \
# Start weed mini (WEED_S3_SSE_KEY enables SSE-S3 encryption)
@AWS_ACCESS_KEY_ID=$(ACCESS_KEY) AWS_SECRET_ACCESS_KEY=$(SECRET_KEY) WEED_S3_SSE_KEY=test-sse-s3-key $(SEAWEEDFS_BINARY) mini \
-dir=/tmp/seaweedfs-test-sse \
-s3.port=$(S3_PORT) \
-s3.config=/tmp/seaweedfs-sse-s3.json \
@@ -337,8 +337,9 @@ start-seaweedfs-ci: check-binary
s3-config-template.json > /tmp/seaweedfs-s3.json
# Start weed mini with embedded S3 using the JSON config (with verbose logging)
# WEED_S3_SSE_KEY enables SSE-S3 encryption for testing (KEK derived via HKDF)
@echo "Starting weed mini with embedded S3..."
@AWS_ACCESS_KEY_ID=$(ACCESS_KEY) AWS_SECRET_ACCESS_KEY=$(SECRET_KEY) GLOG_v=4 $(SEAWEEDFS_BINARY) mini \
@AWS_ACCESS_KEY_ID=$(ACCESS_KEY) AWS_SECRET_ACCESS_KEY=$(SECRET_KEY) WEED_S3_SSE_KEY=test-sse-s3-key GLOG_v=4 $(SEAWEEDFS_BINARY) mini \
-dir=/tmp/seaweedfs-test-sse \
-s3.port=$(S3_PORT) \
-s3.config=/tmp/seaweedfs-s3.json \
@@ -482,7 +483,7 @@ test-volume-encryption: build-weed
-e 's/SECRET_KEY_PLACEHOLDER/$(SECRET_KEY)/g' \
s3-config-template.json > /tmp/seaweedfs-s3.json
@echo "Starting weed mini with S3 volume encryption..."
@AWS_ACCESS_KEY_ID=$(ACCESS_KEY) AWS_SECRET_ACCESS_KEY=$(SECRET_KEY) GLOG_v=4 $(SEAWEEDFS_BINARY) mini \
@AWS_ACCESS_KEY_ID=$(ACCESS_KEY) AWS_SECRET_ACCESS_KEY=$(SECRET_KEY) WEED_S3_SSE_KEY=test-sse-s3-key GLOG_v=4 $(SEAWEEDFS_BINARY) mini \
-dir=/tmp/seaweedfs-test-sse \
-s3.port=$(S3_PORT) \
-s3.config=/tmp/seaweedfs-s3.json \
@@ -289,6 +289,157 @@ func TestVersioningPaginationMultipleObjectsManyVersions(t *testing.T) {
})
}
// TestVersioningPaginationDeepDirectoryHierarchy tests that paginated ListObjectVersions
// correctly skips directory subtrees before the key-marker. This reproduces the
// real-world scenario where Veeam backup objects are spread across many subdirectories
// (e.g., Mailboxes/<uuid>/ItemsData/<file>) and pagination becomes exponentially
// slower as the marker advances through the tree.
//
// Run with: ENABLE_STRESS_TESTS=true go test -v -run TestVersioningPaginationDeepDirectoryHierarchy -timeout 10m
func TestVersioningPaginationDeepDirectoryHierarchy(t *testing.T) {
if os.Getenv("ENABLE_STRESS_TESTS") != "true" {
t.Skip("Skipping stress test. Set ENABLE_STRESS_TESTS=true to run.")
}
client := getS3Client(t)
bucketName := getNewBucketName()
createBucket(t, client, bucketName)
defer deleteBucket(t, client, bucketName)
enableVersioning(t, client, bucketName)
checkVersioningStatus(t, client, bucketName, types.BucketVersioningStatusEnabled)
// Create a deep directory structure mimicking Veeam 365 backup layout:
// Backup/Organizations/<org>/Mailboxes/<mailbox>/ItemsData/<file>
numMailboxes := 20
filesPerMailbox := 5
totalObjects := numMailboxes * filesPerMailbox
orgPrefix := "Backup/Organizations/org-001/Mailboxes"
t.Logf("Creating %d objects across %d subdirectories (depth=6)...", totalObjects, numMailboxes)
startTime := time.Now()
allKeys := make([]string, 0, totalObjects)
for i := 0; i < numMailboxes; i++ {
mailboxId := fmt.Sprintf("mbx-%03d", i)
for j := 0; j < filesPerMailbox; j++ {
key := fmt.Sprintf("%s/%s/ItemsData/file-%03d.dat", orgPrefix, mailboxId, j)
_, err := client.PutObject(context.TODO(), &s3.PutObjectInput{
Bucket: aws.String(bucketName),
Key: aws.String(key),
Body: strings.NewReader(fmt.Sprintf("content-%d-%d", i, j)),
})
require.NoError(t, err)
allKeys = append(allKeys, key)
}
}
t.Logf("Created %d objects in %v", totalObjects, time.Since(startTime))
// Test 1: Paginate through all versions with a broad prefix and small maxKeys.
// This forces multiple pages that must skip earlier subdirectory trees.
t.Run("PaginateAcrossSubdirectories", func(t *testing.T) {
maxKeys := int32(10) // Force many pages to exercise marker skipping
var allVersions []types.ObjectVersion
var keyMarker, versionIdMarker *string
pageCount := 0
pageStartTimes := make([]time.Duration, 0)
for {
pageStart := time.Now()
resp, err := client.ListObjectVersions(context.TODO(), &s3.ListObjectVersionsInput{
Bucket: aws.String(bucketName),
Prefix: aws.String(orgPrefix + "/"),
MaxKeys: aws.Int32(maxKeys),
KeyMarker: keyMarker,
VersionIdMarker: versionIdMarker,
})
pageDuration := time.Since(pageStart)
require.NoError(t, err)
pageCount++
pageStartTimes = append(pageStartTimes, pageDuration)
allVersions = append(allVersions, resp.Versions...)
t.Logf("Page %d: %d versions in %v (marker: %v)",
pageCount, len(resp.Versions), pageDuration,
keyMarker)
if resp.IsTruncated == nil || !*resp.IsTruncated {
break
}
keyMarker = resp.NextKeyMarker
versionIdMarker = resp.NextVersionIdMarker
}
assert.Greater(t, pageCount, 1, "Should require multiple pages")
// Verify listed keys exactly match the keys we created (same elements, same order)
listedKeys := make([]string, 0, len(allVersions))
for _, v := range allVersions {
listedKeys = append(listedKeys, *v.Key)
}
assert.Equal(t, allKeys, listedKeys,
"Listed version keys should exactly match created keys")
// Check that later pages don't take dramatically longer than earlier ones.
// Before the fix, later pages were exponentially slower because they
// re-traversed the entire tree. With the fix, all pages should be similar.
if len(pageStartTimes) >= 4 {
firstQuarter := pageStartTimes[0]
lastQuarter := pageStartTimes[len(pageStartTimes)-1]
t.Logf("First page: %v, Last page: %v, Ratio: %.1fx",
firstQuarter, lastQuarter,
float64(lastQuarter)/float64(firstQuarter))
// Allow generous 10x ratio to avoid flakiness; before the fix
// the ratio was 100x+ on large datasets
if lastQuarter > firstQuarter*10 && lastQuarter > 500*time.Millisecond {
t.Errorf("Last page took %.1fx longer than first page (%v vs %v) — possible pagination regression",
float64(lastQuarter)/float64(firstQuarter), lastQuarter, firstQuarter)
}
}
})
// Test 2: Paginate with delimiter to verify CommonPrefixes interaction
t.Run("PaginateWithDelimiterAcrossSubdirs", func(t *testing.T) {
maxKeys := int32(5)
var allPrefixes []string
var keyMarker *string
pageCount := 0
for {
resp, err := client.ListObjectVersions(context.TODO(), &s3.ListObjectVersionsInput{
Bucket: aws.String(bucketName),
Prefix: aws.String(orgPrefix + "/"),
Delimiter: aws.String("/"),
MaxKeys: aws.Int32(maxKeys),
KeyMarker: keyMarker,
})
require.NoError(t, err)
pageCount++
for _, cp := range resp.CommonPrefixes {
allPrefixes = append(allPrefixes, *cp.Prefix)
}
if resp.IsTruncated == nil || !*resp.IsTruncated {
break
}
keyMarker = resp.NextKeyMarker
}
assert.Greater(t, pageCount, 1, "Should require multiple pages with maxKeys=%d", maxKeys)
// Build the exact expected prefixes
expectedPrefixes := make([]string, 0, numMailboxes)
for i := 0; i < numMailboxes; i++ {
expectedPrefixes = append(expectedPrefixes,
fmt.Sprintf("%s/mbx-%03d/", orgPrefix, i))
}
assert.Equal(t, expectedPrefixes, allPrefixes,
"CommonPrefixes should exactly match expected mailbox prefixes")
})
}
// listAllVersions is a helper to list all versions of a specific object using pagination
func listAllVersions(t *testing.T, client *s3.Client, bucketName, objectKey string) []types.ObjectVersion {
var allVersions []types.ObjectVersion
+7 -4
View File
@@ -115,10 +115,13 @@ func NewTestEnvironment(t *testing.T) *TestEnvironment {
bindIP := testutil.FindBindIP()
masterPort, masterGrpcPort := testutil.MustFreePortPair(t, "Master")
volumePort, volumeGrpcPort := testutil.MustFreePortPair(t, "Volume")
filerPort, filerGrpcPort := testutil.MustFreePortPair(t, "Filer")
s3Port, s3GrpcPort := testutil.MustFreePortPair(t, "S3")
// Allocate all ports in a single batch to prevent the OS from recycling
// a released port, which can cause two services to get the same port.
ports := testutil.MustAllocatePorts(t, 8)
masterPort, masterGrpcPort := ports[0], ports[1]
volumePort, volumeGrpcPort := ports[2], ports[3]
filerPort, filerGrpcPort := ports[4], ports[5]
s3Port, s3GrpcPort := ports[6], ports[7]
return &TestEnvironment{
seaweedDir: seaweedDir,
+8 -5
View File
@@ -89,11 +89,14 @@ func NewTestEnvironment(t *testing.T) *TestEnvironment {
bindIP := testutil.FindBindIP()
masterPort, masterGrpcPort := testutil.MustFreePortPair(t, "Master")
volumePort, volumeGrpcPort := testutil.MustFreePortPair(t, "Volume")
filerPort, filerGrpcPort := testutil.MustFreePortPair(t, "Filer")
s3Port, s3GrpcPort := testutil.MustFreePortPair(t, "S3")
polarisPort, polarisAdminPort := testutil.MustFreePortPair(t, "Polaris")
// Allocate all ports in a single batch to prevent the OS from recycling
// a released port, which can cause two services to get the same port.
ports := testutil.MustAllocatePorts(t, 10)
masterPort, masterGrpcPort := ports[0], ports[1]
volumePort, volumeGrpcPort := ports[2], ports[3]
filerPort, filerGrpcPort := ports[4], ports[5]
s3Port, s3GrpcPort := ports[6], ports[7]
polarisPort, polarisAdminPort := ports[8], ports[9]
return &TestEnvironment{
seaweedDir: seaweedDir,
@@ -93,10 +93,13 @@ func NewTestEnvironment(t *testing.T) *TestEnvironment {
bindIP := testutil.FindBindIP()
masterPort, masterGrpcPort := testutil.MustFreePortPair(t, "Master")
volumePort, volumeGrpcPort := testutil.MustFreePortPair(t, "Volume")
filerPort, filerGrpcPort := testutil.MustFreePortPair(t, "Filer")
s3Port, s3GrpcPort := testutil.MustFreePortPair(t, "S3") // Changed to use testutil.MustFreePortPair
// Allocate all ports in a single batch to prevent the OS from recycling
// a released port, which can cause two services to get the same port.
ports := testutil.MustAllocatePorts(t, 8)
masterPort, masterGrpcPort := ports[0], ports[1]
volumePort, volumeGrpcPort := ports[2], ports[3]
filerPort, filerGrpcPort := ports[4], ports[5]
s3Port, s3GrpcPort := ports[6], ports[7]
return &TestEnvironment{
seaweedDir: seaweedDir,
+30 -19
View File
@@ -15,33 +15,44 @@ func HasDocker() bool {
return cmd.Run() == nil
}
// MustFreePortPair is a convenience wrapper for tests that only need a single pair.
// Prefer MustAllocatePorts when allocating multiple pairs to guarantee uniqueness.
func MustFreePortPair(t *testing.T, name string) (int, int) {
httpPort, grpcPort, err := findAvailablePortPair()
if err != nil {
t.Fatalf("Failed to get free port pair for %s: %v", name, err)
}
return httpPort, grpcPort
ports := MustAllocatePorts(t, 2)
return ports[0], ports[1]
}
func findAvailablePortPair() (int, int, error) {
httpPort, err := GetFreePort()
// MustAllocatePorts allocates count unique free ports atomically.
// All listeners are held open until every port is obtained, preventing
// the OS from recycling a port between successive allocations.
func MustAllocatePorts(t *testing.T, count int) []int {
t.Helper()
ports, err := AllocatePorts(count)
if err != nil {
return 0, 0, err
t.Fatalf("Failed to allocate %d free ports: %v", count, err)
}
grpcPort, err := GetFreePort()
if err != nil {
return 0, 0, err
}
return httpPort, grpcPort, nil
return ports
}
func GetFreePort() (int, error) {
listener, err := net.Listen("tcp", "0.0.0.0:0")
if err != nil {
return 0, err
// AllocatePorts allocates count unique free ports atomically.
func AllocatePorts(count int) ([]int, error) {
listeners := make([]net.Listener, 0, count)
ports := make([]int, 0, count)
for i := 0; i < count; i++ {
l, err := net.Listen("tcp", "0.0.0.0:0")
if err != nil {
for _, ll := range listeners {
_ = ll.Close()
}
return nil, err
}
listeners = append(listeners, l)
ports = append(ports, l.Addr().(*net.TCPAddr).Port)
}
defer listener.Close()
return listener.Addr().(*net.TCPAddr).Port, nil
for _, l := range listeners {
_ = l.Close()
}
return ports, nil
}
func WaitForService(url string, timeout time.Duration) bool {
@@ -0,0 +1,368 @@
package framework
import (
"fmt"
"net"
"os"
"os/exec"
"path/filepath"
"strconv"
"sync"
"testing"
"github.com/seaweedfs/seaweedfs/test/volume_server/matrix"
)
// MixedVolumeCluster wraps a Go master + a mix of Go and Rust volume servers
// for integration testing. Go servers occupy indices [0, goCount) and Rust
// servers occupy indices [goCount, goCount+rustCount).
type MixedVolumeCluster struct {
testingTB testing.TB
profile matrix.Profile
weedBinary string // Go weed binary (master + Go volume servers)
rustVolumeBinary string // Rust volume binary
baseDir string
configDir string
logsDir string
keepLogs bool
masterPort int
masterGrpcPort int
volumePorts []int
volumeGrpcPorts []int
volumePubPorts []int
isRust []bool // which servers are Rust
masterCmd *exec.Cmd
volumeCmds []*exec.Cmd
cleanupOnce sync.Once
}
// StartMixedVolumeCluster starts a cluster with 1 Go master, goCount Go volume
// servers, and rustCount Rust volume servers. Go servers come first in the index.
func StartMixedVolumeCluster(t testing.TB, profile matrix.Profile, goCount, rustCount int) *MixedVolumeCluster {
t.Helper()
if goCount < 0 || rustCount < 0 {
t.Fatalf("goCount and rustCount must be non-negative, got go=%d rust=%d", goCount, rustCount)
}
total := goCount + rustCount
if total < 2 {
t.Fatalf("need at least 2 volume servers, got %d", total)
}
weedBinary, err := FindOrBuildWeedBinary()
if err != nil {
t.Fatalf("resolve weed binary: %v", err)
}
// Only build the Rust binary when Rust servers are requested.
var rustBinary string
if rustCount > 0 {
rustBinary, err = FindOrBuildRustBinary()
if err != nil {
t.Skipf("skipping mixed cluster test: rust binary unavailable: %v", err)
}
}
baseDir, keepLogs, err := newWorkDir()
if err != nil {
t.Fatalf("create temp test directory: %v", err)
}
configDir := filepath.Join(baseDir, "config")
logsDir := filepath.Join(baseDir, "logs")
masterDataDir := filepath.Join(baseDir, "master")
dirs := []string{configDir, logsDir, masterDataDir}
for i := 0; i < total; i++ {
dirs = append(dirs, filepath.Join(baseDir, fmt.Sprintf("volume%d", i)))
}
for _, dir := range dirs {
if mkErr := os.MkdirAll(dir, 0o755); mkErr != nil {
t.Fatalf("create %s: %v", dir, mkErr)
}
}
if err = writeSecurityConfig(configDir, profile); err != nil {
t.Fatalf("write security config: %v", err)
}
masterPort, masterGrpcPort, err := allocateMasterPortPair()
if err != nil {
t.Fatalf("allocate master port pair: %v", err)
}
// 2 ports per server (admin, grpc); add 1 more when public port is split out.
portsPerServer := 2
if profile.SplitPublicPort {
portsPerServer = 3
}
ports, err := allocatePorts(total * portsPerServer)
if err != nil {
t.Fatalf("allocate volume ports: %v", err)
}
isRust := make([]bool, total)
for i := goCount; i < total; i++ {
isRust[i] = true
}
c := &MixedVolumeCluster{
testingTB: t,
profile: profile,
weedBinary: weedBinary,
rustVolumeBinary: rustBinary,
baseDir: baseDir,
configDir: configDir,
logsDir: logsDir,
keepLogs: keepLogs,
masterPort: masterPort,
masterGrpcPort: masterGrpcPort,
volumePorts: make([]int, total),
volumeGrpcPorts: make([]int, total),
volumePubPorts: make([]int, total),
isRust: isRust,
volumeCmds: make([]*exec.Cmd, total),
}
for i := 0; i < total; i++ {
baseIdx := i * portsPerServer
c.volumePorts[i] = ports[baseIdx]
c.volumeGrpcPorts[i] = ports[baseIdx+1]
if profile.SplitPublicPort {
c.volumePubPorts[i] = ports[baseIdx+2]
} else {
c.volumePubPorts[i] = c.volumePorts[i] // reuse admin port
}
}
// Start master
if err = c.startMaster(masterDataDir); err != nil {
c.Stop()
t.Fatalf("start master: %v", err)
}
helper := &Cluster{logsDir: logsDir}
if err = helper.waitForHTTP(c.MasterURL() + "/dir/status"); err != nil {
masterLog := helper.tailLog("master.log")
c.Stop()
t.Fatalf("wait for master readiness: %v\nmaster log tail:\n%s", err, masterLog)
}
// Start volume servers
for i := 0; i < total; i++ {
volumeDataDir := filepath.Join(baseDir, fmt.Sprintf("volume%d", i))
if isRust[i] {
err = c.startRustVolume(i, volumeDataDir)
} else {
err = c.startGoVolume(i, volumeDataDir)
}
if err != nil {
logTail := helper.tailLog(fmt.Sprintf("volume%d.log", i))
c.Stop()
t.Fatalf("start volume server %d (rust=%v): %v\nlog tail:\n%s", i, isRust[i], err, logTail)
}
// Rust uses /healthz, Go uses /status
healthURL := c.VolumeAdminURL(i) + "/status"
if isRust[i] {
healthURL = c.VolumeAdminURL(i) + "/healthz"
}
if err = helper.waitForHTTP(healthURL); err != nil {
logTail := helper.tailLog(fmt.Sprintf("volume%d.log", i))
c.Stop()
t.Fatalf("wait for volume server %d readiness: %v\nlog tail:\n%s", i, err, logTail)
}
if err = helper.waitForTCP(c.VolumeGRPCAddress(i)); err != nil {
logTail := helper.tailLog(fmt.Sprintf("volume%d.log", i))
c.Stop()
t.Fatalf("wait for volume server %d grpc readiness: %v\nlog tail:\n%s", i, err, logTail)
}
}
t.Cleanup(func() {
c.Stop()
})
return c
}
func (c *MixedVolumeCluster) Stop() {
if c == nil {
return
}
c.cleanupOnce.Do(func() {
for i := len(c.volumeCmds) - 1; i >= 0; i-- {
stopProcess(c.volumeCmds[i])
}
stopProcess(c.masterCmd)
if !c.keepLogs && !c.testingTB.Failed() {
_ = os.RemoveAll(c.baseDir)
} else if c.baseDir != "" {
c.testingTB.Logf("mixed volume server integration logs kept at %s", c.baseDir)
}
})
}
func (c *MixedVolumeCluster) startMaster(dataDir string) error {
logFile, err := os.Create(filepath.Join(c.logsDir, "master.log"))
if err != nil {
return err
}
args := []string{
"-config_dir=" + c.configDir,
"master",
"-ip=127.0.0.1",
"-port=" + strconv.Itoa(c.masterPort),
"-port.grpc=" + strconv.Itoa(c.masterGrpcPort),
"-mdir=" + dataDir,
"-peers=none",
"-volumeSizeLimitMB=" + strconv.Itoa(testVolumeSizeLimitMB),
"-defaultReplication=000",
}
c.masterCmd = exec.Command(c.weedBinary, args...)
c.masterCmd.Dir = c.baseDir
c.masterCmd.Stdout = logFile
c.masterCmd.Stderr = logFile
if err = c.masterCmd.Start(); err != nil {
logFile.Close()
return err
}
logFile.Close() // child inherited the fd
return nil
}
func (c *MixedVolumeCluster) startGoVolume(index int, dataDir string) error {
logName := fmt.Sprintf("volume%d.log", index)
logFile, err := os.Create(filepath.Join(c.logsDir, logName))
if err != nil {
return err
}
args := []string{
"-config_dir=" + c.configDir,
"volume",
"-ip=127.0.0.1",
"-port=" + strconv.Itoa(c.volumePorts[index]),
"-port.grpc=" + strconv.Itoa(c.volumeGrpcPorts[index]),
"-port.public=" + strconv.Itoa(c.volumePubPorts[index]),
"-dir=" + dataDir,
"-max=16",
"-master=127.0.0.1:" + strconv.Itoa(c.masterPort),
"-readMode=" + c.profile.ReadMode,
"-concurrentUploadLimitMB=" + strconv.Itoa(c.profile.ConcurrentUploadLimitMB),
"-concurrentDownloadLimitMB=" + strconv.Itoa(c.profile.ConcurrentDownloadLimitMB),
}
if c.profile.InflightUploadTimeout > 0 {
args = append(args, "-inflightUploadDataTimeout="+c.profile.InflightUploadTimeout.String())
}
if c.profile.InflightDownloadTimeout > 0 {
args = append(args, "-inflightDownloadDataTimeout="+c.profile.InflightDownloadTimeout.String())
}
cmd := exec.Command(c.weedBinary, args...)
cmd.Dir = c.baseDir
cmd.Stdout = logFile
cmd.Stderr = logFile
if err = cmd.Start(); err != nil {
logFile.Close()
return err
}
logFile.Close() // child inherited the fd
c.volumeCmds[index] = cmd
return nil
}
func (c *MixedVolumeCluster) startRustVolume(index int, dataDir string) error {
logName := fmt.Sprintf("volume%d.log", index)
logFile, err := os.Create(filepath.Join(c.logsDir, logName))
if err != nil {
return err
}
args := rustVolumeArgs(
c.profile,
c.configDir,
c.masterPort,
c.volumePorts[index],
c.volumeGrpcPorts[index],
c.volumePubPorts[index],
dataDir,
)
cmd := exec.Command(c.rustVolumeBinary, args...)
cmd.Dir = c.baseDir
cmd.Stdout = logFile
cmd.Stderr = logFile
if err = cmd.Start(); err != nil {
logFile.Close()
return err
}
logFile.Close() // child inherited the fd
c.volumeCmds[index] = cmd
return nil
}
// --- accessor methods (mirror MultiVolumeCluster) ---
func (c *MixedVolumeCluster) MasterAddress() string {
return net.JoinHostPort("127.0.0.1", strconv.Itoa(c.masterPort))
}
func (c *MixedVolumeCluster) MasterURL() string {
return "http://" + c.MasterAddress()
}
func (c *MixedVolumeCluster) VolumeAdminAddress(index int) string {
if index < 0 || index >= len(c.volumePorts) {
return ""
}
return net.JoinHostPort("127.0.0.1", strconv.Itoa(c.volumePorts[index]))
}
func (c *MixedVolumeCluster) VolumePublicAddress(index int) string {
if index < 0 || index >= len(c.volumePubPorts) {
return ""
}
return net.JoinHostPort("127.0.0.1", strconv.Itoa(c.volumePubPorts[index]))
}
func (c *MixedVolumeCluster) VolumeGRPCAddress(index int) string {
if index < 0 || index >= len(c.volumeGrpcPorts) {
return ""
}
return net.JoinHostPort("127.0.0.1", strconv.Itoa(c.volumeGrpcPorts[index]))
}
func (c *MixedVolumeCluster) VolumeAdminURL(index int) string {
return "http://" + c.VolumeAdminAddress(index)
}
func (c *MixedVolumeCluster) VolumePublicURL(index int) string {
return "http://" + c.VolumePublicAddress(index)
}
func (c *MixedVolumeCluster) BaseDir() string {
return c.baseDir
}
// VolumeServerAddress returns SeaweedFS server address format: ip:httpPort.grpcPort
func (c *MixedVolumeCluster) VolumeServerAddress(index int) string {
if index < 0 || index >= len(c.volumePorts) {
return ""
}
return fmt.Sprintf("%s.%d", c.VolumeAdminAddress(index), c.volumeGrpcPorts[index])
}
func (c *MixedVolumeCluster) IsRust(index int) bool {
if index < 0 || index >= len(c.isRust) {
return false
}
return c.isRust[index]
}
@@ -0,0 +1,567 @@
package volume_server_grpc_test
import (
"bytes"
"context"
"fmt"
"io"
"net/http"
"testing"
"time"
"github.com/seaweedfs/seaweedfs/test/volume_server/framework"
"github.com/seaweedfs/seaweedfs/test/volume_server/matrix"
"github.com/seaweedfs/seaweedfs/weed/pb/volume_server_pb"
)
// deleteAndWaitForTombstone issues an HTTP DELETE for the given fid on
// volumeURL, asserts a successful response, then polls with GET until
// the file returns 404 (tombstone visible) or the timeout elapses.
func deleteAndWaitForTombstone(t *testing.T, httpClient *http.Client, volumeURL, fid string) {
t.Helper()
req, err := http.NewRequest(http.MethodDelete, fmt.Sprintf("%s/%s", volumeURL, fid), nil)
if err != nil {
t.Fatalf("build delete request for %s: %v", fid, err)
}
resp := framework.DoRequest(t, httpClient, req)
framework.ReadAllAndClose(t, resp)
if resp.StatusCode != http.StatusAccepted && resp.StatusCode != http.StatusOK {
t.Fatalf("delete %s: expected 200 or 202, got %d", fid, resp.StatusCode)
}
// Poll until GET returns 404 (tombstone flushed to disk).
deadline := time.Now().Add(5 * time.Second)
for time.Now().Before(deadline) {
getResp := framework.ReadBytes(t, httpClient, volumeURL, fid)
status := getResp.StatusCode
framework.ReadAllAndClose(t, getResp)
if status == http.StatusNotFound {
return
}
time.Sleep(50 * time.Millisecond)
}
t.Fatalf("delete %s: tombstone not visible after 5s", fid)
}
// TestMixedBalanceCopyGoToRust verifies that VolumeCopy works from a Go volume
// server to a Rust volume server. This is the core operation behind volume
// balancing in a mixed Go+Rust cluster.
func TestMixedBalanceCopyGoToRust(t *testing.T) {
if testing.Short() {
t.Skip("skipping integration test in short mode")
}
cluster := framework.StartMixedVolumeCluster(t, matrix.P1(), 1, 1)
// server 0 = Go, server 1 = Rust
conn0, goClient := framework.DialVolumeServer(t, cluster.VolumeGRPCAddress(0))
defer conn0.Close()
conn1, rustClient := framework.DialVolumeServer(t, cluster.VolumeGRPCAddress(1))
defer conn1.Close()
httpClient := framework.NewHTTPClient()
const volumeID = uint32(50)
// Allocate volume on Go server and upload test data
framework.AllocateVolume(t, goClient, volumeID, "")
testFiles := []struct {
key uint64
cookie uint32
data []byte
}{
{1, 0xAABBCCDD, []byte("hello from Go server")},
{2, 0x11223344, []byte("second file for balance test")},
{3, 0xDEADBEEF, make([]byte, 4096)}, // larger file
}
for _, f := range testFiles {
fid := framework.NewFileID(volumeID, f.key, f.cookie)
resp := framework.UploadBytes(t, httpClient, cluster.VolumeAdminURL(0), fid, f.data)
body := framework.ReadAllAndClose(t, resp)
if resp.StatusCode != http.StatusCreated {
t.Fatalf("upload %s: expected 201, got %d: %s", fid, resp.StatusCode, body)
}
}
ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
defer cancel()
// Read source volume status before copy
sourceStatus, err := goClient.ReadVolumeFileStatus(ctx, &volume_server_pb.ReadVolumeFileStatusRequest{
VolumeId: volumeID,
})
if err != nil {
t.Fatalf("ReadVolumeFileStatus on Go server: %v", err)
}
t.Logf("Source: dat=%d idx=%d files=%d version=%d",
sourceStatus.GetDatFileSize(), sourceStatus.GetIdxFileSize(),
sourceStatus.GetFileCount(), sourceStatus.GetVersion())
// Copy volume from Go (server 0) to Rust (server 1)
copyStream, err := rustClient.VolumeCopy(ctx, &volume_server_pb.VolumeCopyRequest{
VolumeId: volumeID,
SourceDataNode: cluster.VolumeServerAddress(0),
})
if err != nil {
t.Fatalf("VolumeCopy call failed: %v", err)
}
var lastAppendAtNs uint64
for {
resp, recvErr := copyStream.Recv()
if recvErr != nil {
if recvErr != io.EOF {
t.Fatalf("VolumeCopy recv error: %v", recvErr)
}
break
}
if resp.GetLastAppendAtNs() != 0 {
lastAppendAtNs = resp.GetLastAppendAtNs()
}
}
t.Logf("VolumeCopy completed, lastAppendAtNs=%d", lastAppendAtNs)
// Verify: read volume status from Rust server
targetStatus, err := rustClient.ReadVolumeFileStatus(ctx, &volume_server_pb.ReadVolumeFileStatusRequest{
VolumeId: volumeID,
})
if err != nil {
t.Fatalf("ReadVolumeFileStatus on Rust server: %v", err)
}
t.Logf("Target: dat=%d idx=%d files=%d version=%d",
targetStatus.GetDatFileSize(), targetStatus.GetIdxFileSize(),
targetStatus.GetFileCount(), targetStatus.GetVersion())
if sourceStatus.GetDatFileSize() != targetStatus.GetDatFileSize() {
t.Fatalf("dat file size mismatch: source=%d target=%d",
sourceStatus.GetDatFileSize(), targetStatus.GetDatFileSize())
}
if sourceStatus.GetIdxFileSize() != targetStatus.GetIdxFileSize() {
t.Fatalf("idx file size mismatch: source=%d target=%d",
sourceStatus.GetIdxFileSize(), targetStatus.GetIdxFileSize())
}
if sourceStatus.GetFileCount() != targetStatus.GetFileCount() {
t.Fatalf("file count mismatch: source=%d target=%d",
sourceStatus.GetFileCount(), targetStatus.GetFileCount())
}
// Verify data can be read from Rust server
for _, f := range testFiles {
fid := framework.NewFileID(volumeID, f.key, f.cookie)
resp := framework.ReadBytes(t, httpClient, cluster.VolumeAdminURL(1), fid)
body := framework.ReadAllAndClose(t, resp)
if resp.StatusCode != http.StatusOK {
t.Fatalf("read %s from Rust server: expected 200, got %d", fid, resp.StatusCode)
}
if !bytes.Equal(body, f.data) {
t.Fatalf("read %s from Rust server: content mismatch (got %d bytes, want %d)", fid, len(body), len(f.data))
}
}
}
// TestMixedBalanceCopyRustToGo verifies that VolumeCopy works from a Rust
// volume server to a Go volume server.
func TestMixedBalanceCopyRustToGo(t *testing.T) {
if testing.Short() {
t.Skip("skipping integration test in short mode")
}
cluster := framework.StartMixedVolumeCluster(t, matrix.P1(), 1, 1)
conn0, goClient := framework.DialVolumeServer(t, cluster.VolumeGRPCAddress(0))
defer conn0.Close()
conn1, rustClient := framework.DialVolumeServer(t, cluster.VolumeGRPCAddress(1))
defer conn1.Close()
httpClient := framework.NewHTTPClient()
const volumeID = uint32(51)
// Allocate volume on Rust server and upload test data
framework.AllocateVolume(t, rustClient, volumeID, "")
testFiles := []struct {
key uint64
cookie uint32
data []byte
}{
{1, 0xAABBCCDD, []byte("hello from Rust server")},
{2, 0x11223344, []byte("second file for reverse balance")},
}
for _, f := range testFiles {
fid := framework.NewFileID(volumeID, f.key, f.cookie)
resp := framework.UploadBytes(t, httpClient, cluster.VolumeAdminURL(1), fid, f.data)
body := framework.ReadAllAndClose(t, resp)
if resp.StatusCode != http.StatusCreated {
t.Fatalf("upload %s: expected 201, got %d: %s", fid, resp.StatusCode, body)
}
}
ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
defer cancel()
sourceStatus, err := rustClient.ReadVolumeFileStatus(ctx, &volume_server_pb.ReadVolumeFileStatusRequest{
VolumeId: volumeID,
})
if err != nil {
t.Fatalf("ReadVolumeFileStatus on Rust server: %v", err)
}
// Copy volume from Rust (server 1) to Go (server 0)
copyStream, err := goClient.VolumeCopy(ctx, &volume_server_pb.VolumeCopyRequest{
VolumeId: volumeID,
SourceDataNode: cluster.VolumeServerAddress(1),
})
if err != nil {
t.Fatalf("VolumeCopy call failed: %v", err)
}
for {
_, recvErr := copyStream.Recv()
if recvErr != nil {
if recvErr != io.EOF {
t.Fatalf("VolumeCopy recv error: %v", recvErr)
}
break
}
}
targetStatus, err := goClient.ReadVolumeFileStatus(ctx, &volume_server_pb.ReadVolumeFileStatusRequest{
VolumeId: volumeID,
})
if err != nil {
t.Fatalf("ReadVolumeFileStatus on Go server: %v", err)
}
if sourceStatus.GetDatFileSize() != targetStatus.GetDatFileSize() {
t.Fatalf("dat file size mismatch: source=%d target=%d",
sourceStatus.GetDatFileSize(), targetStatus.GetDatFileSize())
}
if sourceStatus.GetIdxFileSize() != targetStatus.GetIdxFileSize() {
t.Fatalf("idx file size mismatch: source=%d target=%d",
sourceStatus.GetIdxFileSize(), targetStatus.GetIdxFileSize())
}
if sourceStatus.GetFileCount() != targetStatus.GetFileCount() {
t.Fatalf("file count mismatch: source=%d target=%d",
sourceStatus.GetFileCount(), targetStatus.GetFileCount())
}
// Verify data can be read from Go server
for _, f := range testFiles {
fid := framework.NewFileID(volumeID, f.key, f.cookie)
resp := framework.ReadBytes(t, httpClient, cluster.VolumeAdminURL(0), fid)
body := framework.ReadAllAndClose(t, resp)
if resp.StatusCode != http.StatusOK {
t.Fatalf("read %s from Go server: expected 200, got %d", fid, resp.StatusCode)
}
if !bytes.Equal(body, f.data) {
t.Fatalf("read %s from Go server: content mismatch (got %d bytes, want %d)", fid, len(body), len(f.data))
}
}
}
// TestMixedBalanceCopyWithDeletes verifies that VolumeCopy correctly handles
// volumes that have both active and deleted needles.
func TestMixedBalanceCopyWithDeletes(t *testing.T) {
if testing.Short() {
t.Skip("skipping integration test in short mode")
}
cluster := framework.StartMixedVolumeCluster(t, matrix.P1(), 1, 1)
conn0, goClient := framework.DialVolumeServer(t, cluster.VolumeGRPCAddress(0))
defer conn0.Close()
conn1, rustClient := framework.DialVolumeServer(t, cluster.VolumeGRPCAddress(1))
defer conn1.Close()
httpClient := framework.NewHTTPClient()
const volumeID = uint32(52)
framework.AllocateVolume(t, goClient, volumeID, "")
// Upload files
for i := uint64(1); i <= 5; i++ {
fid := framework.NewFileID(volumeID, i, 0x12340000+uint32(i))
resp := framework.UploadBytes(t, httpClient, cluster.VolumeAdminURL(0), fid, []byte(fmt.Sprintf("file-%d", i)))
framework.ReadAllAndClose(t, resp)
if resp.StatusCode != http.StatusCreated {
t.Fatalf("upload file %d: expected 201, got %d", i, resp.StatusCode)
}
}
// Delete some files and wait for tombstones to be visible
for _, key := range []uint64{2, 4} {
fid := framework.NewFileID(volumeID, key, 0x12340000+uint32(key))
deleteAndWaitForTombstone(t, httpClient, cluster.VolumeAdminURL(0), fid)
}
ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
defer cancel()
sourceStatus, err := goClient.ReadVolumeFileStatus(ctx, &volume_server_pb.ReadVolumeFileStatusRequest{
VolumeId: volumeID,
})
if err != nil {
t.Fatalf("ReadVolumeFileStatus: %v", err)
}
t.Logf("Source after deletes: dat=%d idx=%d files=%d",
sourceStatus.GetDatFileSize(), sourceStatus.GetIdxFileSize(),
sourceStatus.GetFileCount())
// Copy volume from Go to Rust
copyStream, err := rustClient.VolumeCopy(ctx, &volume_server_pb.VolumeCopyRequest{
VolumeId: volumeID,
SourceDataNode: cluster.VolumeServerAddress(0),
})
if err != nil {
t.Fatalf("VolumeCopy: %v", err)
}
var lastAppendAtNs uint64
for {
resp, recvErr := copyStream.Recv()
if recvErr != nil {
if recvErr != io.EOF {
t.Fatalf("VolumeCopy recv error: %v", recvErr)
}
break
}
if resp.GetLastAppendAtNs() != 0 {
lastAppendAtNs = resp.GetLastAppendAtNs()
}
}
if lastAppendAtNs == 0 {
t.Fatalf("VolumeCopy did not return a lastAppendAtNs timestamp")
}
t.Logf("VolumeCopy completed, lastAppendAtNs=%d", lastAppendAtNs)
targetStatusAfterCopy, err := rustClient.ReadVolumeFileStatus(ctx, &volume_server_pb.ReadVolumeFileStatusRequest{
VolumeId: volumeID,
})
if err != nil {
t.Fatalf("ReadVolumeFileStatus on Rust: %v", err)
}
t.Logf("Target after copy: dat=%d idx=%d files=%d",
targetStatusAfterCopy.GetDatFileSize(), targetStatusAfterCopy.GetIdxFileSize(),
targetStatusAfterCopy.GetFileCount())
if sourceStatus.GetDatFileSize() != targetStatusAfterCopy.GetDatFileSize() {
t.Fatalf("dat file size mismatch: source=%d target=%d",
sourceStatus.GetDatFileSize(), targetStatusAfterCopy.GetDatFileSize())
}
if sourceStatus.GetIdxFileSize() != targetStatusAfterCopy.GetIdxFileSize() {
t.Fatalf("idx file size mismatch: source=%d target=%d",
sourceStatus.GetIdxFileSize(), targetStatusAfterCopy.GetIdxFileSize())
}
// Tail from the copy checkpoint — source is unchanged (deletes happened
// before copy), so tailing should not append any data.
_, err = rustClient.VolumeTailReceiver(ctx, &volume_server_pb.VolumeTailReceiverRequest{
VolumeId: volumeID,
SinceNs: lastAppendAtNs,
IdleTimeoutSeconds: 3,
SourceVolumeServer: cluster.VolumeServerAddress(0),
})
if err != nil {
t.Fatalf("VolumeTailReceiver: %v", err)
}
targetStatusAfterTail, err := rustClient.ReadVolumeFileStatus(ctx, &volume_server_pb.ReadVolumeFileStatusRequest{
VolumeId: volumeID,
})
if err != nil {
t.Fatalf("ReadVolumeFileStatus after tail: %v", err)
}
if targetStatusAfterTail.GetDatFileSize() != targetStatusAfterCopy.GetDatFileSize() {
t.Fatalf("dat grew after tail: before=%d after=%d",
targetStatusAfterCopy.GetDatFileSize(), targetStatusAfterTail.GetDatFileSize())
}
if targetStatusAfterTail.GetIdxFileSize() != targetStatusAfterCopy.GetIdxFileSize() {
t.Fatalf("idx grew after tail: before=%d after=%d",
targetStatusAfterCopy.GetIdxFileSize(), targetStatusAfterTail.GetIdxFileSize())
}
// Verify surviving files are readable from Rust
for _, key := range []uint64{1, 3, 5} {
fid := framework.NewFileID(volumeID, key, 0x12340000+uint32(key))
resp := framework.ReadBytes(t, httpClient, cluster.VolumeAdminURL(1), fid)
body := framework.ReadAllAndClose(t, resp)
if resp.StatusCode != http.StatusOK {
t.Fatalf("read surviving file %d from Rust: expected 200, got %d", key, resp.StatusCode)
}
expected := fmt.Sprintf("file-%d", key)
if string(body) != expected {
t.Fatalf("data mismatch for file %d: got %q want %q", key, body, expected)
}
}
// Verify deleted files return 404
for _, key := range []uint64{2, 4} {
fid := framework.NewFileID(volumeID, key, 0x12340000+uint32(key))
resp := framework.ReadBytes(t, httpClient, cluster.VolumeAdminURL(1), fid)
framework.ReadAllAndClose(t, resp)
if resp.StatusCode != http.StatusNotFound {
t.Fatalf("read deleted file %d from Rust: expected 404, got %d", key, resp.StatusCode)
}
}
}
// TestMixedBalanceFullMoveGoToRust exercises the complete volume balance move
// flow: mark readonly → copy → tail → verify sizes → delete source.
// This mirrors the steps in balance_task.go Execute().
func TestMixedBalanceFullMoveGoToRust(t *testing.T) {
if testing.Short() {
t.Skip("skipping integration test in short mode")
}
cluster := framework.StartMixedVolumeCluster(t, matrix.P1(), 1, 1)
conn0, goClient := framework.DialVolumeServer(t, cluster.VolumeGRPCAddress(0))
defer conn0.Close()
conn1, rustClient := framework.DialVolumeServer(t, cluster.VolumeGRPCAddress(1))
defer conn1.Close()
httpClient := framework.NewHTTPClient()
const volumeID = uint32(53)
framework.AllocateVolume(t, goClient, volumeID, "")
// Upload test data
for i := uint64(1); i <= 5; i++ {
fid := framework.NewFileID(volumeID, i, 0xABCD0000+uint32(i))
resp := framework.UploadBytes(t, httpClient, cluster.VolumeAdminURL(0), fid, []byte(fmt.Sprintf("balance-move-file-%d", i)))
framework.ReadAllAndClose(t, resp)
if resp.StatusCode != http.StatusCreated {
t.Fatalf("upload file %d: expected 201, got %d", i, resp.StatusCode)
}
}
ctx, cancel := context.WithTimeout(context.Background(), 60*time.Second)
defer cancel()
// Step 1: Copy volume to Rust server (source stays writable so we can
// delete after copy to exercise the tail tombstone path)
copyStream, err := rustClient.VolumeCopy(ctx, &volume_server_pb.VolumeCopyRequest{
VolumeId: volumeID,
SourceDataNode: cluster.VolumeServerAddress(0),
})
if err != nil {
t.Fatalf("VolumeCopy: %v", err)
}
var lastAppendAtNs uint64
for {
resp, recvErr := copyStream.Recv()
if recvErr != nil {
if recvErr != io.EOF {
t.Fatalf("VolumeCopy recv: %v", recvErr)
}
break
}
if resp.GetLastAppendAtNs() != 0 {
lastAppendAtNs = resp.GetLastAppendAtNs()
}
}
t.Logf("Copy done, lastAppendAtNs=%d", lastAppendAtNs)
// Step 2: Delete file 3 on the source AFTER copy. This creates a
// tombstone needle that the tail step must propagate to the Rust server.
deleteAndWaitForTombstone(t, httpClient, cluster.VolumeAdminURL(0),
framework.NewFileID(volumeID, 3, 0xABCD0003))
// Step 3: Mark source readonly so no further writes arrive during tail
_, err = goClient.VolumeMarkReadonly(ctx, &volume_server_pb.VolumeMarkReadonlyRequest{
VolumeId: volumeID,
})
if err != nil {
t.Fatalf("mark readonly: %v", err)
}
// Read source status (the reference for post-tail verification)
sourceStatus, err := goClient.ReadVolumeFileStatus(ctx, &volume_server_pb.ReadVolumeFileStatusRequest{
VolumeId: volumeID,
})
if err != nil {
t.Fatalf("read source status: %v", err)
}
t.Logf("Source: dat=%d idx=%d files=%d",
sourceStatus.GetDatFileSize(), sourceStatus.GetIdxFileSize(),
sourceStatus.GetFileCount())
// Step 4: Tail for updates — this must pick up the delete tombstone
_, err = rustClient.VolumeTailReceiver(ctx, &volume_server_pb.VolumeTailReceiverRequest{
VolumeId: volumeID,
SinceNs: lastAppendAtNs,
IdleTimeoutSeconds: 5,
SourceVolumeServer: cluster.VolumeServerAddress(0),
})
if err != nil {
t.Fatalf("VolumeTailReceiver: %v", err)
}
// Step 5: Verify dat/idx sizes match after tail.
// We compare file sizes (byte-level correctness) rather than file_count
// because the tail writes the tombstone via the write path (which
// increments file_count) while the source used the delete path (which
// only increments deletion_count). This is consistent with Go behavior.
targetStatus, err := rustClient.ReadVolumeFileStatus(ctx, &volume_server_pb.ReadVolumeFileStatusRequest{
VolumeId: volumeID,
})
if err != nil {
t.Fatalf("read target status: %v", err)
}
t.Logf("Target: dat=%d idx=%d files=%d",
targetStatus.GetDatFileSize(), targetStatus.GetIdxFileSize(),
targetStatus.GetFileCount())
if sourceStatus.GetDatFileSize() != targetStatus.GetDatFileSize() {
t.Fatalf("dat size mismatch after tail: source=%d target=%d",
sourceStatus.GetDatFileSize(), targetStatus.GetDatFileSize())
}
if sourceStatus.GetIdxFileSize() != targetStatus.GetIdxFileSize() {
t.Fatalf("idx size mismatch after tail: source=%d target=%d",
sourceStatus.GetIdxFileSize(), targetStatus.GetIdxFileSize())
}
// Step 6: Delete volume from source
_, err = goClient.VolumeDelete(ctx, &volume_server_pb.VolumeDeleteRequest{
VolumeId: volumeID,
})
if err != nil {
t.Fatalf("delete source volume: %v", err)
}
// Verify source volume is gone
_, err = goClient.ReadVolumeFileStatus(ctx, &volume_server_pb.ReadVolumeFileStatusRequest{
VolumeId: volumeID,
})
if err == nil {
t.Fatalf("expected error reading deleted source volume, got nil")
}
t.Logf("Source volume deleted successfully (error as expected: %v)", err)
// Verify all surviving data is readable from Rust (the new home)
for _, key := range []uint64{1, 2, 4, 5} {
fid := framework.NewFileID(volumeID, key, 0xABCD0000+uint32(key))
resp := framework.ReadBytes(t, httpClient, cluster.VolumeAdminURL(1), fid)
body := framework.ReadAllAndClose(t, resp)
if resp.StatusCode != http.StatusOK {
t.Fatalf("read file %d from Rust after move: expected 200, got %d", key, resp.StatusCode)
}
expected := fmt.Sprintf("balance-move-file-%d", key)
if string(body) != expected {
t.Fatalf("data mismatch for file %d after move: got %q want %q", key, body, expected)
}
}
// Verify deleted file is 404 on Rust (tombstone propagated via tail)
{
fid := framework.NewFileID(volumeID, 3, 0xABCD0003)
resp := framework.ReadBytes(t, httpClient, cluster.VolumeAdminURL(1), fid)
framework.ReadAllAndClose(t, resp)
if resp.StatusCode != http.StatusNotFound {
t.Fatalf("deleted file 3 on Rust: expected 404, got %d", resp.StatusCode)
}
}
t.Logf("Full balance move completed: volume %d moved from Go to Rust, source purged", volumeID)
}
+1
View File
@@ -45,6 +45,7 @@ type ObjectStoreUser struct {
SecretKey string `json:"secret_key"`
Permissions []string `json:"permissions"`
PolicyNames []string `json:"policy_names"`
IsStatic bool `json:"is_static"` // loaded from static config file, not editable
}
type ObjectStoreUsersData struct {
+19
View File
@@ -867,6 +867,24 @@ func (s *AdminServer) DeleteS3Bucket(bucketName string) error {
})
}
// IsStaticUser checks if a user is a static identity by loading the
// configuration from the credential manager and checking the IsStatic flag.
func (s *AdminServer) IsStaticUser(username string) bool {
if s.credentialManager == nil {
return false
}
s3cfg, err := s.credentialManager.LoadConfiguration(context.Background())
if err != nil {
return false
}
for _, ident := range s3cfg.Identities {
if ident.Name == username {
return ident.IsStatic
}
}
return false
}
// GetObjectStoreUsers retrieves object store users from identity.json
func (s *AdminServer) GetObjectStoreUsers(ctx context.Context) ([]ObjectStoreUser, error) {
if s.credentialManager == nil {
@@ -890,6 +908,7 @@ func (s *AdminServer) GetObjectStoreUsers(ctx context.Context) ([]ObjectStoreUse
user := ObjectStoreUser{
Username: identity.Name,
Permissions: identity.Actions,
IsStatic: identity.IsStatic,
}
// Set email from account if available
-5
View File
@@ -447,11 +447,6 @@ type QueueStats = maintenance.QueueStats
type WorkerDetailsData = maintenance.WorkerDetailsData
type WorkerPerformance = maintenance.WorkerPerformance
// GetTaskIcon returns the icon CSS class for a task type from its UI provider
func GetTaskIcon(taskType MaintenanceTaskType) string {
return maintenance.GetTaskIcon(taskType)
}
// Status constants (these are still static)
const (
TaskStatusPending = maintenance.TaskStatusPending
+9 -6
View File
@@ -3,13 +3,13 @@ package dash
import (
"context"
"crypto/rand"
"encoding/base64"
"errors"
"fmt"
"strings"
"time"
"github.com/seaweedfs/seaweedfs/weed/credential"
"github.com/seaweedfs/seaweedfs/weed/iam"
"github.com/seaweedfs/seaweedfs/weed/pb/iam_pb"
)
@@ -175,7 +175,7 @@ func (s *AdminServer) GetObjectStoreUserDetails(username string) (*UserDetails,
ctx := context.Background()
// Get user using credential manager
// Get user using credential manager (resolves static users via filer gRPC)
identity, err := s.credentialManager.GetUser(ctx, username)
if err != nil {
if err == credential.ErrUserNotFound {
@@ -435,10 +435,13 @@ func generateAccessKey() string {
}
func generateSecretKey() string {
// Generate 40-character secret key (AWS standard)
b := make([]byte, 30) // 30 bytes = 40 characters in base64
rand.Read(b)
return base64.StdEncoding.EncodeToString(b)
// Use the IAM helper to generate URL-safe secret keys (no +, / characters)
// that won't break S3 signature authentication
key, err := iam.GenerateSecretAccessKey()
if err != nil {
panic(fmt.Sprintf("failed to generate secret key: %v", err))
}
return key
}
func generateAccountId() string {
+10 -3
View File
@@ -74,9 +74,16 @@ func TestGenerateSecretKey(t *testing.T) {
key1 := generateSecretKey()
key2 := generateSecretKey()
// Check length (base64 encoding of 30 bytes = 40 characters)
if len(key1) != 40 {
t.Errorf("Expected secret key length 40, got %d", len(key1))
// Check length (IAM standard secret key length)
if len(key1) != 42 {
t.Errorf("Expected secret key length 42, got %d", len(key1))
}
// Check that key contains only URL-safe characters (no +, /)
for _, c := range key1 {
if c == '+' || c == '/' || c == '=' {
t.Errorf("Secret key contains non-URL-safe character: %c", c)
}
}
// Check uniqueness
-23
View File
@@ -312,29 +312,6 @@ func (h *ClusterHandlers) ShowClusterFilers(w http.ResponseWriter, r *http.Reque
}
}
// ShowClusterBrokers renders the cluster message brokers page
func (h *ClusterHandlers) ShowClusterBrokers(w http.ResponseWriter, r *http.Request) {
// Get cluster brokers data
brokersData, err := h.adminServer.GetClusterBrokers()
if err != nil {
writeJSONError(w, http.StatusInternalServerError, "Failed to get cluster brokers: "+err.Error())
return
}
username := usernameOrDefault(r)
brokersData.Username = username
// Render HTML template
w.Header().Set("Content-Type", "text/html")
brokersComponent := app.ClusterBrokers(*brokersData)
viewCtx := layout.NewViewContext(r, username, dash.CSRFTokenFromContext(r.Context()))
layoutComponent := layout.Layout(viewCtx, brokersComponent)
if err := layoutComponent.Render(r.Context(), w); err != nil {
writeJSONError(w, http.StatusInternalServerError, "Failed to render template: "+err.Error())
return
}
}
// GetClusterTopology returns the cluster topology as JSON
func (h *ClusterHandlers) GetClusterTopology(w http.ResponseWriter, r *http.Request) {
topology, err := h.adminServer.GetClusterTopology()
-28
View File
@@ -78,34 +78,6 @@ func (h *MessageQueueHandlers) ShowTopics(w http.ResponseWriter, r *http.Request
}
}
// ShowSubscribers renders the message queue subscribers page
func (h *MessageQueueHandlers) ShowSubscribers(w http.ResponseWriter, r *http.Request) {
// Get subscribers data
subscribersData, err := h.adminServer.GetSubscribers()
if err != nil {
writeJSONError(w, http.StatusInternalServerError, "Failed to get subscribers: "+err.Error())
return
}
// Set username
username := dash.UsernameFromContext(r.Context())
if username == "" {
username = "admin"
}
subscribersData.Username = username
// Render HTML template
w.Header().Set("Content-Type", "text/html")
subscribersComponent := app.Subscribers(*subscribersData)
viewCtx := layout.NewViewContext(r, username, dash.CSRFTokenFromContext(r.Context()))
layoutComponent := layout.Layout(viewCtx, subscribersComponent)
err = layoutComponent.Render(r.Context(), w)
if err != nil {
writeJSONError(w, http.StatusInternalServerError, "Failed to render template: "+err.Error())
return
}
}
// ShowTopicDetails renders the topic details page
func (h *MessageQueueHandlers) ShowTopicDetails(w http.ResponseWriter, r *http.Request) {
// Get topic parameters from URL
+10
View File
@@ -93,6 +93,11 @@ func (h *UserHandlers) UpdateUser(w http.ResponseWriter, r *http.Request) {
return
}
if h.adminServer.IsStaticUser(username) {
writeJSONError(w, http.StatusForbidden, "Cannot modify static user "+username+" (loaded from config file)")
return
}
var req dash.UpdateUserRequest
if err := decodeJSONBody(newJSONMaxReader(w, r), &req); err != nil {
writeJSONError(w, http.StatusBadRequest, "Invalid request: "+err.Error())
@@ -120,6 +125,11 @@ func (h *UserHandlers) DeleteUser(w http.ResponseWriter, r *http.Request) {
return
}
if h.adminServer.IsStaticUser(username) {
writeJSONError(w, http.StatusForbidden, "Cannot delete static user "+username+" (loaded from config file)")
return
}
err := h.adminServer.DeleteObjectStoreUser(username)
if err != nil {
glog.Errorf("Failed to delete user %s: %v", username, err)
@@ -1,124 +0,0 @@
package maintenance
import (
"fmt"
"github.com/seaweedfs/seaweedfs/weed/pb/worker_pb"
)
// VerifyProtobufConfig demonstrates that the protobuf configuration system is working
func VerifyProtobufConfig() error {
// Create configuration manager
configManager := NewMaintenanceConfigManager()
config := configManager.GetConfig()
// Verify basic configuration
if !config.Enabled {
return fmt.Errorf("expected config to be enabled by default")
}
if config.ScanIntervalSeconds != 30*60 {
return fmt.Errorf("expected scan interval to be 1800 seconds, got %d", config.ScanIntervalSeconds)
}
// Verify policy configuration
if config.Policy == nil {
return fmt.Errorf("expected policy to be configured")
}
if config.Policy.GlobalMaxConcurrent != 4 {
return fmt.Errorf("expected global max concurrent to be 4, got %d", config.Policy.GlobalMaxConcurrent)
}
// Verify task policies
vacuumPolicy := config.Policy.TaskPolicies["vacuum"]
if vacuumPolicy == nil {
return fmt.Errorf("expected vacuum policy to be configured")
}
if !vacuumPolicy.Enabled {
return fmt.Errorf("expected vacuum policy to be enabled")
}
// Verify typed configuration access
vacuumConfig := vacuumPolicy.GetVacuumConfig()
if vacuumConfig == nil {
return fmt.Errorf("expected vacuum config to be accessible")
}
if vacuumConfig.GarbageThreshold != 0.3 {
return fmt.Errorf("expected garbage threshold to be 0.3, got %f", vacuumConfig.GarbageThreshold)
}
// Verify helper functions work
if !IsTaskEnabled(config.Policy, "vacuum") {
return fmt.Errorf("expected vacuum task to be enabled via helper function")
}
maxConcurrent := GetMaxConcurrent(config.Policy, "vacuum")
if maxConcurrent != 2 {
return fmt.Errorf("expected vacuum max concurrent to be 2, got %d", maxConcurrent)
}
// Verify erasure coding configuration
ecPolicy := config.Policy.TaskPolicies["erasure_coding"]
if ecPolicy == nil {
return fmt.Errorf("expected EC policy to be configured")
}
ecConfig := ecPolicy.GetErasureCodingConfig()
if ecConfig == nil {
return fmt.Errorf("expected EC config to be accessible")
}
// Verify configurable EC fields only
if ecConfig.FullnessRatio <= 0 || ecConfig.FullnessRatio > 1 {
return fmt.Errorf("expected EC config to have valid fullness ratio (0-1), got %f", ecConfig.FullnessRatio)
}
return nil
}
// GetProtobufConfigSummary returns a summary of the current protobuf configuration
func GetProtobufConfigSummary() string {
configManager := NewMaintenanceConfigManager()
config := configManager.GetConfig()
summary := fmt.Sprintf("SeaweedFS Protobuf Maintenance Configuration:\n")
summary += fmt.Sprintf(" Enabled: %v\n", config.Enabled)
summary += fmt.Sprintf(" Scan Interval: %d seconds\n", config.ScanIntervalSeconds)
summary += fmt.Sprintf(" Max Retries: %d\n", config.MaxRetries)
summary += fmt.Sprintf(" Global Max Concurrent: %d\n", config.Policy.GlobalMaxConcurrent)
summary += fmt.Sprintf(" Task Policies: %d configured\n", len(config.Policy.TaskPolicies))
for taskType, policy := range config.Policy.TaskPolicies {
summary += fmt.Sprintf(" %s: enabled=%v, max_concurrent=%d\n",
taskType, policy.Enabled, policy.MaxConcurrent)
}
return summary
}
// CreateCustomConfig demonstrates creating a custom protobuf configuration
func CreateCustomConfig() *worker_pb.MaintenanceConfig {
return &worker_pb.MaintenanceConfig{
Enabled: true,
ScanIntervalSeconds: 60 * 60, // 1 hour
MaxRetries: 5,
Policy: &worker_pb.MaintenancePolicy{
GlobalMaxConcurrent: 8,
TaskPolicies: map[string]*worker_pb.TaskPolicy{
"custom_vacuum": {
Enabled: true,
MaxConcurrent: 4,
TaskConfig: &worker_pb.TaskPolicy_VacuumConfig{
VacuumConfig: &worker_pb.VacuumTaskConfig{
GarbageThreshold: 0.5,
MinVolumeAgeHours: 48,
},
},
},
},
},
}
}
@@ -1,24 +1,9 @@
package maintenance
import (
"fmt"
"time"
"github.com/seaweedfs/seaweedfs/weed/pb/worker_pb"
)
// MaintenanceConfigManager handles protobuf-based configuration
type MaintenanceConfigManager struct {
config *worker_pb.MaintenanceConfig
}
// NewMaintenanceConfigManager creates a new config manager with defaults
func NewMaintenanceConfigManager() *MaintenanceConfigManager {
return &MaintenanceConfigManager{
config: DefaultMaintenanceConfigProto(),
}
}
// DefaultMaintenanceConfigProto returns default configuration as protobuf
func DefaultMaintenanceConfigProto() *worker_pb.MaintenanceConfig {
return &worker_pb.MaintenanceConfig{
@@ -34,253 +19,3 @@ func DefaultMaintenanceConfigProto() *worker_pb.MaintenanceConfig {
Policy: nil,
}
}
// GetConfig returns the current configuration
func (mcm *MaintenanceConfigManager) GetConfig() *worker_pb.MaintenanceConfig {
return mcm.config
}
// Type-safe configuration accessors
// GetVacuumConfig returns vacuum-specific configuration for a task type
func (mcm *MaintenanceConfigManager) GetVacuumConfig(taskType string) *worker_pb.VacuumTaskConfig {
if policy := mcm.getTaskPolicy(taskType); policy != nil {
if vacuumConfig := policy.GetVacuumConfig(); vacuumConfig != nil {
return vacuumConfig
}
}
// Return defaults if not configured
return &worker_pb.VacuumTaskConfig{
GarbageThreshold: 0.3,
MinVolumeAgeHours: 24,
}
}
// GetErasureCodingConfig returns EC-specific configuration for a task type
func (mcm *MaintenanceConfigManager) GetErasureCodingConfig(taskType string) *worker_pb.ErasureCodingTaskConfig {
if policy := mcm.getTaskPolicy(taskType); policy != nil {
if ecConfig := policy.GetErasureCodingConfig(); ecConfig != nil {
return ecConfig
}
}
// Return defaults if not configured
return &worker_pb.ErasureCodingTaskConfig{
FullnessRatio: 0.95,
QuietForSeconds: 3600,
MinVolumeSizeMb: 100,
CollectionFilter: "",
}
}
// GetBalanceConfig returns balance-specific configuration for a task type
func (mcm *MaintenanceConfigManager) GetBalanceConfig(taskType string) *worker_pb.BalanceTaskConfig {
if policy := mcm.getTaskPolicy(taskType); policy != nil {
if balanceConfig := policy.GetBalanceConfig(); balanceConfig != nil {
return balanceConfig
}
}
// Return defaults if not configured
return &worker_pb.BalanceTaskConfig{
ImbalanceThreshold: 0.2,
MinServerCount: 2,
}
}
// GetReplicationConfig returns replication-specific configuration for a task type
func (mcm *MaintenanceConfigManager) GetReplicationConfig(taskType string) *worker_pb.ReplicationTaskConfig {
if policy := mcm.getTaskPolicy(taskType); policy != nil {
if replicationConfig := policy.GetReplicationConfig(); replicationConfig != nil {
return replicationConfig
}
}
// Return defaults if not configured
return &worker_pb.ReplicationTaskConfig{
TargetReplicaCount: 2,
}
}
// Typed convenience methods for getting task configurations
// GetVacuumTaskConfigForType returns vacuum configuration for a specific task type
func (mcm *MaintenanceConfigManager) GetVacuumTaskConfigForType(taskType string) *worker_pb.VacuumTaskConfig {
return GetVacuumTaskConfig(mcm.config.Policy, MaintenanceTaskType(taskType))
}
// GetErasureCodingTaskConfigForType returns erasure coding configuration for a specific task type
func (mcm *MaintenanceConfigManager) GetErasureCodingTaskConfigForType(taskType string) *worker_pb.ErasureCodingTaskConfig {
return GetErasureCodingTaskConfig(mcm.config.Policy, MaintenanceTaskType(taskType))
}
// GetBalanceTaskConfigForType returns balance configuration for a specific task type
func (mcm *MaintenanceConfigManager) GetBalanceTaskConfigForType(taskType string) *worker_pb.BalanceTaskConfig {
return GetBalanceTaskConfig(mcm.config.Policy, MaintenanceTaskType(taskType))
}
// GetReplicationTaskConfigForType returns replication configuration for a specific task type
func (mcm *MaintenanceConfigManager) GetReplicationTaskConfigForType(taskType string) *worker_pb.ReplicationTaskConfig {
return GetReplicationTaskConfig(mcm.config.Policy, MaintenanceTaskType(taskType))
}
// Helper methods
func (mcm *MaintenanceConfigManager) getTaskPolicy(taskType string) *worker_pb.TaskPolicy {
if mcm.config.Policy != nil && mcm.config.Policy.TaskPolicies != nil {
return mcm.config.Policy.TaskPolicies[taskType]
}
return nil
}
// IsTaskEnabled returns whether a task type is enabled
func (mcm *MaintenanceConfigManager) IsTaskEnabled(taskType string) bool {
if policy := mcm.getTaskPolicy(taskType); policy != nil {
return policy.Enabled
}
return false
}
// GetMaxConcurrent returns the max concurrent limit for a task type
func (mcm *MaintenanceConfigManager) GetMaxConcurrent(taskType string) int32 {
if policy := mcm.getTaskPolicy(taskType); policy != nil {
return policy.MaxConcurrent
}
return 1 // Default
}
// GetRepeatInterval returns the repeat interval for a task type in seconds
func (mcm *MaintenanceConfigManager) GetRepeatInterval(taskType string) int32 {
if policy := mcm.getTaskPolicy(taskType); policy != nil {
return policy.RepeatIntervalSeconds
}
return mcm.config.Policy.DefaultRepeatIntervalSeconds
}
// GetCheckInterval returns the check interval for a task type in seconds
func (mcm *MaintenanceConfigManager) GetCheckInterval(taskType string) int32 {
if policy := mcm.getTaskPolicy(taskType); policy != nil {
return policy.CheckIntervalSeconds
}
return mcm.config.Policy.DefaultCheckIntervalSeconds
}
// Duration accessor methods
// GetScanInterval returns the scan interval as a time.Duration
func (mcm *MaintenanceConfigManager) GetScanInterval() time.Duration {
return time.Duration(mcm.config.ScanIntervalSeconds) * time.Second
}
// GetWorkerTimeout returns the worker timeout as a time.Duration
func (mcm *MaintenanceConfigManager) GetWorkerTimeout() time.Duration {
return time.Duration(mcm.config.WorkerTimeoutSeconds) * time.Second
}
// GetTaskTimeout returns the task timeout as a time.Duration
func (mcm *MaintenanceConfigManager) GetTaskTimeout() time.Duration {
return time.Duration(mcm.config.TaskTimeoutSeconds) * time.Second
}
// GetRetryDelay returns the retry delay as a time.Duration
func (mcm *MaintenanceConfigManager) GetRetryDelay() time.Duration {
return time.Duration(mcm.config.RetryDelaySeconds) * time.Second
}
// GetCleanupInterval returns the cleanup interval as a time.Duration
func (mcm *MaintenanceConfigManager) GetCleanupInterval() time.Duration {
return time.Duration(mcm.config.CleanupIntervalSeconds) * time.Second
}
// GetTaskRetention returns the task retention period as a time.Duration
func (mcm *MaintenanceConfigManager) GetTaskRetention() time.Duration {
return time.Duration(mcm.config.TaskRetentionSeconds) * time.Second
}
// ValidateMaintenanceConfigWithSchema validates protobuf maintenance configuration using ConfigField rules
func ValidateMaintenanceConfigWithSchema(config *worker_pb.MaintenanceConfig) error {
if config == nil {
return fmt.Errorf("configuration cannot be nil")
}
// Get the schema to access field validation rules
schema := GetMaintenanceConfigSchema()
// Validate each field individually using the ConfigField rules
if err := validateFieldWithSchema(schema, "enabled", config.Enabled); err != nil {
return err
}
if err := validateFieldWithSchema(schema, "scan_interval_seconds", int(config.ScanIntervalSeconds)); err != nil {
return err
}
if err := validateFieldWithSchema(schema, "worker_timeout_seconds", int(config.WorkerTimeoutSeconds)); err != nil {
return err
}
if err := validateFieldWithSchema(schema, "task_timeout_seconds", int(config.TaskTimeoutSeconds)); err != nil {
return err
}
if err := validateFieldWithSchema(schema, "retry_delay_seconds", int(config.RetryDelaySeconds)); err != nil {
return err
}
if err := validateFieldWithSchema(schema, "max_retries", int(config.MaxRetries)); err != nil {
return err
}
if err := validateFieldWithSchema(schema, "cleanup_interval_seconds", int(config.CleanupIntervalSeconds)); err != nil {
return err
}
if err := validateFieldWithSchema(schema, "task_retention_seconds", int(config.TaskRetentionSeconds)); err != nil {
return err
}
// Validate policy fields if present
if config.Policy != nil {
// Note: These field names might need to be adjusted based on the actual schema
if err := validatePolicyField("global_max_concurrent", int(config.Policy.GlobalMaxConcurrent)); err != nil {
return err
}
if err := validatePolicyField("default_repeat_interval_seconds", int(config.Policy.DefaultRepeatIntervalSeconds)); err != nil {
return err
}
if err := validatePolicyField("default_check_interval_seconds", int(config.Policy.DefaultCheckIntervalSeconds)); err != nil {
return err
}
}
return nil
}
// validateFieldWithSchema validates a single field using its ConfigField definition
func validateFieldWithSchema(schema *MaintenanceConfigSchema, fieldName string, value interface{}) error {
field := schema.GetFieldByName(fieldName)
if field == nil {
// Field not in schema, skip validation
return nil
}
return field.ValidateValue(value)
}
// validatePolicyField validates policy fields (simplified validation for now)
func validatePolicyField(fieldName string, value int) error {
switch fieldName {
case "global_max_concurrent":
if value < 1 || value > 20 {
return fmt.Errorf("Global Max Concurrent must be between 1 and 20, got %d", value)
}
case "default_repeat_interval":
if value < 1 || value > 168 {
return fmt.Errorf("Default Repeat Interval must be between 1 and 168 hours, got %d", value)
}
case "default_check_interval":
if value < 1 || value > 168 {
return fmt.Errorf("Default Check Interval must be between 1 and 168 hours, got %d", value)
}
}
return nil
}
@@ -1055,28 +1055,6 @@ func (mq *MaintenanceQueue) getMaxConcurrentForTaskType(taskType MaintenanceTask
return 1
}
// getRunningTasks returns all currently running tasks
func (mq *MaintenanceQueue) getRunningTasks() []*MaintenanceTask {
var runningTasks []*MaintenanceTask
for _, task := range mq.tasks {
if task.Status == TaskStatusAssigned || task.Status == TaskStatusInProgress {
runningTasks = append(runningTasks, task)
}
}
return runningTasks
}
// getAvailableWorkers returns all workers that can take more work
func (mq *MaintenanceQueue) getAvailableWorkers() []*MaintenanceWorker {
var availableWorkers []*MaintenanceWorker
for _, worker := range mq.workers {
if worker.Status == "active" && worker.CurrentLoad < worker.MaxConcurrent {
availableWorkers = append(availableWorkers, worker)
}
}
return availableWorkers
}
// trackPendingOperation adds a task to the pending operations tracker
func (mq *MaintenanceQueue) trackPendingOperation(task *MaintenanceTask) {
if mq.integration == nil {
-309
View File
@@ -2,15 +2,11 @@ package maintenance
import (
"html/template"
"sort"
"sync"
"time"
"github.com/seaweedfs/seaweedfs/weed/glog"
"github.com/seaweedfs/seaweedfs/weed/pb/master_pb"
"github.com/seaweedfs/seaweedfs/weed/pb/worker_pb"
"github.com/seaweedfs/seaweedfs/weed/worker/tasks"
"github.com/seaweedfs/seaweedfs/weed/worker/types"
)
// AdminClient interface defines what the maintenance system needs from the admin server
@@ -21,51 +17,6 @@ type AdminClient interface {
// MaintenanceTaskType represents different types of maintenance operations
type MaintenanceTaskType string
// GetRegisteredMaintenanceTaskTypes returns all registered task types as MaintenanceTaskType values
// sorted alphabetically for consistent menu ordering
func GetRegisteredMaintenanceTaskTypes() []MaintenanceTaskType {
typesRegistry := tasks.GetGlobalTypesRegistry()
var taskTypes []MaintenanceTaskType
for workerTaskType := range typesRegistry.GetAllDetectors() {
maintenanceTaskType := MaintenanceTaskType(string(workerTaskType))
taskTypes = append(taskTypes, maintenanceTaskType)
}
// Sort task types alphabetically to ensure consistent menu ordering
sort.Slice(taskTypes, func(i, j int) bool {
return string(taskTypes[i]) < string(taskTypes[j])
})
return taskTypes
}
// GetMaintenanceTaskType returns a specific task type if it's registered, or empty string if not found
func GetMaintenanceTaskType(taskTypeName string) MaintenanceTaskType {
typesRegistry := tasks.GetGlobalTypesRegistry()
for workerTaskType := range typesRegistry.GetAllDetectors() {
if string(workerTaskType) == taskTypeName {
return MaintenanceTaskType(taskTypeName)
}
}
return MaintenanceTaskType("")
}
// IsMaintenanceTaskTypeRegistered checks if a task type is registered
func IsMaintenanceTaskTypeRegistered(taskType MaintenanceTaskType) bool {
typesRegistry := tasks.GetGlobalTypesRegistry()
for workerTaskType := range typesRegistry.GetAllDetectors() {
if string(workerTaskType) == string(taskType) {
return true
}
}
return false
}
// MaintenanceTaskPriority represents task execution priority
type MaintenanceTaskPriority int
@@ -200,14 +151,6 @@ func GetTaskPolicy(mp *MaintenancePolicy, taskType MaintenanceTaskType) *TaskPol
return mp.TaskPolicies[string(taskType)]
}
// SetTaskPolicy sets the policy for a specific task type
func SetTaskPolicy(mp *MaintenancePolicy, taskType MaintenanceTaskType, policy *TaskPolicy) {
if mp.TaskPolicies == nil {
mp.TaskPolicies = make(map[string]*TaskPolicy)
}
mp.TaskPolicies[string(taskType)] = policy
}
// IsTaskEnabled returns whether a task type is enabled
func IsTaskEnabled(mp *MaintenancePolicy, taskType MaintenanceTaskType) bool {
policy := GetTaskPolicy(mp, taskType)
@@ -235,84 +178,6 @@ func GetRepeatInterval(mp *MaintenancePolicy, taskType MaintenanceTaskType) int
return int(policy.RepeatIntervalSeconds)
}
// GetVacuumTaskConfig returns the vacuum task configuration
func GetVacuumTaskConfig(mp *MaintenancePolicy, taskType MaintenanceTaskType) *worker_pb.VacuumTaskConfig {
policy := GetTaskPolicy(mp, taskType)
if policy == nil {
return nil
}
return policy.GetVacuumConfig()
}
// GetErasureCodingTaskConfig returns the erasure coding task configuration
func GetErasureCodingTaskConfig(mp *MaintenancePolicy, taskType MaintenanceTaskType) *worker_pb.ErasureCodingTaskConfig {
policy := GetTaskPolicy(mp, taskType)
if policy == nil {
return nil
}
return policy.GetErasureCodingConfig()
}
// GetBalanceTaskConfig returns the balance task configuration
func GetBalanceTaskConfig(mp *MaintenancePolicy, taskType MaintenanceTaskType) *worker_pb.BalanceTaskConfig {
policy := GetTaskPolicy(mp, taskType)
if policy == nil {
return nil
}
return policy.GetBalanceConfig()
}
// GetReplicationTaskConfig returns the replication task configuration
func GetReplicationTaskConfig(mp *MaintenancePolicy, taskType MaintenanceTaskType) *worker_pb.ReplicationTaskConfig {
policy := GetTaskPolicy(mp, taskType)
if policy == nil {
return nil
}
return policy.GetReplicationConfig()
}
// Note: GetTaskConfig was removed - use typed getters: GetVacuumTaskConfig, GetErasureCodingTaskConfig, GetBalanceTaskConfig, or GetReplicationTaskConfig
// SetVacuumTaskConfig sets the vacuum task configuration
func SetVacuumTaskConfig(mp *MaintenancePolicy, taskType MaintenanceTaskType, config *worker_pb.VacuumTaskConfig) {
policy := GetTaskPolicy(mp, taskType)
if policy != nil {
policy.TaskConfig = &worker_pb.TaskPolicy_VacuumConfig{
VacuumConfig: config,
}
}
}
// SetErasureCodingTaskConfig sets the erasure coding task configuration
func SetErasureCodingTaskConfig(mp *MaintenancePolicy, taskType MaintenanceTaskType, config *worker_pb.ErasureCodingTaskConfig) {
policy := GetTaskPolicy(mp, taskType)
if policy != nil {
policy.TaskConfig = &worker_pb.TaskPolicy_ErasureCodingConfig{
ErasureCodingConfig: config,
}
}
}
// SetBalanceTaskConfig sets the balance task configuration
func SetBalanceTaskConfig(mp *MaintenancePolicy, taskType MaintenanceTaskType, config *worker_pb.BalanceTaskConfig) {
policy := GetTaskPolicy(mp, taskType)
if policy != nil {
policy.TaskConfig = &worker_pb.TaskPolicy_BalanceConfig{
BalanceConfig: config,
}
}
}
// SetReplicationTaskConfig sets the replication task configuration
func SetReplicationTaskConfig(mp *MaintenancePolicy, taskType MaintenanceTaskType, config *worker_pb.ReplicationTaskConfig) {
policy := GetTaskPolicy(mp, taskType)
if policy != nil {
policy.TaskConfig = &worker_pb.TaskPolicy_ReplicationConfig{
ReplicationConfig: config,
}
}
}
// SetTaskConfig sets a configuration value for a task type (legacy method - use typed setters above)
// Note: SetTaskConfig was removed - use typed setters: SetVacuumTaskConfig, SetErasureCodingTaskConfig, SetBalanceTaskConfig, or SetReplicationTaskConfig
@@ -475,180 +340,6 @@ type ClusterReplicationTask struct {
Metadata map[string]string `json:"metadata,omitempty"`
}
// BuildMaintenancePolicyFromTasks creates a maintenance policy with configurations
// from all registered tasks using their UI providers
func BuildMaintenancePolicyFromTasks() *MaintenancePolicy {
policy := &MaintenancePolicy{
TaskPolicies: make(map[string]*TaskPolicy),
GlobalMaxConcurrent: 4,
DefaultRepeatIntervalSeconds: 6 * 3600, // 6 hours in seconds
DefaultCheckIntervalSeconds: 12 * 3600, // 12 hours in seconds
}
// Get all registered task types from the UI registry
uiRegistry := tasks.GetGlobalUIRegistry()
typesRegistry := tasks.GetGlobalTypesRegistry()
for taskType, provider := range uiRegistry.GetAllProviders() {
// Convert task type to maintenance task type
maintenanceTaskType := MaintenanceTaskType(string(taskType))
// Get the default configuration from the UI provider
defaultConfig := provider.GetCurrentConfig()
// Create task policy from UI configuration
taskPolicy := &TaskPolicy{
Enabled: true, // Default enabled
MaxConcurrent: 2, // Default concurrency
RepeatIntervalSeconds: policy.DefaultRepeatIntervalSeconds,
CheckIntervalSeconds: policy.DefaultCheckIntervalSeconds,
}
// Extract configuration using TaskConfig interface - no more map conversions!
if taskConfig, ok := defaultConfig.(interface{ ToTaskPolicy() *worker_pb.TaskPolicy }); ok {
// Use protobuf directly for clean, type-safe config extraction
pbTaskPolicy := taskConfig.ToTaskPolicy()
taskPolicy.Enabled = pbTaskPolicy.Enabled
taskPolicy.MaxConcurrent = pbTaskPolicy.MaxConcurrent
if pbTaskPolicy.RepeatIntervalSeconds > 0 {
taskPolicy.RepeatIntervalSeconds = pbTaskPolicy.RepeatIntervalSeconds
}
if pbTaskPolicy.CheckIntervalSeconds > 0 {
taskPolicy.CheckIntervalSeconds = pbTaskPolicy.CheckIntervalSeconds
}
}
// Also get defaults from scheduler if available (using types.TaskScheduler explicitly)
var scheduler types.TaskScheduler = typesRegistry.GetScheduler(taskType)
if scheduler != nil {
if taskPolicy.MaxConcurrent <= 0 {
taskPolicy.MaxConcurrent = int32(scheduler.GetMaxConcurrent())
}
// Convert default repeat interval to seconds
if repeatInterval := scheduler.GetDefaultRepeatInterval(); repeatInterval > 0 {
taskPolicy.RepeatIntervalSeconds = int32(repeatInterval.Seconds())
}
}
// Also get defaults from detector if available (using types.TaskDetector explicitly)
var detector types.TaskDetector = typesRegistry.GetDetector(taskType)
if detector != nil {
// Convert scan interval to check interval (seconds)
if scanInterval := detector.ScanInterval(); scanInterval > 0 {
taskPolicy.CheckIntervalSeconds = int32(scanInterval.Seconds())
}
}
policy.TaskPolicies[string(maintenanceTaskType)] = taskPolicy
glog.V(3).Infof("Built policy for task type %s: enabled=%v, max_concurrent=%d",
maintenanceTaskType, taskPolicy.Enabled, taskPolicy.MaxConcurrent)
}
glog.V(2).Infof("Built maintenance policy with %d task configurations", len(policy.TaskPolicies))
return policy
}
// SetPolicyFromTasks sets the maintenance policy from registered tasks
func SetPolicyFromTasks(policy *MaintenancePolicy) {
if policy == nil {
return
}
// Build new policy from tasks
newPolicy := BuildMaintenancePolicyFromTasks()
// Copy task policies
policy.TaskPolicies = newPolicy.TaskPolicies
glog.V(1).Infof("Updated maintenance policy with %d task configurations from registered tasks", len(policy.TaskPolicies))
}
// GetTaskIcon returns the icon CSS class for a task type from its UI provider
func GetTaskIcon(taskType MaintenanceTaskType) string {
typesRegistry := tasks.GetGlobalTypesRegistry()
uiRegistry := tasks.GetGlobalUIRegistry()
// Convert MaintenanceTaskType to TaskType
for workerTaskType := range typesRegistry.GetAllDetectors() {
if string(workerTaskType) == string(taskType) {
// Get the UI provider for this task type
provider := uiRegistry.GetProvider(workerTaskType)
if provider != nil {
return provider.GetIcon()
}
break
}
}
// Default icon if no UI provider found
return "fas fa-cog text-muted"
}
// GetTaskDisplayName returns the display name for a task type from its UI provider
func GetTaskDisplayName(taskType MaintenanceTaskType) string {
typesRegistry := tasks.GetGlobalTypesRegistry()
uiRegistry := tasks.GetGlobalUIRegistry()
// Convert MaintenanceTaskType to TaskType
for workerTaskType := range typesRegistry.GetAllDetectors() {
if string(workerTaskType) == string(taskType) {
// Get the UI provider for this task type
provider := uiRegistry.GetProvider(workerTaskType)
if provider != nil {
return provider.GetDisplayName()
}
break
}
}
// Fallback to the task type string
return string(taskType)
}
// GetTaskDescription returns the description for a task type from its UI provider
func GetTaskDescription(taskType MaintenanceTaskType) string {
typesRegistry := tasks.GetGlobalTypesRegistry()
uiRegistry := tasks.GetGlobalUIRegistry()
// Convert MaintenanceTaskType to TaskType
for workerTaskType := range typesRegistry.GetAllDetectors() {
if string(workerTaskType) == string(taskType) {
// Get the UI provider for this task type
provider := uiRegistry.GetProvider(workerTaskType)
if provider != nil {
return provider.GetDescription()
}
break
}
}
// Fallback to a generic description
return "Configure detailed settings for " + string(taskType) + " tasks."
}
// BuildMaintenanceMenuItems creates menu items for all registered task types
func BuildMaintenanceMenuItems() []*MaintenanceMenuItem {
var menuItems []*MaintenanceMenuItem
// Get all registered task types
registeredTypes := GetRegisteredMaintenanceTaskTypes()
for _, taskType := range registeredTypes {
menuItem := &MaintenanceMenuItem{
TaskType: taskType,
DisplayName: GetTaskDisplayName(taskType),
Description: GetTaskDescription(taskType),
Icon: GetTaskIcon(taskType),
IsEnabled: IsMaintenanceTaskTypeRegistered(taskType),
Path: "/maintenance/config/" + string(taskType),
}
menuItems = append(menuItems, menuItem)
}
return menuItems
}
// Helper functions to extract configuration fields
// Note: Removed getVacuumConfigField, getErasureCodingConfigField, getBalanceConfigField, getReplicationConfigField
@@ -1,421 +0,0 @@
package maintenance
import (
"context"
"fmt"
"os"
"sync"
"time"
"github.com/seaweedfs/seaweedfs/weed/glog"
"github.com/seaweedfs/seaweedfs/weed/worker"
"github.com/seaweedfs/seaweedfs/weed/worker/tasks"
"github.com/seaweedfs/seaweedfs/weed/worker/types"
// Import task packages to trigger their auto-registration
_ "github.com/seaweedfs/seaweedfs/weed/worker/tasks/balance"
_ "github.com/seaweedfs/seaweedfs/weed/worker/tasks/erasure_coding"
_ "github.com/seaweedfs/seaweedfs/weed/worker/tasks/vacuum"
)
// MaintenanceWorkerService manages maintenance task execution
// TaskExecutor defines the function signature for task execution
type TaskExecutor func(*MaintenanceWorkerService, *MaintenanceTask) error
// TaskExecutorFactory creates a task executor for a given worker service
type TaskExecutorFactory func() TaskExecutor
// Global registry for task executor factories
var taskExecutorFactories = make(map[MaintenanceTaskType]TaskExecutorFactory)
var executorRegistryMutex sync.RWMutex
var executorRegistryInitOnce sync.Once
// initializeExecutorFactories dynamically registers executor factories for all auto-registered task types
func initializeExecutorFactories() {
executorRegistryInitOnce.Do(func() {
// Get all registered task types from the global registry
typesRegistry := tasks.GetGlobalTypesRegistry()
var taskTypes []MaintenanceTaskType
for workerTaskType := range typesRegistry.GetAllDetectors() {
// Convert types.TaskType to MaintenanceTaskType by string conversion
maintenanceTaskType := MaintenanceTaskType(string(workerTaskType))
taskTypes = append(taskTypes, maintenanceTaskType)
}
// Register generic executor for all task types
for _, taskType := range taskTypes {
RegisterTaskExecutorFactory(taskType, createGenericTaskExecutor)
}
glog.V(1).Infof("Dynamically registered generic task executor for %d task types: %v", len(taskTypes), taskTypes)
})
}
// RegisterTaskExecutorFactory registers a factory function for creating task executors
func RegisterTaskExecutorFactory(taskType MaintenanceTaskType, factory TaskExecutorFactory) {
executorRegistryMutex.Lock()
defer executorRegistryMutex.Unlock()
taskExecutorFactories[taskType] = factory
glog.V(2).Infof("Registered executor factory for task type: %s", taskType)
}
// GetTaskExecutorFactory returns the factory for a task type
func GetTaskExecutorFactory(taskType MaintenanceTaskType) (TaskExecutorFactory, bool) {
// Ensure executor factories are initialized
initializeExecutorFactories()
executorRegistryMutex.RLock()
defer executorRegistryMutex.RUnlock()
factory, exists := taskExecutorFactories[taskType]
return factory, exists
}
// GetSupportedExecutorTaskTypes returns all task types with registered executor factories
func GetSupportedExecutorTaskTypes() []MaintenanceTaskType {
// Ensure executor factories are initialized
initializeExecutorFactories()
executorRegistryMutex.RLock()
defer executorRegistryMutex.RUnlock()
taskTypes := make([]MaintenanceTaskType, 0, len(taskExecutorFactories))
for taskType := range taskExecutorFactories {
taskTypes = append(taskTypes, taskType)
}
return taskTypes
}
// createGenericTaskExecutor creates a generic task executor that uses the task registry
func createGenericTaskExecutor() TaskExecutor {
return func(mws *MaintenanceWorkerService, task *MaintenanceTask) error {
return mws.executeGenericTask(task)
}
}
// init does minimal initialization - actual registration happens lazily
func init() {
// Executor factory registration will happen lazily when first accessed
glog.V(1).Infof("Maintenance worker initialized - executor factories will be registered on first access")
}
type MaintenanceWorkerService struct {
workerID string
address string
adminServer string
capabilities []MaintenanceTaskType
maxConcurrent int
currentTasks map[string]*MaintenanceTask
queue *MaintenanceQueue
adminClient AdminClient
running bool
stopChan chan struct{}
// Task execution registry
taskExecutors map[MaintenanceTaskType]TaskExecutor
// Task registry for creating task instances
taskRegistry *tasks.TaskRegistry
}
// NewMaintenanceWorkerService creates a new maintenance worker service
func NewMaintenanceWorkerService(workerID, address, adminServer string) *MaintenanceWorkerService {
// Get all registered maintenance task types dynamically
capabilities := GetRegisteredMaintenanceTaskTypes()
worker := &MaintenanceWorkerService{
workerID: workerID,
address: address,
adminServer: adminServer,
capabilities: capabilities,
maxConcurrent: 2, // Default concurrent task limit
currentTasks: make(map[string]*MaintenanceTask),
stopChan: make(chan struct{}),
taskExecutors: make(map[MaintenanceTaskType]TaskExecutor),
taskRegistry: tasks.GetGlobalTaskRegistry(), // Use global registry with auto-registered tasks
}
// Initialize task executor registry
worker.initializeTaskExecutors()
glog.V(1).Infof("Created maintenance worker with %d registered task types", len(worker.taskRegistry.GetAll()))
return worker
}
// executeGenericTask executes a task using the task registry instead of hardcoded methods
func (mws *MaintenanceWorkerService) executeGenericTask(task *MaintenanceTask) error {
glog.V(2).Infof("Executing generic task %s: %s for volume %d", task.ID, task.Type, task.VolumeID)
// Validate that task has proper typed parameters
if task.TypedParams == nil {
return fmt.Errorf("task %s has no typed parameters - task was not properly planned (insufficient destinations)", task.ID)
}
// Convert MaintenanceTask to types.TaskType
taskType := types.TaskType(string(task.Type))
// Create task instance using the registry
taskInstance, err := mws.taskRegistry.Get(taskType).Create(task.TypedParams)
if err != nil {
return fmt.Errorf("failed to create task instance: %w", err)
}
// Update progress to show task has started
mws.updateTaskProgress(task.ID, 5)
// Execute the task
err = taskInstance.Execute(context.Background(), task.TypedParams)
if err != nil {
return fmt.Errorf("task execution failed: %w", err)
}
// Update progress to show completion
mws.updateTaskProgress(task.ID, 100)
glog.V(2).Infof("Generic task %s completed successfully", task.ID)
return nil
}
// initializeTaskExecutors sets up the task execution registry dynamically
func (mws *MaintenanceWorkerService) initializeTaskExecutors() {
mws.taskExecutors = make(map[MaintenanceTaskType]TaskExecutor)
// Get all registered executor factories and create executors
executorRegistryMutex.RLock()
defer executorRegistryMutex.RUnlock()
for taskType, factory := range taskExecutorFactories {
executor := factory()
mws.taskExecutors[taskType] = executor
glog.V(3).Infof("Initialized executor for task type: %s", taskType)
}
glog.V(2).Infof("Initialized %d task executors", len(mws.taskExecutors))
}
// RegisterTaskExecutor allows dynamic registration of new task executors
func (mws *MaintenanceWorkerService) RegisterTaskExecutor(taskType MaintenanceTaskType, executor TaskExecutor) {
if mws.taskExecutors == nil {
mws.taskExecutors = make(map[MaintenanceTaskType]TaskExecutor)
}
mws.taskExecutors[taskType] = executor
glog.V(1).Infof("Registered executor for task type: %s", taskType)
}
// GetSupportedTaskTypes returns all task types that this worker can execute
func (mws *MaintenanceWorkerService) GetSupportedTaskTypes() []MaintenanceTaskType {
return GetSupportedExecutorTaskTypes()
}
// Start begins the worker service
func (mws *MaintenanceWorkerService) Start() error {
mws.running = true
// Register with admin server
worker := &MaintenanceWorker{
ID: mws.workerID,
Address: mws.address,
Capabilities: mws.capabilities,
MaxConcurrent: mws.maxConcurrent,
}
if mws.queue != nil {
mws.queue.RegisterWorker(worker)
}
// Start worker loop
go mws.workerLoop()
glog.Infof("Maintenance worker %s started at %s", mws.workerID, mws.address)
return nil
}
// Stop terminates the worker service
func (mws *MaintenanceWorkerService) Stop() {
mws.running = false
close(mws.stopChan)
// Wait for current tasks to complete or timeout
timeout := time.NewTimer(30 * time.Second)
defer timeout.Stop()
for len(mws.currentTasks) > 0 {
select {
case <-timeout.C:
glog.Warningf("Worker %s stopping with %d tasks still running", mws.workerID, len(mws.currentTasks))
return
case <-time.After(time.Second):
// Check again
}
}
glog.Infof("Maintenance worker %s stopped", mws.workerID)
}
// workerLoop is the main worker event loop
func (mws *MaintenanceWorkerService) workerLoop() {
heartbeatTicker := time.NewTicker(30 * time.Second)
defer heartbeatTicker.Stop()
taskRequestTicker := time.NewTicker(5 * time.Second)
defer taskRequestTicker.Stop()
for mws.running {
select {
case <-mws.stopChan:
return
case <-heartbeatTicker.C:
mws.sendHeartbeat()
case <-taskRequestTicker.C:
mws.requestTasks()
}
}
}
// sendHeartbeat sends heartbeat to admin server
func (mws *MaintenanceWorkerService) sendHeartbeat() {
if mws.queue != nil {
mws.queue.UpdateWorkerHeartbeat(mws.workerID)
}
}
// requestTasks requests new tasks from the admin server
func (mws *MaintenanceWorkerService) requestTasks() {
if len(mws.currentTasks) >= mws.maxConcurrent {
return // Already at capacity
}
if mws.queue != nil {
task := mws.queue.GetNextTask(mws.workerID, mws.capabilities)
if task != nil {
mws.executeTask(task)
}
}
}
// executeTask executes a maintenance task
func (mws *MaintenanceWorkerService) executeTask(task *MaintenanceTask) {
mws.currentTasks[task.ID] = task
go func() {
defer func() {
delete(mws.currentTasks, task.ID)
}()
glog.Infof("Worker %s executing task %s: %s", mws.workerID, task.ID, task.Type)
// Execute task using dynamic executor registry
var err error
if executor, exists := mws.taskExecutors[task.Type]; exists {
err = executor(mws, task)
} else {
err = fmt.Errorf("unsupported task type: %s", task.Type)
glog.Errorf("No executor registered for task type: %s", task.Type)
}
// Report task completion
if mws.queue != nil {
errorMsg := ""
if err != nil {
errorMsg = err.Error()
}
mws.queue.CompleteTask(task.ID, errorMsg)
}
if err != nil {
glog.Errorf("Worker %s failed to execute task %s: %v", mws.workerID, task.ID, err)
} else {
glog.Infof("Worker %s completed task %s successfully", mws.workerID, task.ID)
}
}()
}
// updateTaskProgress updates the progress of a task
func (mws *MaintenanceWorkerService) updateTaskProgress(taskID string, progress float64) {
if mws.queue != nil {
mws.queue.UpdateTaskProgress(taskID, progress)
}
}
// GetStatus returns the current status of the worker
func (mws *MaintenanceWorkerService) GetStatus() map[string]interface{} {
return map[string]interface{}{
"worker_id": mws.workerID,
"address": mws.address,
"running": mws.running,
"capabilities": mws.capabilities,
"max_concurrent": mws.maxConcurrent,
"current_tasks": len(mws.currentTasks),
"task_details": mws.currentTasks,
}
}
// SetQueue sets the maintenance queue for the worker
func (mws *MaintenanceWorkerService) SetQueue(queue *MaintenanceQueue) {
mws.queue = queue
}
// SetAdminClient sets the admin client for the worker
func (mws *MaintenanceWorkerService) SetAdminClient(client AdminClient) {
mws.adminClient = client
}
// SetCapabilities sets the worker capabilities
func (mws *MaintenanceWorkerService) SetCapabilities(capabilities []MaintenanceTaskType) {
mws.capabilities = capabilities
}
// SetMaxConcurrent sets the maximum concurrent tasks
func (mws *MaintenanceWorkerService) SetMaxConcurrent(max int) {
mws.maxConcurrent = max
}
// SetHeartbeatInterval sets the heartbeat interval (placeholder for future use)
func (mws *MaintenanceWorkerService) SetHeartbeatInterval(interval time.Duration) {
// Future implementation for configurable heartbeat
}
// SetTaskRequestInterval sets the task request interval (placeholder for future use)
func (mws *MaintenanceWorkerService) SetTaskRequestInterval(interval time.Duration) {
// Future implementation for configurable task requests
}
// MaintenanceWorkerCommand represents a standalone maintenance worker command
type MaintenanceWorkerCommand struct {
workerService *MaintenanceWorkerService
}
// NewMaintenanceWorkerCommand creates a new worker command
func NewMaintenanceWorkerCommand(workerID, address, adminServer string) *MaintenanceWorkerCommand {
return &MaintenanceWorkerCommand{
workerService: NewMaintenanceWorkerService(workerID, address, adminServer),
}
}
// Run starts the maintenance worker as a standalone service
func (mwc *MaintenanceWorkerCommand) Run() error {
// Generate or load persistent worker ID if not provided
if mwc.workerService.workerID == "" {
// Get current working directory for worker ID persistence
wd, err := os.Getwd()
if err != nil {
return fmt.Errorf("failed to get working directory: %w", err)
}
workerID, err := worker.GenerateOrLoadWorkerID(wd)
if err != nil {
return fmt.Errorf("failed to generate or load worker ID: %w", err)
}
mwc.workerService.workerID = workerID
}
// Start the worker service
err := mwc.workerService.Start()
if err != nil {
return fmt.Errorf("failed to start maintenance worker: %w", err)
}
// Wait for interrupt signal
select {}
}
+9 -3
View File
@@ -122,6 +122,7 @@ type Plugin struct {
type streamSession struct {
workerID string
outgoing chan *plugin_pb.AdminToWorkerMessage
done chan struct{}
closeOnce sync.Once
}
@@ -274,6 +275,7 @@ func (r *Plugin) WorkerStream(stream plugin_pb.PluginControlService_WorkerStream
session := &streamSession{
workerID: workerID,
outgoing: make(chan *plugin_pb.AdminToWorkerMessage, r.outgoingBuffer),
done: make(chan struct{}),
}
r.putSession(session)
defer r.cleanupSession(workerID)
@@ -908,8 +910,10 @@ func (r *Plugin) sendLoop(
return nil
case <-r.shutdownCh:
return nil
case msg, ok := <-session.outgoing:
if !ok {
case <-session.done:
return nil
case msg := <-session.outgoing:
if msg == nil {
return nil
}
if err := stream.Send(msg); err != nil {
@@ -930,6 +934,8 @@ func (r *Plugin) sendToWorker(workerID string, message *plugin_pb.AdminToWorkerM
select {
case <-r.shutdownCh:
return fmt.Errorf("plugin is shutting down")
case <-session.done:
return fmt.Errorf("worker %s session is closed", workerID)
case session.outgoing <- message:
return nil
case <-time.After(r.sendTimeout):
@@ -1425,7 +1431,7 @@ func CloneConfigValueMap(in map[string]*plugin_pb.ConfigValue) map[string]*plugi
func (s *streamSession) close() {
s.closeOnce.Do(func() {
close(s.outgoing)
close(s.done)
})
}
+6 -6
View File
@@ -26,7 +26,7 @@ func TestRunDetectionSendsCancelOnContextDone(t *testing.T) {
{JobType: jobType, CanDetect: true, MaxDetectionConcurrency: 1},
},
})
session := &streamSession{workerID: workerID, outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 4)}
session := &streamSession{workerID: workerID, outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 4), done: make(chan struct{})}
pluginSvc.putSession(session)
ctx, cancel := context.WithCancel(context.Background())
@@ -77,7 +77,7 @@ func TestExecuteJobSendsCancelOnContextDone(t *testing.T) {
{JobType: jobType, CanExecute: true, MaxExecutionConcurrency: 1},
},
})
session := &streamSession{workerID: workerID, outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 4)}
session := &streamSession{workerID: workerID, outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 4), done: make(chan struct{})}
pluginSvc.putSession(session)
job := &plugin_pb.JobSpec{JobId: "job-1", JobType: jobType}
@@ -135,8 +135,8 @@ func TestAdminScriptExecutionBlocksOtherDetection(t *testing.T) {
{JobType: "vacuum", CanDetect: true, MaxDetectionConcurrency: 1},
},
})
adminSession := &streamSession{workerID: adminWorkerID, outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 8)}
otherSession := &streamSession{workerID: otherWorkerID, outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 8)}
adminSession := &streamSession{workerID: adminWorkerID, outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 8), done: make(chan struct{})}
otherSession := &streamSession{workerID: otherWorkerID, outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 8), done: make(chan struct{})}
pluginSvc.putSession(adminSession)
pluginSvc.putSession(otherSession)
@@ -214,8 +214,8 @@ func TestAdminScriptExecutionBlocksOtherExecution(t *testing.T) {
{JobType: "vacuum", CanExecute: true, MaxExecutionConcurrency: 1},
},
})
adminSession := &streamSession{workerID: adminWorkerID, outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 8)}
otherSession := &streamSession{workerID: otherWorkerID, outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 8)}
adminSession := &streamSession{workerID: adminWorkerID, outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 8), done: make(chan struct{})}
otherSession := &streamSession{workerID: otherWorkerID, outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 8), done: make(chan struct{})}
pluginSvc.putSession(adminSession)
pluginSvc.putSession(otherSession)
+4 -4
View File
@@ -22,7 +22,7 @@ func TestRunDetectionIncludesLatestSuccessfulRun(t *testing.T) {
{JobType: jobType, CanDetect: true, MaxDetectionConcurrency: 1},
},
})
session := &streamSession{workerID: "worker-a", outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 1)}
session := &streamSession{workerID: "worker-a", outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 1), done: make(chan struct{})}
pluginSvc.putSession(session)
oldSuccess := time.Date(2026, 1, 1, 0, 0, 0, 0, time.UTC)
@@ -80,7 +80,7 @@ func TestRunDetectionOmitsLastSuccessfulRunWhenNoSuccessHistory(t *testing.T) {
{JobType: jobType, CanDetect: true, MaxDetectionConcurrency: 1},
},
})
session := &streamSession{workerID: "worker-a", outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 1)}
session := &streamSession{workerID: "worker-a", outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 1), done: make(chan struct{})}
pluginSvc.putSession(session)
if err := pluginSvc.store.AppendRunRecord(jobType, &JobRunRecord{
@@ -130,7 +130,7 @@ func TestRunDetectionWithReportCapturesDetectionActivities(t *testing.T) {
{JobType: jobType, CanDetect: true, MaxDetectionConcurrency: 1},
},
})
session := &streamSession{workerID: "worker-a", outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 1)}
session := &streamSession{workerID: "worker-a", outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 1), done: make(chan struct{})}
pluginSvc.putSession(session)
reportCh := make(chan *DetectionReport, 1)
@@ -210,7 +210,7 @@ func TestRunDetectionAdminScriptUsesLastCompletedRun(t *testing.T) {
{JobType: jobType, CanDetect: true, MaxDetectionConcurrency: 1},
},
})
session := &streamSession{workerID: "worker-admin-script", outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 1)}
session := &streamSession{workerID: "worker-admin-script", outgoing: make(chan *plugin_pb.AdminToWorkerMessage, 1), done: make(chan struct{})}
pluginSvc.putSession(session)
successCompleted := time.Date(2026, 2, 1, 10, 0, 0, 0, time.UTC)
-86
View File
@@ -95,16 +95,6 @@ func (r *Plugin) laneSchedulerLoop(ls *schedulerLaneState) {
}
}
// schedulerLoop is kept for backward compatibility; it delegates to
// laneSchedulerLoop with the default lane. New code should not call this.
func (r *Plugin) schedulerLoop() {
ls := r.lanes[LaneDefault]
if ls == nil {
ls = newLaneState(LaneDefault)
}
r.laneSchedulerLoop(ls)
}
// runLaneSchedulerIteration runs one scheduling pass for a single lane,
// processing only the job types assigned to that lane.
//
@@ -229,82 +219,6 @@ func (r *Plugin) runLaneSchedulerIterationConcurrent(ls *schedulerLaneState, job
return hadJobs.Load()
}
// runSchedulerIteration is kept for backward compatibility. It runs a
// single iteration across ALL job types (equivalent to the old single-loop
// behavior). It is only used by the legacy schedulerLoop() fallback.
func (r *Plugin) runSchedulerIteration() bool {
ls := r.lanes[LaneDefault]
if ls == nil {
ls = newLaneState(LaneDefault)
}
// For backward compat, the old function processes all job types.
r.expireStaleJobs(time.Now().UTC())
jobTypes := r.registry.DetectableJobTypes()
if len(jobTypes) == 0 {
r.setSchedulerLoopState("", "idle")
return false
}
r.setSchedulerLoopState("", "waiting_for_lock")
releaseLock, err := r.acquireAdminLock("plugin scheduler iteration")
if err != nil {
glog.Warningf("Plugin scheduler failed to acquire lock: %v", err)
r.setSchedulerLoopState("", "idle")
return false
}
if releaseLock != nil {
defer releaseLock()
}
active := make(map[string]struct{}, len(jobTypes))
hadJobs := false
for _, jobType := range jobTypes {
active[jobType] = struct{}{}
policy, enabled, err := r.loadSchedulerPolicy(jobType)
if err != nil {
glog.Warningf("Plugin scheduler failed to load policy for %s: %v", jobType, err)
continue
}
if !enabled {
r.clearSchedulerJobType(jobType)
continue
}
initialDelay := time.Duration(0)
if runInfo := r.snapshotSchedulerRun(jobType); runInfo.lastRunStartedAt.IsZero() {
initialDelay = 5 * time.Second
}
if !r.markDetectionDue(jobType, policy.DetectionInterval, initialDelay) {
continue
}
detected := r.runJobTypeIteration(jobType, policy)
if detected {
hadJobs = true
}
}
r.pruneSchedulerState(active)
r.pruneDetectorLeases(active)
r.setSchedulerLoopState("", "idle")
return hadJobs
}
// wakeLane wakes the scheduler goroutine for a specific lane.
func (r *Plugin) wakeLane(lane SchedulerLane) {
if r == nil {
return
}
if ls, ok := r.lanes[lane]; ok {
select {
case ls.wakeCh <- struct{}{}:
default:
}
}
}
// wakeAllLanes wakes all lane scheduler goroutines.
func (r *Plugin) wakeAllLanes() {
if r == nil {
-10
View File
@@ -210,16 +210,6 @@ func (r *Plugin) setSchedulerLoopStateForJobType(jobType, phase string) {
}
}
func (r *Plugin) recordSchedulerIterationComplete(hadJobs bool) {
if r == nil {
return
}
r.schedulerLoopMu.Lock()
r.schedulerLoopState.lastIterationHadJobs = hadJobs
r.schedulerLoopState.lastIterationCompleted = time.Now().UTC()
r.schedulerLoopMu.Unlock()
}
func (r *Plugin) snapshotSchedulerLoopState() schedulerLoopState {
if r == nil {
return schedulerLoopState{}
+17 -10
View File
@@ -125,6 +125,9 @@ templ ObjectStoreUsers(data dash.ObjectStoreUsersData) {
<div class="d-flex align-items-center">
<i class="fas fa-user me-2 text-muted"></i>
<strong>{user.Username}</strong>
if user.IsStatic {
<span class="badge bg-secondary ms-2" title="Loaded from config file (read-only)">static</span>
}
</div>
</td>
<td>{user.Email}</td>
@@ -133,24 +136,28 @@ templ ObjectStoreUsers(data dash.ObjectStoreUsersData) {
</td>
<td>
<div class="btn-group btn-group-sm" role="group">
<button type="button" class="btn btn-outline-info"
<button type="button" class="btn btn-outline-info"
data-action="show-user-details" data-username={ user.Username }>
<i class="fas fa-info-circle"></i>
</button>
<button type="button" class="btn btn-outline-primary"
data-action="edit-user" data-username={ user.Username }>
<i class="fas fa-edit"></i>
</button>
if user.Username != "anonymous" {
if !user.IsStatic {
<button type="button" class="btn btn-outline-primary"
data-action="edit-user" data-username={ user.Username }>
<i class="fas fa-edit"></i>
</button>
}
if user.Username != "anonymous" && !user.IsStatic {
<button type="button" class="btn btn-outline-secondary"
data-action="manage-access-keys" data-username={ user.Username }>
<i class="fas fa-key"></i>
</button>
}
<button type="button" class="btn btn-outline-danger"
data-action="delete-user" data-username={ user.Username }>
<i class="fas fa-trash"></i>
</button>
if !user.IsStatic {
<button type="button" class="btn btn-outline-danger"
data-action="delete-user" data-username={ user.Username }>
<i class="fas fa-trash"></i>
</button>
}
</div>
</td>
</tr>
File diff suppressed because one or more lines are too long
+2 -2
View File
@@ -163,7 +163,7 @@ templ S3Buckets(data dash.S3BucketsData) {
for _, bucket := range data.Buckets {
<tr>
<td>
<a href={templ.SafeURL(fmt.Sprintf("/files?path=/buckets/%s", bucket.Name))}
<a href={dash.PUrl(ctx, fmt.Sprintf("/files?path=/buckets/%s", bucket.Name))}
class="text-decoration-none">
<i class="fas fa-cube me-2"></i>
{bucket.Name}
@@ -236,7 +236,7 @@ templ S3Buckets(data dash.S3BucketsData) {
</td>
<td>
<div class="btn-group btn-group-sm" role="group">
<a href={templ.SafeURL(fmt.Sprintf("/files?path=/buckets/%s", bucket.Name))}
<a href={dash.PUrl(ctx, fmt.Sprintf("/files?path=/buckets/%s", bucket.Name))}
class="btn btn-outline-success btn-sm"
title="Browse Files">
<i class="fas fa-folder-open"></i>
+4 -4
View File
@@ -171,9 +171,9 @@ func S3Buckets(data dash.S3BucketsData) templ.Component {
return templ_7745c5c3_Err
}
var templ_7745c5c3_Var5 templ.SafeURL
templ_7745c5c3_Var5, templ_7745c5c3_Err = templ.JoinURLErrs(templ.SafeURL(fmt.Sprintf("/files?path=/buckets/%s", bucket.Name)))
templ_7745c5c3_Var5, templ_7745c5c3_Err = templ.JoinURLErrs(dash.PUrl(ctx, fmt.Sprintf("/files?path=/buckets/%s", bucket.Name)))
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3_buckets.templ`, Line: 166, Col: 123}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3_buckets.templ`, Line: 166, Col: 124}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var5))
if templ_7745c5c3_Err != nil {
@@ -439,9 +439,9 @@ func S3Buckets(data dash.S3BucketsData) templ.Component {
return templ_7745c5c3_Err
}
var templ_7745c5c3_Var19 templ.SafeURL
templ_7745c5c3_Var19, templ_7745c5c3_Err = templ.JoinURLErrs(templ.SafeURL(fmt.Sprintf("/files?path=/buckets/%s", bucket.Name)))
templ_7745c5c3_Var19, templ_7745c5c3_Err = templ.JoinURLErrs(dash.PUrl(ctx, fmt.Sprintf("/files?path=/buckets/%s", bucket.Name)))
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3_buckets.templ`, Line: 239, Col: 127}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3_buckets.templ`, Line: 239, Col: 128}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var19))
if templ_7745c5c3_Err != nil {
+2 -1
View File
@@ -2,6 +2,7 @@ package app
import (
"fmt"
"net/url"
"github.com/seaweedfs/seaweedfs/weed/admin/dash"
"github.com/seaweedfs/seaweedfs/weed/s3api/s3tables"
@@ -152,7 +153,7 @@ templ S3TablesBuckets(data dash.S3TablesBucketsData) {
<div class="btn-group btn-group-sm" role="group">
{{ bucketName, parseErr := s3tables.ParseBucketNameFromARN(bucket.ARN) }}
if parseErr == nil {
<a class="btn btn-outline-primary btn-sm" href={ templ.SafeURL(fmt.Sprintf("/object-store/s3tables/buckets/%s/namespaces", bucketName)) }>
<a class="btn btn-outline-primary btn-sm" href={ dash.PUrl(ctx, fmt.Sprintf("/object-store/s3tables/buckets/%s/namespaces", url.PathEscape(bucketName))) }>
<i class="fas fa-folder-open"></i>
</a>
} else {
+21 -20
View File
@@ -10,6 +10,7 @@ import templruntime "github.com/a-h/templ/runtime"
import (
"fmt"
"net/url"
"github.com/seaweedfs/seaweedfs/weed/admin/dash"
"github.com/seaweedfs/seaweedfs/weed/s3api/s3tables"
@@ -48,7 +49,7 @@ func S3TablesBuckets(data dash.S3TablesBucketsData) templ.Component {
var templ_7745c5c3_Var2 templ.SafeURL
templ_7745c5c3_Var2, templ_7745c5c3_Err = templ.JoinURLErrs(templ.SafeURL(fmt.Sprintf("http://localhost:%d/v1/config", data.IcebergPort)))
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 23, Col: 124}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 24, Col: 124}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var2))
if templ_7745c5c3_Err != nil {
@@ -66,7 +67,7 @@ func S3TablesBuckets(data dash.S3TablesBucketsData) templ.Component {
var templ_7745c5c3_Var3 string
templ_7745c5c3_Var3, templ_7745c5c3_Err = templ.JoinStringErrs(fmt.Sprintf("%d", data.IcebergPort))
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 30, Col: 91}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 31, Col: 91}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var3))
if templ_7745c5c3_Err != nil {
@@ -84,7 +85,7 @@ func S3TablesBuckets(data dash.S3TablesBucketsData) templ.Component {
var templ_7745c5c3_Var4 string
templ_7745c5c3_Var4, templ_7745c5c3_Err = templ.JoinStringErrs(fmt.Sprintf("%d", data.IcebergPort))
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 37, Col: 109}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 38, Col: 109}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var4))
if templ_7745c5c3_Err != nil {
@@ -97,7 +98,7 @@ func S3TablesBuckets(data dash.S3TablesBucketsData) templ.Component {
var templ_7745c5c3_Var5 string
templ_7745c5c3_Var5, templ_7745c5c3_Err = templ.JoinStringErrs(fmt.Sprintf("%d", data.IcebergPort))
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 39, Col: 107}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 40, Col: 107}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var5))
if templ_7745c5c3_Err != nil {
@@ -120,7 +121,7 @@ func S3TablesBuckets(data dash.S3TablesBucketsData) templ.Component {
var templ_7745c5c3_Var6 string
templ_7745c5c3_Var6, templ_7745c5c3_Err = templ.JoinStringErrs(fmt.Sprintf("%d", data.TotalBuckets))
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 69, Col: 47}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 70, Col: 47}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var6))
if templ_7745c5c3_Err != nil {
@@ -133,7 +134,7 @@ func S3TablesBuckets(data dash.S3TablesBucketsData) templ.Component {
var templ_7745c5c3_Var7 string
templ_7745c5c3_Var7, templ_7745c5c3_Err = templ.JoinStringErrs(data.LastUpdated.Format("2006-01-02 15:04"))
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 88, Col: 54}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 89, Col: 54}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var7))
if templ_7745c5c3_Err != nil {
@@ -147,7 +148,7 @@ func S3TablesBuckets(data dash.S3TablesBucketsData) templ.Component {
var templ_7745c5c3_Var8 string
templ_7745c5c3_Var8, templ_7745c5c3_Err = templ.JoinStringErrs(fmt.Sprintf("%d", data.IcebergPort))
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 108, Col: 47}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 109, Col: 47}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var8))
if templ_7745c5c3_Err != nil {
@@ -171,7 +172,7 @@ func S3TablesBuckets(data dash.S3TablesBucketsData) templ.Component {
var templ_7745c5c3_Var9 string
templ_7745c5c3_Var9, templ_7745c5c3_Err = templ.JoinStringErrs(bucket.Name)
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 146, Col: 28}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 147, Col: 28}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var9))
if templ_7745c5c3_Err != nil {
@@ -184,7 +185,7 @@ func S3TablesBuckets(data dash.S3TablesBucketsData) templ.Component {
var templ_7745c5c3_Var10 string
templ_7745c5c3_Var10, templ_7745c5c3_Err = templ.JoinStringErrs(bucket.OwnerAccountID)
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 147, Col: 38}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 148, Col: 38}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var10))
if templ_7745c5c3_Err != nil {
@@ -197,7 +198,7 @@ func S3TablesBuckets(data dash.S3TablesBucketsData) templ.Component {
var templ_7745c5c3_Var11 string
templ_7745c5c3_Var11, templ_7745c5c3_Err = templ.JoinStringErrs(bucket.ARN)
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 148, Col: 52}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 149, Col: 52}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var11))
if templ_7745c5c3_Err != nil {
@@ -210,7 +211,7 @@ func S3TablesBuckets(data dash.S3TablesBucketsData) templ.Component {
var templ_7745c5c3_Var12 string
templ_7745c5c3_Var12, templ_7745c5c3_Err = templ.JoinStringErrs(bucket.Name)
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 149, Col: 52}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 150, Col: 52}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var12))
if templ_7745c5c3_Err != nil {
@@ -223,7 +224,7 @@ func S3TablesBuckets(data dash.S3TablesBucketsData) templ.Component {
var templ_7745c5c3_Var13 string
templ_7745c5c3_Var13, templ_7745c5c3_Err = templ.JoinStringErrs(bucket.CreatedAt.Format("2006-01-02 15:04"))
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 150, Col: 60}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 151, Col: 60}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var13))
if templ_7745c5c3_Err != nil {
@@ -240,9 +241,9 @@ func S3TablesBuckets(data dash.S3TablesBucketsData) templ.Component {
return templ_7745c5c3_Err
}
var templ_7745c5c3_Var14 templ.SafeURL
templ_7745c5c3_Var14, templ_7745c5c3_Err = templ.JoinURLErrs(templ.SafeURL(fmt.Sprintf("/object-store/s3tables/buckets/%s/namespaces", bucketName)))
templ_7745c5c3_Var14, templ_7745c5c3_Err = templ.JoinURLErrs(dash.PUrl(ctx, fmt.Sprintf("/object-store/s3tables/buckets/%s/namespaces", url.PathEscape(bucketName))))
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 155, Col: 149}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 156, Col: 166}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var14))
if templ_7745c5c3_Err != nil {
@@ -265,7 +266,7 @@ func S3TablesBuckets(data dash.S3TablesBucketsData) templ.Component {
var templ_7745c5c3_Var15 string
templ_7745c5c3_Var15, templ_7745c5c3_Err = templ.JoinStringErrs(bucket.ARN)
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 163, Col: 122}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 164, Col: 122}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var15))
if templ_7745c5c3_Err != nil {
@@ -278,7 +279,7 @@ func S3TablesBuckets(data dash.S3TablesBucketsData) templ.Component {
var templ_7745c5c3_Var16 string
templ_7745c5c3_Var16, templ_7745c5c3_Err = templ.JoinStringErrs(bucket.ARN)
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 166, Col: 126}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 167, Col: 126}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var16))
if templ_7745c5c3_Err != nil {
@@ -291,7 +292,7 @@ func S3TablesBuckets(data dash.S3TablesBucketsData) templ.Component {
var templ_7745c5c3_Var17 string
templ_7745c5c3_Var17, templ_7745c5c3_Err = templ.JoinStringErrs(bucket.ARN)
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 169, Col: 128}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 170, Col: 128}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var17))
if templ_7745c5c3_Err != nil {
@@ -304,7 +305,7 @@ func S3TablesBuckets(data dash.S3TablesBucketsData) templ.Component {
var templ_7745c5c3_Var18 string
templ_7745c5c3_Var18, templ_7745c5c3_Err = templ.JoinStringErrs(bucket.Name)
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 169, Col: 161}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 170, Col: 161}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var18))
if templ_7745c5c3_Err != nil {
@@ -342,7 +343,7 @@ CREATE SECRET (
SELECT * FROM iceberg_scan('s3://my-table-bucket/my-namespace/my-table');`)
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 219, Col: 74}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 220, Col: 74}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var19))
if templ_7745c5c3_Err != nil {
@@ -365,7 +366,7 @@ catalog = load_catalog(
namespaces = catalog.list_namespaces()`)
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 235, Col: 39}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_buckets.templ`, Line: 236, Col: 39}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var20))
if templ_7745c5c3_Err != nil {
+2 -2
View File
@@ -24,7 +24,7 @@ templ S3TablesTables(data dash.S3TablesTablesData) {
<div class="mb-3">
{{ bucketName, parseErr := s3tables.ParseBucketNameFromARN(data.BucketARN) }}
if parseErr == nil {
<a href={ templ.SafeURL(fmt.Sprintf("/object-store/s3tables/buckets/%s/namespaces", bucketName)) } class="btn btn-sm btn-outline-secondary">
<a href={ dash.PUrl(ctx, fmt.Sprintf("/object-store/s3tables/buckets/%s/namespaces", url.PathEscape(bucketName))) } class="btn btn-sm btn-outline-secondary">
<i class="fas fa-arrow-left me-1"></i>Back to Namespaces
</a>
} else {
@@ -126,7 +126,7 @@ templ S3TablesTables(data dash.S3TablesTablesData) {
<td>
<div class="btn-group btn-group-sm" role="group">
if parseErr == nil {
<a class="btn btn-outline-primary btn-sm" href={ templ.SafeURL(fmt.Sprintf("/object-store/s3tables/buckets/%s/namespaces/%s/tables/%s", url.PathEscape(bucketName), url.PathEscape(data.Namespace), url.PathEscape(tableName))) } title="View Iceberg Details">
<a class="btn btn-outline-primary btn-sm" href={ dash.PUrl(ctx, fmt.Sprintf("/object-store/s3tables/buckets/%s/namespaces/%s/tables/%s", url.PathEscape(bucketName), url.PathEscape(data.Namespace), url.PathEscape(tableName))) } title="View Iceberg Details">
<i class="fas fa-eye"></i>
</a>
} else {
+4 -4
View File
@@ -48,9 +48,9 @@ func S3TablesTables(data dash.S3TablesTablesData) templ.Component {
return templ_7745c5c3_Err
}
var templ_7745c5c3_Var2 templ.SafeURL
templ_7745c5c3_Var2, templ_7745c5c3_Err = templ.JoinURLErrs(templ.SafeURL(fmt.Sprintf("/object-store/s3tables/buckets/%s/namespaces", bucketName)))
templ_7745c5c3_Var2, templ_7745c5c3_Err = templ.JoinURLErrs(dash.PUrl(ctx, fmt.Sprintf("/object-store/s3tables/buckets/%s/namespaces", url.PathEscape(bucketName))))
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_tables.templ`, Line: 27, Col: 99}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_tables.templ`, Line: 27, Col: 116}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var2))
if templ_7745c5c3_Err != nil {
@@ -308,9 +308,9 @@ func S3TablesTables(data dash.S3TablesTablesData) templ.Component {
return templ_7745c5c3_Err
}
var templ_7745c5c3_Var17 templ.SafeURL
templ_7745c5c3_Var17, templ_7745c5c3_Err = templ.JoinURLErrs(templ.SafeURL(fmt.Sprintf("/object-store/s3tables/buckets/%s/namespaces/%s/tables/%s", url.PathEscape(bucketName), url.PathEscape(data.Namespace), url.PathEscape(tableName))))
templ_7745c5c3_Var17, templ_7745c5c3_Err = templ.JoinURLErrs(dash.PUrl(ctx, fmt.Sprintf("/object-store/s3tables/buckets/%s/namespaces/%s/tables/%s", url.PathEscape(bucketName), url.PathEscape(data.Namespace), url.PathEscape(tableName))))
if templ_7745c5c3_Err != nil {
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_tables.templ`, Line: 129, Col: 237}
return templ.Error{Err: templ_7745c5c3_Err, FileName: `view/app/s3tables_tables.templ`, Line: 129, Col: 238}
}
_, templ_7745c5c3_Err = templ_7745c5c3_Buffer.WriteString(templ.EscapeString(templ_7745c5c3_Var17))
if templ_7745c5c3_Err != nil {
-14
View File
@@ -6,20 +6,6 @@ import (
"strings"
)
// getStatusColor returns Bootstrap color class for status
func getStatusColor(status string) string {
switch status {
case "active", "healthy":
return "success"
case "warning":
return "warning"
case "critical", "unreachable":
return "danger"
default:
return "secondary"
}
}
// formatBytes converts bytes to human readable format
func formatBytes(bytes int64) string {
if bytes == 0 {
-12
View File
@@ -95,18 +95,6 @@ func NewCluster() *Cluster {
}
}
func (cluster *Cluster) getGroupMembers(filerGroup FilerGroupName, nodeType string, createIfNotFound bool) *GroupMembers {
switch nodeType {
case FilerType:
return cluster.filerGroups.getGroupMembers(filerGroup, createIfNotFound)
case BrokerType:
return cluster.brokerGroups.getGroupMembers(filerGroup, createIfNotFound)
case S3Type:
return cluster.s3Groups.getGroupMembers(filerGroup, createIfNotFound)
}
return nil
}
func (cluster *Cluster) AddClusterNode(ns, nodeType string, dataCenter DataCenter, rack Rack, address pb.ServerAddress, version string) []*master_pb.KeepConnectedResponse {
filerGroup := FilerGroupName(ns)
switch nodeType {
+25 -5
View File
@@ -30,6 +30,7 @@ import (
"github.com/seaweedfs/seaweedfs/weed/pb"
"github.com/seaweedfs/seaweedfs/weed/security"
"github.com/seaweedfs/seaweedfs/weed/util"
"github.com/seaweedfs/seaweedfs/weed/util/grace"
)
var (
@@ -48,6 +49,10 @@ type AdminOptions struct {
dataDir *string
icebergPort *int
urlPrefix *string
debug *bool
debugPort *int
cpuProfile *string
memProfile *string
}
func init() {
@@ -64,6 +69,10 @@ func init() {
a.readOnlyPassword = cmdAdmin.Flag.String("readOnlyPassword", "", "read-only user password (optional, for view-only access; requires adminPassword to be set)")
a.icebergPort = cmdAdmin.Flag.Int("iceberg.port", 8181, "Iceberg REST Catalog port (0 to hide in UI)")
a.urlPrefix = cmdAdmin.Flag.String("urlPrefix", "", "URL path prefix when running behind a reverse proxy under a subdirectory (e.g. /seaweedfs)")
a.debug = cmdAdmin.Flag.Bool("debug", false, "serves runtime profiling data via pprof on the port specified by -debug.port")
a.debugPort = cmdAdmin.Flag.Int("debug.port", 6060, "http port for debugging")
a.cpuProfile = cmdAdmin.Flag.String("cpuprofile", "", "cpu profile output file")
a.memProfile = cmdAdmin.Flag.String("memprofile", "", "memory profile output file")
}
var cmdAdmin = &Command{
@@ -140,6 +149,16 @@ var cmdAdmin = &Command{
- All static assets, API endpoints, and navigation links will use the prefix
- Session cookies are scoped to the prefix path
Debugging and Profiling:
- Use -debug to start a pprof HTTP server for live profiling (localhost only)
- Set -debug.port to choose the pprof port (default 6060)
- Profiles are accessible at http://127.0.0.1:<debug.port>/debug/pprof/
- Use -cpuprofile and -memprofile to write profiles to files on shutdown
- WARNING: -debug exposes runtime internals; use only in trusted environments
- Examples:
weed admin -debug -debug.port=6060 -master="localhost:9333"
weed admin -cpuprofile=cpu.prof -memprofile=mem.prof -master="localhost:9333"
Configuration File:
- The security.toml file is read from ".", "$HOME/.seaweedfs/",
"/usr/local/etc/seaweedfs/", or "/etc/seaweedfs/", in that order
@@ -149,6 +168,12 @@ var cmdAdmin = &Command{
}
func runAdmin(cmd *Command, args []string) bool {
if *a.debug {
grace.StartDebugServer(*a.debugPort)
}
grace.SetupProfiling(*a.cpuProfile, *a.memProfile)
// Load security configuration
util.LoadSecurityConfiguration()
@@ -511,11 +536,6 @@ func recoveryMiddleware(next http.Handler) http.Handler {
})
}
// GetAdminOptions returns the admin command options for testing
func GetAdminOptions() *AdminOptions {
return &AdminOptions{}
}
// loadOrGenerateSessionKeys loads or creates authentication/encryption keys for session cookies.
func loadOrGenerateSessionKeys(dataDir string) ([]byte, []byte, error) {
const keyLen = 32
-13
View File
@@ -132,16 +132,3 @@ func fetchContent(masterFn operation.GetMasterFn, grpcDialOption grpc.DialOption
content, e = io.ReadAll(rc.Body)
return
}
func WriteFile(filename string, data []byte, perm os.FileMode) error {
f, err := os.OpenFile(filename, os.O_WRONLY|os.O_CREATE|os.O_TRUNC, perm)
if err != nil {
return err
}
n, err := f.Write(data)
f.Close()
if err == nil && n < len(data) {
err = io.ErrShortWrite
}
return err
}
+11
View File
@@ -81,6 +81,7 @@ type FilerOptions struct {
exposeDirectoryData *bool
tusBasePath *string
certProvider certprovider.Provider
s3ConfigFile *string // optional path to static S3 identity config
}
func init() {
@@ -146,6 +147,7 @@ func init() {
filerS3Options.iamReadOnly = cmdFiler.Flag.Bool("s3.iam.readOnly", true, "disable IAM write operations on this server")
filerS3Options.portIceberg = cmdFiler.Flag.Int("s3.port.iceberg", 8181, "Iceberg REST Catalog server listen port (0 to disable)")
filerS3Options.externalUrl = cmdFiler.Flag.String("s3.externalUrl", "", "the external URL clients use to connect (e.g. https://api.example.com:9000). Used for S3 signature verification behind a reverse proxy. Falls back to S3_EXTERNAL_URL env var.")
filerS3Options.defaultFileMode = cmdFiler.Flag.String("s3.defaultFileMode", "", "default file mode for S3 uploaded objects, e.g. 0660, 0644, 0666")
// start webdav on filer
filerStartWebDav = cmdFiler.Flag.Bool("webdav", false, "whether to start webdav gateway")
@@ -342,6 +344,15 @@ func (fo *FilerOptions) startFiler() {
glog.V(0).Infof("Initialized credential manager: %s", credentialManager.GetStoreName())
}
// Load static S3 identities from config file if specified
if fo.s3ConfigFile != nil && *fo.s3ConfigFile != "" {
if credentialManager != nil {
if err := credentialManager.LoadS3ConfigFile(*fo.s3ConfigFile); err != nil {
glog.Warningf("Failed to load S3 config file for static identities: %v", err)
}
}
}
fs, nfs_err := weed_server.NewFilerServer(defaultMux, publicVolumeMux, &weed_server.FilerOption{
Masters: fo.masters,
FilerGroup: *fo.filerGroup,
+5
View File
@@ -250,6 +250,7 @@ func initMiniS3Flags() {
miniS3Options.auditLogConfig = cmdMini.Flag.String("s3.auditLogConfig", "", "path to the audit log config file")
miniS3Options.allowDeleteBucketNotEmpty = miniS3AllowDeleteBucketNotEmpty
miniS3Options.externalUrl = cmdMini.Flag.String("s3.externalUrl", "", "the external URL clients use to connect (e.g. https://api.example.com:9000). Used for S3 signature verification behind a reverse proxy. Falls back to S3_EXTERNAL_URL env var.")
miniS3Options.defaultFileMode = cmdMini.Flag.String("s3.defaultFileMode", "", "default file mode for S3 uploaded objects, e.g. 0660, 0644, 0666")
// In mini mode, S3 uses the shared debug server started at line 681, not its own separate debug server
miniS3Options.debug = new(bool) // explicitly false
miniS3Options.debugPort = cmdMini.Flag.Int("s3.debug.port", 6060, "http port for debugging (unused in mini mode)")
@@ -819,6 +820,10 @@ func runMini(cmd *Command, args []string) bool {
miniFilerOptions.disableHttp = miniDisableHttp
miniMasterOptions.disableHttp = miniDisableHttp
// Share the S3 static identity config file with the filer so its
// credential manager can also serve static users.
miniFilerOptions.s3ConfigFile = miniS3Config
filerAddress := string(pb.NewServerAddress(*miniIp, *miniFilerOptions.port, *miniFilerOptions.portGrpc))
miniS3Options.filer = &filerAddress
miniWebDavOptions.filer = &filerAddress
+23
View File
@@ -10,6 +10,7 @@ import (
"net/http"
"os"
"runtime"
"strconv"
"strings"
"time"
@@ -36,6 +37,9 @@ var (
s3StandaloneOptions S3Options
)
// S3Options holds CLI flags for the S3 gateway.
// Flags are registered in multiple commands: s3.go (standalone), server.go, filer.go, and mini.go.
// When adding a new field, update all four flag registration sites.
type S3Options struct {
filer *string
bindIp *string
@@ -68,6 +72,7 @@ type S3Options struct {
debugPort *int
cipher *bool
externalUrl *string
defaultFileMode *string
}
func init() {
@@ -103,6 +108,7 @@ func init() {
s3StandaloneOptions.debugPort = cmdS3.Flag.Int("debug.port", 6060, "http port for debugging")
s3StandaloneOptions.cipher = cmdS3.Flag.Bool("encryptVolumeData", false, "encrypt data on volume servers")
s3StandaloneOptions.externalUrl = cmdS3.Flag.String("externalUrl", "", "the external URL clients use to connect (e.g. https://api.example.com:9000). Used for S3 signature verification behind a reverse proxy. Falls back to S3_EXTERNAL_URL env var.")
s3StandaloneOptions.defaultFileMode = cmdS3.Flag.String("defaultFileMode", "", "default file mode for S3 uploaded objects, e.g. 0660, 0644, 0666")
}
var cmdS3 = &Command{
@@ -232,6 +238,17 @@ func (s3opt *S3Options) resolveExternalUrl() string {
return os.Getenv("S3_EXTERNAL_URL")
}
func (s3opt *S3Options) parseDefaultFileMode() (uint32, error) {
if s3opt.defaultFileMode == nil || *s3opt.defaultFileMode == "" {
return 0, nil
}
mode, err := strconv.ParseUint(*s3opt.defaultFileMode, 8, 32)
if err != nil {
return 0, fmt.Errorf("invalid defaultFileMode %q: %v", *s3opt.defaultFileMode, err)
}
return uint32(mode), nil
}
func (s3opt *S3Options) startS3Server() bool {
filerAddresses := pb.ServerAddresses(*s3opt.filer).ToAddresses()
@@ -298,6 +315,11 @@ func (s3opt *S3Options) startS3Server() bool {
*s3opt.bindIp = "0.0.0.0"
}
defaultFileMode, fileModeErr := s3opt.parseDefaultFileMode()
if fileModeErr != nil {
glog.Fatalf("S3 API Server startup error: %v", fileModeErr)
}
s3ApiServer, s3ApiServer_err = s3api.NewS3ApiServer(router, &s3api.S3ApiServerOption{
Filers: filerAddresses,
Masters: masterAddresses,
@@ -320,6 +342,7 @@ func (s3opt *S3Options) startS3Server() bool {
BindIp: *s3opt.bindIp,
GrpcPort: *s3opt.portGrpc,
ExternalUrl: s3opt.resolveExternalUrl(),
DefaultFileMode: defaultFileMode,
})
if s3ApiServer_err != nil {
glog.Fatalf("S3 API Server startup error: %v", s3ApiServer_err)
+13
View File
@@ -179,6 +179,19 @@ password = ""
user = ""
password = ""
# SSE-S3 server-side encryption key management
# These settings configure the Key Encryption Key (KEK) for S3 SSE-S3 encryption.
# Set exactly one of kek or key. If neither is set, SSE-S3 is disabled.
# Can also be set via env vars: WEED_S3_SSE_KEK, WEED_S3_SSE_KEY
[s3.sse]
# hex-encoded 256-bit key, same format as the legacy /etc/s3/sse_kek filer file.
# Use this to migrate from a filer-stored KEK: copy the value from /etc/s3/sse_kek.
# Generate a new one with: openssl rand -hex 32
kek = ""
# any secret string; a 256-bit key is derived automatically via HKDF-SHA256.
# Cannot be used while /etc/s3/sse_kek exists on the filer — delete it first.
key = ""
# white list. It's checking request ip address.
[guard]
white_list = ""
+3
View File
@@ -180,6 +180,7 @@ func init() {
s3Options.iamReadOnly = cmdServer.Flag.Bool("s3.iam.readOnly", true, "disable IAM write operations on this server")
s3Options.cipher = cmdServer.Flag.Bool("s3.encryptVolumeData", false, "encrypt data on volume servers for S3 uploads")
s3Options.externalUrl = cmdServer.Flag.String("s3.externalUrl", "", "the external URL clients use to connect (e.g. https://api.example.com:9000). Used for S3 signature verification behind a reverse proxy. Falls back to S3_EXTERNAL_URL env var.")
s3Options.defaultFileMode = cmdServer.Flag.String("s3.defaultFileMode", "", "default file mode for S3 uploaded objects, e.g. 0660, 0644, 0666")
sftpOptions.port = cmdServer.Flag.Int("sftp.port", 2022, "SFTP server listen port")
sftpOptions.sshPrivateKey = cmdServer.Flag.String("sftp.sshPrivateKey", "", "path to the SSH private key file for host authentication")
@@ -371,6 +372,8 @@ func runServer(cmd *Command, args []string) bool {
} else if *serverIamConfig != "" && *s3Options.iamConfig != *serverIamConfig {
glog.V(0).Infof("both -s3.iam.config(%s) and -iam.config(%s) provided; using -s3.iam.config", *s3Options.iamConfig, *serverIamConfig)
}
// Share the S3 static identity config file with the filer
filerOptions.s3ConfigFile = s3Options.config
go func() {
time.Sleep(2 * time.Second)
s3Options.localFilerSocket = filerOptions.localSocket
-36
View File
@@ -57,42 +57,6 @@ func LoadCredentialConfiguration() (*CredentialConfig, error) {
}, nil
}
// GetCredentialStoreConfig extracts credential store configuration from command line flags
// This is used when credential store is configured via command line instead of credential.toml
func GetCredentialStoreConfig(store string, config util.Configuration, prefix string) *CredentialConfig {
if store == "" {
return nil
}
return &CredentialConfig{
Store: store,
Config: config,
Prefix: prefix,
}
}
// MergeCredentialConfig merges command line credential config with credential.toml config
// Command line flags take priority over credential.toml
func MergeCredentialConfig(cmdLineStore string, cmdLineConfig util.Configuration, cmdLinePrefix string) (*CredentialConfig, error) {
// If command line credential store is specified, use it
if cmdLineStore != "" {
glog.V(0).Infof("Using command line credential configuration: store=%s", cmdLineStore)
return GetCredentialStoreConfig(cmdLineStore, cmdLineConfig, cmdLinePrefix), nil
}
// Otherwise, try to load from credential.toml
config, err := LoadCredentialConfiguration()
if err != nil {
return nil, err
}
if config == nil {
glog.V(1).Info("No credential store configured")
}
return config, nil
}
// NewCredentialManagerWithDefaults creates a credential manager with fallback to defaults
// If explicitStore is provided, it will be used regardless of credential.toml
// If explicitStore is empty, it tries credential.toml first, then defaults to "filer_etc"
+127 -5
View File
@@ -3,14 +3,18 @@ package credential
import (
"context"
"fmt"
"os"
"strings"
"sync"
"github.com/seaweedfs/seaweedfs/weed/glog"
"github.com/seaweedfs/seaweedfs/weed/pb"
"github.com/seaweedfs/seaweedfs/weed/pb/iam_pb"
"github.com/seaweedfs/seaweedfs/weed/s3api/policy_engine"
"github.com/seaweedfs/seaweedfs/weed/util"
"github.com/seaweedfs/seaweedfs/weed/wdclient"
"google.golang.org/grpc"
"google.golang.org/protobuf/encoding/protojson"
)
// FilerAddressSetter is an interface for credential stores that need a dynamic filer address
@@ -21,6 +25,15 @@ type FilerAddressSetter interface {
// CredentialManager manages user credentials using a configurable store
type CredentialManager struct {
Store CredentialStore
// staticMu protects staticIdentities and staticNames, which are written
// by SetStaticIdentities (startup + config reload) and read concurrently
// by LoadConfiguration, SaveConfiguration, GetStaticUsernames, and IsStaticIdentity.
staticMu sync.RWMutex
// staticIdentities holds identities loaded from a static config file (-s3.config).
// These are included in LoadConfiguration so that listing operations
// return all configured identities, not just dynamic ones from the store.
staticIdentities []*iam_pb.Identity
staticNames map[string]bool
}
// NewCredentialManager creates a new credential manager with the specified store
@@ -74,13 +87,97 @@ func (cm *CredentialManager) GetStoreName() string {
return ""
}
// LoadConfiguration loads the S3 API configuration
func (cm *CredentialManager) LoadConfiguration(ctx context.Context) (*iam_pb.S3ApiConfiguration, error) {
return cm.Store.LoadConfiguration(ctx)
// SetStaticIdentities registers identities loaded from a static config file.
// These identities are included in LoadConfiguration and ListUsers results
// but are never persisted to the dynamic store.
func (cm *CredentialManager) SetStaticIdentities(identities []*iam_pb.Identity) {
filtered := make([]*iam_pb.Identity, 0, len(identities))
names := make(map[string]bool, len(identities))
for _, ident := range identities {
if ident != nil {
filtered = append(filtered, ident)
names[ident.Name] = true
}
}
cm.staticMu.Lock()
cm.staticIdentities = filtered
cm.staticNames = names
cm.staticMu.Unlock()
}
// SaveConfiguration saves the S3 API configuration
// IsStaticIdentity returns true if the named identity was loaded from static config.
func (cm *CredentialManager) IsStaticIdentity(name string) bool {
cm.staticMu.RLock()
defer cm.staticMu.RUnlock()
return cm.staticNames[name]
}
// GetStaticIdentity returns the protobuf identity for a static user, or nil.
func (cm *CredentialManager) GetStaticIdentity(name string) *iam_pb.Identity {
cm.staticMu.RLock()
defer cm.staticMu.RUnlock()
for _, ident := range cm.staticIdentities {
if ident.Name == name {
return ident
}
}
return nil
}
// GetStaticUsernames returns the names of all static identities.
func (cm *CredentialManager) GetStaticUsernames() []string {
cm.staticMu.RLock()
defer cm.staticMu.RUnlock()
names := make([]string, 0, len(cm.staticIdentities))
for _, ident := range cm.staticIdentities {
names = append(names, ident.Name)
}
return names
}
// LoadConfiguration loads the S3 API configuration from the store and merges
// in any static identities so that listing operations show all users.
func (cm *CredentialManager) LoadConfiguration(ctx context.Context) (*iam_pb.S3ApiConfiguration, error) {
config, err := cm.Store.LoadConfiguration(ctx)
if err != nil {
return config, err
}
// Merge static identities that are not already in the dynamic config
cm.staticMu.RLock()
staticIdents := cm.staticIdentities
cm.staticMu.RUnlock()
if len(staticIdents) > 0 {
dynamicNames := make(map[string]bool, len(config.Identities))
for _, ident := range config.Identities {
dynamicNames[ident.Name] = true
}
for _, si := range staticIdents {
if !dynamicNames[si.Name] {
config.Identities = append(config.Identities, si)
}
}
}
return config, nil
}
// SaveConfiguration saves the S3 API configuration.
// Static identities are filtered out before saving to the store.
// The caller's config is not mutated.
func (cm *CredentialManager) SaveConfiguration(ctx context.Context, config *iam_pb.S3ApiConfiguration) error {
cm.staticMu.RLock()
staticNames := cm.staticNames
cm.staticMu.RUnlock()
if len(staticNames) > 0 {
var dynamicOnly []*iam_pb.Identity
for _, ident := range config.Identities {
if !staticNames[ident.Name] {
dynamicOnly = append(dynamicOnly, ident)
}
}
configCopy := *config
configCopy.Identities = dynamicOnly
return cm.Store.SaveConfiguration(ctx, &configCopy)
}
return cm.Store.SaveConfiguration(ctx, config)
}
@@ -104,7 +201,12 @@ func (cm *CredentialManager) DeleteUser(ctx context.Context, username string) er
return cm.Store.DeleteUser(ctx, username)
}
// ListUsers returns all usernames
// ListUsers returns usernames from the dynamic store via cm.Store.ListUsers.
// On store error the error is returned directly without merging static entries.
// Static identities (cm.staticIdentities) are NOT included here because
// internal callers (e.g. DeletePolicy) look up each user in the store and
// would fail on non-existent static entries. External callers that need the
// full list should merge GetStaticUsernames separately.
func (cm *CredentialManager) ListUsers(ctx context.Context) ([]string, error) {
return cm.Store.ListUsers(ctx)
}
@@ -169,6 +271,26 @@ func (cm *CredentialManager) UpdatePolicy(ctx context.Context, name string, docu
return cm.Store.PutPolicy(ctx, name, document)
}
// LoadS3ConfigFile reads a static S3 identity config file and registers
// the identities so they appear in LoadConfiguration and listing results.
func (cm *CredentialManager) LoadS3ConfigFile(path string) error {
content, err := os.ReadFile(path)
if err != nil {
return fmt.Errorf("read %s: %w", path, err)
}
config := &iam_pb.S3ApiConfiguration{}
opts := protojson.UnmarshalOptions{DiscardUnknown: true, AllowPartial: true}
if err := opts.Unmarshal(content, config); err != nil {
return fmt.Errorf("parse %s: %w", path, err)
}
for _, ident := range config.Identities {
ident.IsStatic = true
}
cm.SetStaticIdentities(config.Identities)
glog.V(1).Infof("Loaded %d static identities from %s", len(config.Identities), path)
return nil
}
// Shutdown performs cleanup
func (cm *CredentialManager) Shutdown() {
if cm.Store != nil {
@@ -207,32 +207,6 @@ func (store *FilerEtcStore) loadPoliciesFromMultiFile(ctx context.Context, polic
})
}
func (store *FilerEtcStore) migratePoliciesToMultiFile(ctx context.Context, policies map[string]policy_engine.PolicyDocument) error {
glog.Infof("Migrating IAM policies to multi-file layout...")
// 1. Save all policies to individual files
for name, policy := range policies {
if err := store.savePolicy(ctx, name, policy); err != nil {
return err
}
}
// 2. Rename legacy file
return store.withFilerClient(func(client filer_pb.SeaweedFilerClient) error {
_, err := client.AtomicRenameEntry(ctx, &filer_pb.AtomicRenameEntryRequest{
OldDirectory: filer.IamConfigDirectory,
OldName: filer.IamPoliciesFile,
NewDirectory: filer.IamConfigDirectory,
NewName: IamLegacyPoliciesOldFile,
})
if err != nil {
glog.Errorf("Failed to rename legacy IAM policies file %s/%s to %s: %v",
filer.IamConfigDirectory, filer.IamPoliciesFile, IamLegacyPoliciesOldFile, err)
}
return err
})
}
func (store *FilerEtcStore) savePolicy(ctx context.Context, name string, document policy_engine.PolicyDocument) error {
if err := validatePolicyName(name); err != nil {
return err
-221
View File
@@ -1,221 +0,0 @@
package credential
import (
"context"
"fmt"
"github.com/seaweedfs/seaweedfs/weed/glog"
"github.com/seaweedfs/seaweedfs/weed/pb/iam_pb"
"github.com/seaweedfs/seaweedfs/weed/util"
)
// MigrateCredentials migrates credentials from one store to another
func MigrateCredentials(fromStoreName, toStoreName CredentialStoreTypeName, configuration util.Configuration, fromPrefix, toPrefix string) error {
ctx := context.Background()
// Create source credential manager
fromCM, err := NewCredentialManager(fromStoreName, configuration, fromPrefix)
if err != nil {
return fmt.Errorf("failed to create source credential manager (%s): %v", fromStoreName, err)
}
defer fromCM.Shutdown()
// Create destination credential manager
toCM, err := NewCredentialManager(toStoreName, configuration, toPrefix)
if err != nil {
return fmt.Errorf("failed to create destination credential manager (%s): %v", toStoreName, err)
}
defer toCM.Shutdown()
// Load configuration from source
glog.Infof("Loading configuration from %s store...", fromStoreName)
config, err := fromCM.LoadConfiguration(ctx)
if err != nil {
return fmt.Errorf("failed to load configuration from source store: %w", err)
}
if config == nil || len(config.Identities) == 0 {
glog.Info("No identities found in source store")
return nil
}
glog.Infof("Found %d identities in source store", len(config.Identities))
// Migrate each identity
var migrated, failed int
for _, identity := range config.Identities {
glog.V(1).Infof("Migrating user: %s", identity.Name)
// Check if user already exists in destination
existingUser, err := toCM.GetUser(ctx, identity.Name)
if err != nil && err != ErrUserNotFound {
glog.Errorf("Failed to check if user %s exists in destination: %v", identity.Name, err)
failed++
continue
}
if existingUser != nil {
glog.Warningf("User %s already exists in destination store, skipping", identity.Name)
continue
}
// Create user in destination
err = toCM.CreateUser(ctx, identity)
if err != nil {
glog.Errorf("Failed to create user %s in destination store: %v", identity.Name, err)
failed++
continue
}
migrated++
glog.V(1).Infof("Successfully migrated user: %s", identity.Name)
}
glog.Infof("Migration completed: %d migrated, %d failed", migrated, failed)
if failed > 0 {
return fmt.Errorf("migration completed with %d failures", failed)
}
return nil
}
// ExportCredentials exports credentials from a store to a configuration
func ExportCredentials(storeName CredentialStoreTypeName, configuration util.Configuration, prefix string) (*iam_pb.S3ApiConfiguration, error) {
ctx := context.Background()
// Create credential manager
cm, err := NewCredentialManager(storeName, configuration, prefix)
if err != nil {
return nil, fmt.Errorf("failed to create credential manager (%s): %v", storeName, err)
}
defer cm.Shutdown()
// Load configuration
config, err := cm.LoadConfiguration(ctx)
if err != nil {
return nil, fmt.Errorf("failed to load configuration: %w", err)
}
return config, nil
}
// ImportCredentials imports credentials from a configuration to a store
func ImportCredentials(storeName CredentialStoreTypeName, configuration util.Configuration, prefix string, config *iam_pb.S3ApiConfiguration) error {
ctx := context.Background()
// Create credential manager
cm, err := NewCredentialManager(storeName, configuration, prefix)
if err != nil {
return fmt.Errorf("failed to create credential manager (%s): %v", storeName, err)
}
defer cm.Shutdown()
// Import each identity
var imported, failed int
for _, identity := range config.Identities {
glog.V(1).Infof("Importing user: %s", identity.Name)
// Check if user already exists
existingUser, err := cm.GetUser(ctx, identity.Name)
if err != nil && err != ErrUserNotFound {
glog.Errorf("Failed to check if user %s exists: %v", identity.Name, err)
failed++
continue
}
if existingUser != nil {
glog.Warningf("User %s already exists, skipping", identity.Name)
continue
}
// Create user
err = cm.CreateUser(ctx, identity)
if err != nil {
glog.Errorf("Failed to create user %s: %v", identity.Name, err)
failed++
continue
}
imported++
glog.V(1).Infof("Successfully imported user: %s", identity.Name)
}
glog.Infof("Import completed: %d imported, %d failed", imported, failed)
if failed > 0 {
return fmt.Errorf("import completed with %d failures", failed)
}
return nil
}
// ValidateCredentials validates that all credentials in a store are accessible
func ValidateCredentials(storeName CredentialStoreTypeName, configuration util.Configuration, prefix string) error {
ctx := context.Background()
// Create credential manager
cm, err := NewCredentialManager(storeName, configuration, prefix)
if err != nil {
return fmt.Errorf("failed to create credential manager (%s): %v", storeName, err)
}
defer cm.Shutdown()
// Load configuration
config, err := cm.LoadConfiguration(ctx)
if err != nil {
return fmt.Errorf("failed to load configuration: %w", err)
}
if config == nil || len(config.Identities) == 0 {
glog.Info("No identities found in store")
return nil
}
glog.Infof("Validating %d identities...", len(config.Identities))
// Validate each identity
var validated, failed int
for _, identity := range config.Identities {
// Check if user can be retrieved
user, err := cm.GetUser(ctx, identity.Name)
if err != nil {
glog.Errorf("Failed to retrieve user %s: %v", identity.Name, err)
failed++
continue
}
if user == nil {
glog.Errorf("User %s not found", identity.Name)
failed++
continue
}
// Validate access keys
for _, credential := range identity.Credentials {
accessKeyUser, err := cm.GetUserByAccessKey(ctx, credential.AccessKey)
if err != nil {
glog.Errorf("Failed to retrieve user by access key %s: %v", credential.AccessKey, err)
failed++
continue
}
if accessKeyUser == nil || accessKeyUser.Name != identity.Name {
glog.Errorf("Access key %s does not map to correct user %s", credential.AccessKey, identity.Name)
failed++
continue
}
}
validated++
glog.V(1).Infof("Successfully validated user: %s", identity.Name)
}
glog.Infof("Validation completed: %d validated, %d failed", validated, failed)
if failed > 0 {
return fmt.Errorf("validation completed with %d failures", failed)
}
return nil
}
-4
View File
@@ -246,10 +246,6 @@ func NewLogFileEntryCollector(f *Filer, startPosition log_buffer.MessagePosition
}
}
func (c *LogFileEntryCollector) hasMore() bool {
return c.dayEntryQueue.Len() > 0
}
func (c *LogFileEntryCollector) collectMore(v *OrderedLogVisitor) (err error) {
dayEntry := c.dayEntryQueue.Dequeue()
if dayEntry == nil {
+30
View File
@@ -128,6 +128,9 @@ func (fsw *FilerStoreWrapper) Initialize(configuration util.Configuration, prefi
}
func (fsw *FilerStoreWrapper) InsertEntry(ctx context.Context, entry *Entry) error {
if err := ctx.Err(); err != nil {
return err
}
ctx = context.WithoutCancel(ctx)
actualStore := fsw.getActualStore(entry.FullPath)
stats.FilerStoreCounter.WithLabelValues(actualStore.GetName(), "insert").Inc()
@@ -155,6 +158,9 @@ func (fsw *FilerStoreWrapper) InsertEntry(ctx context.Context, entry *Entry) err
// InsertEntryKnownAbsent skips the pre-insert FindEntry path when the caller has
// already established that the target path does not exist.
func (fsw *FilerStoreWrapper) InsertEntryKnownAbsent(ctx context.Context, entry *Entry) error {
if err := ctx.Err(); err != nil {
return err
}
ctx = context.WithoutCancel(ctx)
actualStore := fsw.getActualStore(entry.FullPath)
stats.FilerStoreCounter.WithLabelValues(actualStore.GetName(), "insert").Inc()
@@ -178,6 +184,9 @@ func (fsw *FilerStoreWrapper) InsertEntryKnownAbsent(ctx context.Context, entry
}
func (fsw *FilerStoreWrapper) UpdateEntry(ctx context.Context, entry *Entry) error {
if err := ctx.Err(); err != nil {
return err
}
ctx = context.WithoutCancel(ctx)
actualStore := fsw.getActualStore(entry.FullPath)
stats.FilerStoreCounter.WithLabelValues(actualStore.GetName(), "update").Inc()
@@ -236,6 +245,9 @@ func (fsw *FilerStoreWrapper) FindEntry(ctx context.Context, fp util.FullPath) (
}
func (fsw *FilerStoreWrapper) DeleteEntry(ctx context.Context, fp util.FullPath) (err error) {
if err := ctx.Err(); err != nil {
return err
}
ctx = context.WithoutCancel(ctx)
actualStore := fsw.getActualStore(fp)
stats.FilerStoreCounter.WithLabelValues(actualStore.GetName(), "delete").Inc()
@@ -264,6 +276,9 @@ func (fsw *FilerStoreWrapper) DeleteEntry(ctx context.Context, fp util.FullPath)
}
func (fsw *FilerStoreWrapper) DeleteOneEntry(ctx context.Context, existingEntry *Entry) (err error) {
if err := ctx.Err(); err != nil {
return err
}
ctx = context.WithoutCancel(ctx)
actualStore := fsw.getActualStore(existingEntry.FullPath)
stats.FilerStoreCounter.WithLabelValues(actualStore.GetName(), "delete").Inc()
@@ -288,6 +303,9 @@ func (fsw *FilerStoreWrapper) DeleteOneEntry(ctx context.Context, existingEntry
}
func (fsw *FilerStoreWrapper) DeleteFolderChildren(ctx context.Context, fp util.FullPath) (err error) {
if err := ctx.Err(); err != nil {
return err
}
ctx = context.WithoutCancel(ctx)
actualStore := fsw.getActualStore(fp + "/")
stats.FilerStoreCounter.WithLabelValues(actualStore.GetName(), "deleteFolderChildren").Inc()
@@ -394,11 +412,17 @@ func (fsw *FilerStoreWrapper) prefixFilterEntries(ctx context.Context, dirPath u
}
func (fsw *FilerStoreWrapper) BeginTransaction(ctx context.Context) (context.Context, error) {
if err := ctx.Err(); err != nil {
return nil, err
}
ctx = context.WithoutCancel(ctx)
return fsw.getDefaultStore().BeginTransaction(ctx)
}
func (fsw *FilerStoreWrapper) CommitTransaction(ctx context.Context) error {
if err := ctx.Err(); err != nil {
return err
}
ctx = context.WithoutCancel(ctx)
return fsw.getDefaultStore().CommitTransaction(ctx)
}
@@ -413,6 +437,9 @@ func (fsw *FilerStoreWrapper) Shutdown() {
}
func (fsw *FilerStoreWrapper) KvPut(ctx context.Context, key []byte, value []byte) (err error) {
if err := ctx.Err(); err != nil {
return err
}
ctx = context.WithoutCancel(ctx)
return fsw.getDefaultStore().KvPut(ctx, key, value)
}
@@ -421,6 +448,9 @@ func (fsw *FilerStoreWrapper) KvGet(ctx context.Context, key []byte) (value []by
return fsw.getDefaultStore().KvGet(ctx, key)
}
func (fsw *FilerStoreWrapper) KvDelete(ctx context.Context, key []byte) (err error) {
if err := ctx.Err(); err != nil {
return err
}
ctx = context.WithoutCancel(ctx)
return fsw.getDefaultStore().KvDelete(ctx, key)
}
+136
View File
@@ -2,8 +2,10 @@ package filer
import (
"context"
"errors"
"os"
"testing"
"time"
"github.com/seaweedfs/seaweedfs/weed/util"
"github.com/stretchr/testify/assert"
@@ -69,3 +71,137 @@ func TestFilerStoreWrapperMimeNormalization(t *testing.T) {
}
}
}
// cancelledCtx returns a context that is already cancelled.
func cancelledCtx() context.Context {
ctx, cancel := context.WithCancel(context.Background())
cancel()
return ctx
}
// expiredCtx returns a context whose deadline has already passed.
func expiredCtx() context.Context {
ctx, cancel := context.WithDeadline(context.Background(), time.Now().Add(-time.Second))
cancel() // release resources immediately as it's already expired
return ctx
}
func TestFilerStoreWrapperWriteOpsRejectCancelledContext(t *testing.T) {
newEntry := func(path string) *Entry {
return &Entry{
FullPath: util.FullPath(path),
Attr: Attr{Mode: 0o660, Mime: "application/octet-stream"},
}
}
// Each write operation that should be guarded.
writeOps := []struct {
name string
run func(*FilerStoreWrapper, context.Context) error
}{
{"InsertEntry", func(fsw *FilerStoreWrapper, ctx context.Context) error {
return fsw.InsertEntry(ctx, newEntry("/test/a"))
}},
{"InsertEntryKnownAbsent", func(fsw *FilerStoreWrapper, ctx context.Context) error {
return fsw.InsertEntryKnownAbsent(ctx, newEntry("/test/b"))
}},
{"UpdateEntry", func(fsw *FilerStoreWrapper, ctx context.Context) error {
_ = fsw.InsertEntry(context.Background(), newEntry("/test/c"))
return fsw.UpdateEntry(ctx, newEntry("/test/c"))
}},
{"DeleteEntry", func(fsw *FilerStoreWrapper, ctx context.Context) error {
_ = fsw.InsertEntry(context.Background(), newEntry("/test/d"))
return fsw.DeleteEntry(ctx, "/test/d")
}},
{"DeleteOneEntry", func(fsw *FilerStoreWrapper, ctx context.Context) error {
e := newEntry("/test/e")
_ = fsw.InsertEntry(context.Background(), e)
return fsw.DeleteOneEntry(ctx, e)
}},
{"DeleteFolderChildren", func(fsw *FilerStoreWrapper, ctx context.Context) error {
_ = fsw.InsertEntry(context.Background(), newEntry("/test/folder/child"))
return fsw.DeleteFolderChildren(ctx, "/test/folder")
}},
{"BeginTransaction", func(fsw *FilerStoreWrapper, ctx context.Context) error {
_, err := fsw.BeginTransaction(ctx)
return err
}},
{"CommitTransaction", func(fsw *FilerStoreWrapper, ctx context.Context) error {
return fsw.CommitTransaction(ctx)
}},
{"KvPut", func(fsw *FilerStoreWrapper, ctx context.Context) error {
return fsw.KvPut(ctx, []byte("k"), []byte("v"))
}},
{"KvDelete", func(fsw *FilerStoreWrapper, ctx context.Context) error {
_ = fsw.KvPut(context.Background(), []byte("k"), []byte("v"))
return fsw.KvDelete(ctx, []byte("k"))
}},
}
badContexts := []struct {
name string
ctx context.Context
wantError error
}{
{"cancelled", cancelledCtx(), context.Canceled},
{"deadline exceeded", expiredCtx(), context.DeadlineExceeded},
}
for _, op := range writeOps {
for _, bc := range badContexts {
t.Run(op.name+"/"+bc.name, func(t *testing.T) {
wrapper := NewFilerStoreWrapper(newStubFilerStore())
err := op.run(wrapper, bc.ctx)
require.Error(t, err)
assert.True(t, errors.Is(err, bc.wantError), "got %v, want %v", err, bc.wantError)
})
}
}
}
func TestFilerStoreWrapperWriteOpsSucceedWithActiveContext(t *testing.T) {
wrapper := NewFilerStoreWrapper(newStubFilerStore())
ctx := context.Background()
entry := &Entry{
FullPath: util.FullPath("/test/obj"),
Attr: Attr{Mode: 0o660},
}
require.NoError(t, wrapper.InsertEntry(ctx, entry))
require.NoError(t, wrapper.UpdateEntry(ctx, entry))
require.NoError(t, wrapper.DeleteOneEntry(ctx, entry))
require.NoError(t, wrapper.InsertEntryKnownAbsent(ctx, entry))
require.NoError(t, wrapper.DeleteEntry(ctx, entry.FullPath))
require.NoError(t, wrapper.KvPut(ctx, []byte("k"), []byte("v")))
require.NoError(t, wrapper.KvDelete(ctx, []byte("k")))
txCtx, err := wrapper.BeginTransaction(ctx)
require.NoError(t, err)
require.NoError(t, wrapper.CommitTransaction(txCtx))
}
func TestFilerStoreWrapperReadOpsSucceedWithCancelledContext(t *testing.T) {
wrapper := NewFilerStoreWrapper(newStubFilerStore())
entry := &Entry{
FullPath: util.FullPath("/test/readable"),
Attr: Attr{Mode: 0o660},
}
require.NoError(t, wrapper.InsertEntry(context.Background(), entry))
require.NoError(t, wrapper.KvPut(context.Background(), []byte("rk"), []byte("rv")))
ctx := cancelledCtx()
_, err := wrapper.FindEntry(ctx, entry.FullPath)
assert.NoError(t, err)
_, err = wrapper.KvGet(ctx, []byte("rk"))
assert.NoError(t, err)
}
// RollbackTransaction must succeed even when the context is cancelled or
// expired, because it is a cleanup operation called after failures.
func TestFilerStoreWrapperRollbackSucceedsWithCancelledContext(t *testing.T) {
wrapper := NewFilerStoreWrapper(newStubFilerStore())
assert.NoError(t, wrapper.RollbackTransaction(cancelledCtx()))
assert.NoError(t, wrapper.RollbackTransaction(expiredCtx()))
}
-37
View File
@@ -2,7 +2,6 @@ package filer
import (
"context"
"sync"
"github.com/seaweedfs/seaweedfs/weed/glog"
"github.com/seaweedfs/seaweedfs/weed/pb/filer_pb"
@@ -36,39 +35,3 @@ func Replay(filerStore FilerStore, resp *filer_pb.SubscribeMetadataResponse) err
return nil
}
// ParallelProcessDirectoryStructure processes each entry in parallel, and also ensure parent directories are processed first.
// This also assumes the parent directories are in the entryChan already.
func ParallelProcessDirectoryStructure(entryChan chan *Entry, concurrency int, eachEntryFn func(entry *Entry) error) (firstErr error) {
executors := util.NewLimitedConcurrentExecutor(concurrency)
var wg sync.WaitGroup
for entry := range entryChan {
wg.Add(1)
if entry.IsDirectory() {
func() {
defer wg.Done()
if err := eachEntryFn(entry); err != nil {
if firstErr == nil {
firstErr = err
}
}
}()
} else {
executors.Execute(func() {
defer wg.Done()
if err := eachEntryFn(entry); err != nil {
if firstErr == nil {
firstErr = err
}
}
})
}
if firstErr != nil {
break
}
}
wg.Wait()
return
}
+3 -1
View File
@@ -173,14 +173,16 @@ func (store *UniversalRedisStore) ListDirectoryEntries(ctx context.Context, dirP
members = members[:limit]
}
var entry *filer.Entry
// fetch entry meta
for _, fileName := range members {
path := util.NewFullPath(string(dirPath), fileName)
entry, err := store.FindEntry(ctx, path)
entry, err = store.FindEntry(ctx, path)
lastFileName = fileName
if err != nil {
glog.V(0).InfofCtx(ctx, "list %s : %v", path, err)
if err == filer_pb.ErrNotFound {
err = nil
continue
}
} else {
-9
View File
@@ -16,15 +16,6 @@ type ItemList struct {
prefix string
}
func newItemList(client redis.UniversalClient, prefix string, store skiplist.ListStore, batchSize int) *ItemList {
return &ItemList{
skipList: skiplist.New(store),
batchSize: batchSize,
client: client,
prefix: prefix,
}
}
/*
Be reluctant to create new nodes. Try to fit into either previous node or next node.
Prefer to add to previous node.
@@ -1,48 +0,0 @@
package redis_lua
import (
"github.com/redis/go-redis/v9"
"github.com/seaweedfs/seaweedfs/weed/filer"
"github.com/seaweedfs/seaweedfs/weed/util"
)
func init() {
filer.Stores = append(filer.Stores, &RedisLuaClusterStore{})
}
type RedisLuaClusterStore struct {
UniversalRedisLuaStore
}
func (store *RedisLuaClusterStore) GetName() string {
return "redis_lua_cluster"
}
func (store *RedisLuaClusterStore) Initialize(configuration util.Configuration, prefix string) (err error) {
configuration.SetDefault(prefix+"useReadOnly", false)
configuration.SetDefault(prefix+"routeByLatency", false)
return store.initialize(
configuration.GetStringSlice(prefix+"addresses"),
configuration.GetString(prefix+"username"),
configuration.GetString(prefix+"password"),
configuration.GetString(prefix+"keyPrefix"),
configuration.GetBool(prefix+"useReadOnly"),
configuration.GetBool(prefix+"routeByLatency"),
configuration.GetStringSlice(prefix+"superLargeDirectories"),
)
}
func (store *RedisLuaClusterStore) initialize(addresses []string, username string, password string, keyPrefix string, readOnly, routeByLatency bool, superLargeDirectories []string) (err error) {
store.Client = redis.NewClusterClient(&redis.ClusterOptions{
Addrs: addresses,
Username: username,
Password: password,
ReadOnly: readOnly,
RouteByLatency: routeByLatency,
})
store.keyPrefix = keyPrefix
store.loadSuperLargeDirectories(superLargeDirectories)
return
}
@@ -1,48 +0,0 @@
package redis_lua
import (
"time"
"github.com/redis/go-redis/v9"
"github.com/seaweedfs/seaweedfs/weed/filer"
"github.com/seaweedfs/seaweedfs/weed/util"
)
func init() {
filer.Stores = append(filer.Stores, &RedisLuaSentinelStore{})
}
type RedisLuaSentinelStore struct {
UniversalRedisLuaStore
}
func (store *RedisLuaSentinelStore) GetName() string {
return "redis_lua_sentinel"
}
func (store *RedisLuaSentinelStore) Initialize(configuration util.Configuration, prefix string) (err error) {
return store.initialize(
configuration.GetStringSlice(prefix+"addresses"),
configuration.GetString(prefix+"masterName"),
configuration.GetString(prefix+"username"),
configuration.GetString(prefix+"password"),
configuration.GetInt(prefix+"database"),
configuration.GetString(prefix+"keyPrefix"),
)
}
func (store *RedisLuaSentinelStore) initialize(addresses []string, masterName string, username string, password string, database int, keyPrefix string) (err error) {
store.Client = redis.NewFailoverClient(&redis.FailoverOptions{
MasterName: masterName,
SentinelAddrs: addresses,
Username: username,
Password: password,
DB: database,
MinRetryBackoff: time.Millisecond * 100,
MaxRetryBackoff: time.Minute * 1,
ReadTimeout: time.Second * 30,
WriteTimeout: time.Second * 5,
})
store.keyPrefix = keyPrefix
return
}
-42
View File
@@ -1,42 +0,0 @@
package redis_lua
import (
"github.com/redis/go-redis/v9"
"github.com/seaweedfs/seaweedfs/weed/filer"
"github.com/seaweedfs/seaweedfs/weed/util"
)
func init() {
filer.Stores = append(filer.Stores, &RedisLuaStore{})
}
type RedisLuaStore struct {
UniversalRedisLuaStore
}
func (store *RedisLuaStore) GetName() string {
return "redis_lua"
}
func (store *RedisLuaStore) Initialize(configuration util.Configuration, prefix string) (err error) {
return store.initialize(
configuration.GetString(prefix+"address"),
configuration.GetString(prefix+"username"),
configuration.GetString(prefix+"password"),
configuration.GetInt(prefix+"database"),
configuration.GetString(prefix+"keyPrefix"),
configuration.GetStringSlice(prefix+"superLargeDirectories"),
)
}
func (store *RedisLuaStore) initialize(hostPort string, username string, password string, database int, keyPrefix string, superLargeDirectories []string) (err error) {
store.Client = redis.NewClient(&redis.Options{
Addr: hostPort,
Username: username,
Password: password,
DB: database,
})
store.keyPrefix = keyPrefix
store.loadSuperLargeDirectories(superLargeDirectories)
return
}
@@ -1,19 +0,0 @@
-- KEYS[1]: full path of entry
local fullpath = KEYS[1]
-- KEYS[2]: full path of entry
local fullpath_list_key = KEYS[2]
-- KEYS[3]: dir of the entry
local dir_list_key = KEYS[3]
-- ARGV[1]: isSuperLargeDirectory
local isSuperLargeDirectory = ARGV[1] == "1"
-- ARGV[2]: name of the entry
local name = ARGV[2]
redis.call("DEL", fullpath, fullpath_list_key)
if not isSuperLargeDirectory and name ~= "" then
redis.call("ZREM", dir_list_key, name)
end
return 0
@@ -1,15 +0,0 @@
-- KEYS[1]: full path of entry
local fullpath = KEYS[1]
if fullpath ~= "" and string.sub(fullpath, -1) == "/" then
fullpath = string.sub(fullpath, 0, -2)
end
local files = redis.call("ZRANGE", fullpath .. "\0", "0", "-1")
for _, name in ipairs(files) do
local file_path = fullpath .. "/" .. name
redis.call("DEL", file_path, file_path .. "\0")
end
return 0
@@ -1,25 +0,0 @@
package stored_procedure
import (
_ "embed"
"github.com/redis/go-redis/v9"
)
func init() {
InsertEntryScript = redis.NewScript(insertEntry)
DeleteEntryScript = redis.NewScript(deleteEntry)
DeleteFolderChildrenScript = redis.NewScript(deleteFolderChildren)
}
//go:embed insert_entry.lua
var insertEntry string
var InsertEntryScript *redis.Script
//go:embed delete_entry.lua
var deleteEntry string
var DeleteEntryScript *redis.Script
//go:embed delete_folder_children.lua
var deleteFolderChildren string
var DeleteFolderChildrenScript *redis.Script
@@ -1,27 +0,0 @@
-- KEYS[1]: full path of entry
local full_path = KEYS[1]
-- KEYS[2]: dir of the entry
local dir_list_key = KEYS[2]
-- ARGV[1]: content of the entry
local entry = ARGV[1]
-- ARGV[2]: TTL of the entry
local ttlSec = tonumber(ARGV[2])
-- ARGV[3]: isSuperLargeDirectory
local isSuperLargeDirectory = ARGV[3] == "1"
-- ARGV[4]: zscore of the entry in zset
local zscore = tonumber(ARGV[4])
-- ARGV[5]: name of the entry
local name = ARGV[5]
if ttlSec > 0 then
redis.call("SET", full_path, entry, "EX", ttlSec)
else
redis.call("SET", full_path, entry)
end
if not isSuperLargeDirectory and name ~= "" then
redis.call("ZADD", dir_list_key, "NX", zscore, name)
end
return 0
@@ -1,206 +0,0 @@
package redis_lua
import (
"context"
"fmt"
"time"
"github.com/redis/go-redis/v9"
"github.com/seaweedfs/seaweedfs/weed/filer"
"github.com/seaweedfs/seaweedfs/weed/filer/redis_lua/stored_procedure"
"github.com/seaweedfs/seaweedfs/weed/glog"
"github.com/seaweedfs/seaweedfs/weed/pb/filer_pb"
"github.com/seaweedfs/seaweedfs/weed/util"
)
const (
DIR_LIST_MARKER = "\x00"
)
type UniversalRedisLuaStore struct {
Client redis.UniversalClient
keyPrefix string
superLargeDirectoryHash map[string]bool
}
func (store *UniversalRedisLuaStore) isSuperLargeDirectory(dir string) (isSuperLargeDirectory bool) {
_, isSuperLargeDirectory = store.superLargeDirectoryHash[dir]
return
}
func (store *UniversalRedisLuaStore) loadSuperLargeDirectories(superLargeDirectories []string) {
// set directory hash
store.superLargeDirectoryHash = make(map[string]bool)
for _, dir := range superLargeDirectories {
store.superLargeDirectoryHash[dir] = true
}
}
func (store *UniversalRedisLuaStore) getKey(key string) string {
if store.keyPrefix == "" {
return key
}
return store.keyPrefix + key
}
func (store *UniversalRedisLuaStore) BeginTransaction(ctx context.Context) (context.Context, error) {
return ctx, nil
}
func (store *UniversalRedisLuaStore) CommitTransaction(ctx context.Context) error {
return nil
}
func (store *UniversalRedisLuaStore) RollbackTransaction(ctx context.Context) error {
return nil
}
func (store *UniversalRedisLuaStore) InsertEntry(ctx context.Context, entry *filer.Entry) (err error) {
value, err := entry.EncodeAttributesAndChunks()
if err != nil {
return fmt.Errorf("encoding %s %+v: %v", entry.FullPath, entry.Attr, err)
}
if len(entry.GetChunks()) > filer.CountEntryChunksForGzip {
value = util.MaybeGzipData(value)
}
dir, name := entry.FullPath.DirAndName()
err = stored_procedure.InsertEntryScript.Run(ctx, store.Client,
[]string{store.getKey(string(entry.FullPath)), store.getKey(genDirectoryListKey(dir))},
value, entry.TtlSec,
store.isSuperLargeDirectory(dir), 0, name).Err()
if err != nil {
return fmt.Errorf("persisting %s : %v", entry.FullPath, err)
}
return nil
}
func (store *UniversalRedisLuaStore) UpdateEntry(ctx context.Context, entry *filer.Entry) (err error) {
return store.InsertEntry(ctx, entry)
}
func (store *UniversalRedisLuaStore) FindEntry(ctx context.Context, fullpath util.FullPath) (entry *filer.Entry, err error) {
data, err := store.Client.Get(ctx, store.getKey(string(fullpath))).Result()
if err == redis.Nil {
return nil, filer_pb.ErrNotFound
}
if err != nil {
return nil, fmt.Errorf("get %s : %v", fullpath, err)
}
entry = &filer.Entry{
FullPath: fullpath,
}
err = entry.DecodeAttributesAndChunks(util.MaybeDecompressData([]byte(data)))
if err != nil {
return entry, fmt.Errorf("decode %s : %v", entry.FullPath, err)
}
return entry, nil
}
func (store *UniversalRedisLuaStore) DeleteEntry(ctx context.Context, fullpath util.FullPath) (err error) {
dir, name := fullpath.DirAndName()
err = stored_procedure.DeleteEntryScript.Run(ctx, store.Client,
[]string{store.getKey(string(fullpath)), store.getKey(genDirectoryListKey(string(fullpath))), store.getKey(genDirectoryListKey(dir))},
store.isSuperLargeDirectory(dir), name).Err()
if err != nil {
return fmt.Errorf("DeleteEntry %s : %v", fullpath, err)
}
return nil
}
func (store *UniversalRedisLuaStore) DeleteFolderChildren(ctx context.Context, fullpath util.FullPath) (err error) {
if store.isSuperLargeDirectory(string(fullpath)) {
return nil
}
err = stored_procedure.DeleteFolderChildrenScript.Run(ctx, store.Client,
[]string{store.getKey(string(fullpath))}).Err()
if err != nil {
return fmt.Errorf("DeleteFolderChildren %s : %v", fullpath, err)
}
return nil
}
func (store *UniversalRedisLuaStore) ListDirectoryPrefixedEntries(ctx context.Context, dirPath util.FullPath, startFileName string, includeStartFile bool, limit int64, prefix string, eachEntryFunc filer.ListEachEntryFunc) (lastFileName string, err error) {
return lastFileName, filer.ErrUnsupportedListDirectoryPrefixed
}
func (store *UniversalRedisLuaStore) ListDirectoryEntries(ctx context.Context, dirPath util.FullPath, startFileName string, includeStartFile bool, limit int64, eachEntryFunc filer.ListEachEntryFunc) (lastFileName string, err error) {
dirListKey := store.getKey(genDirectoryListKey(string(dirPath)))
min := "-"
if startFileName != "" {
if includeStartFile {
min = "[" + startFileName
} else {
min = "(" + startFileName
}
}
members, err := store.Client.ZRangeByLex(ctx, dirListKey, &redis.ZRangeBy{
Min: min,
Max: "+",
Offset: 0,
Count: limit,
}).Result()
if err != nil {
return lastFileName, fmt.Errorf("list %s : %v", dirPath, err)
}
// fetch entry meta
for _, fileName := range members {
path := util.NewFullPath(string(dirPath), fileName)
entry, err := store.FindEntry(ctx, path)
lastFileName = fileName
if err != nil {
glog.V(0).InfofCtx(ctx, "list %s : %v", path, err)
if err == filer_pb.ErrNotFound {
continue
}
} else {
if entry.TtlSec > 0 {
if entry.Attr.Crtime.Add(time.Duration(entry.TtlSec) * time.Second).Before(time.Now()) {
store.DeleteEntry(ctx, path)
continue
}
}
resEachEntryFunc, resEachEntryFuncErr := eachEntryFunc(entry)
if resEachEntryFuncErr != nil {
err = fmt.Errorf("failed to process eachEntryFunc: %w", resEachEntryFuncErr)
break
}
if !resEachEntryFunc {
break
}
}
}
return lastFileName, err
}
func genDirectoryListKey(dir string) (dirList string) {
return dir + DIR_LIST_MARKER
}
func (store *UniversalRedisLuaStore) Shutdown() {
store.Client.Close()
}
@@ -1,42 +0,0 @@
package redis_lua
import (
"context"
"fmt"
"github.com/redis/go-redis/v9"
"github.com/seaweedfs/seaweedfs/weed/filer"
)
func (store *UniversalRedisLuaStore) KvPut(ctx context.Context, key []byte, value []byte) (err error) {
_, err = store.Client.Set(ctx, string(key), value, 0).Result()
if err != nil {
return fmt.Errorf("kv put: %w", err)
}
return nil
}
func (store *UniversalRedisLuaStore) KvGet(ctx context.Context, key []byte) (value []byte, err error) {
data, err := store.Client.Get(ctx, string(key)).Result()
if err == redis.Nil {
return nil, filer.ErrKvNotFound
}
return []byte(data), err
}
func (store *UniversalRedisLuaStore) KvDelete(ctx context.Context, key []byte) (err error) {
_, err = store.Client.Del(ctx, string(key)).Result()
if err != nil {
return fmt.Errorf("kv delete: %w", err)
}
return nil
}
+60 -31
View File
@@ -102,10 +102,6 @@ func PrepareStreamContent(masterClient wdclient.HasLookupFileIdFunction, jwtFunc
type VolumeServerJwtFunction func(fileId string) string
func noJwtFunc(string) string {
return ""
}
type CacheInvalidator interface {
InvalidateCache(fileId string)
}
@@ -249,6 +245,66 @@ func PrepareStreamContentWithThrottler(ctx context.Context, masterClient wdclien
}, nil
}
// PrepareStreamContentWithPrefetch is like PrepareStreamContentWithThrottler but uses
// concurrent chunk prefetching to overlap network I/O. When prefetchAhead > 1, fetch
// goroutines establish HTTP connections to volume servers ahead of time, streaming data
// through io.Pipe with minimal memory overhead.
//
// prefetchAhead controls the number of chunks fetched concurrently:
// - 0 or 1: falls back to sequential fetching (same as PrepareStreamContentWithThrottler)
// - 2+: uses pipe-based prefetch pipeline with that many concurrent fetches
func PrepareStreamContentWithPrefetch(ctx context.Context, masterClient wdclient.HasLookupFileIdFunction, jwtFunc VolumeServerJwtFunction, chunks []*filer_pb.FileChunk, offset int64, size int64, downloadMaxBytesPs int64, prefetchAhead int) (DoStreamContent, error) {
if prefetchAhead <= 1 {
return PrepareStreamContentWithThrottler(ctx, masterClient, jwtFunc, chunks, offset, size, downloadMaxBytesPs)
}
glog.V(4).InfofCtx(ctx, "prepare to stream content with prefetch=%d for chunks: %d", prefetchAhead, len(chunks))
chunkViews := ViewFromChunks(ctx, masterClient.GetLookupFileIdFunction(), chunks, offset, size)
fileId2Url := make(map[string][]string)
for x := chunkViews.Front(); x != nil; x = x.Next {
chunkView := x.Value
var urlStrings []string
var err error
for _, backoff := range getLookupFileIdBackoffSchedule {
if err := ctx.Err(); err != nil {
return nil, err
}
urlStrings, err = masterClient.GetLookupFileIdFunction()(ctx, chunkView.FileId)
if err == nil && len(urlStrings) > 0 {
break
}
if err := ctx.Err(); err != nil {
return nil, err
}
glog.V(4).InfofCtx(ctx, "waiting for chunk: %s", chunkView.FileId)
timer := time.NewTimer(backoff)
select {
case <-ctx.Done():
if !timer.Stop() {
<-timer.C
}
return nil, ctx.Err()
case <-timer.C:
}
}
if err != nil {
glog.V(1).InfofCtx(ctx, "operation LookupFileId %s failed, err: %v", chunkView.FileId, err)
return nil, err
} else if len(urlStrings) == 0 {
errUrlNotFound := fmt.Errorf("operation LookupFileId %s failed, err: urls not found", chunkView.FileId)
glog.ErrorCtx(ctx, errUrlNotFound)
return nil, errUrlNotFound
}
fileId2Url[chunkView.FileId] = urlStrings
}
return func(writer io.Writer) error {
return streamChunksPrefetched(ctx, writer, chunkViews, fileId2Url, jwtFunc, masterClient, offset, size, downloadMaxBytesPs, prefetchAhead)
}, nil
}
func StreamContent(masterClient wdclient.HasLookupFileIdFunction, writer io.Writer, chunks []*filer_pb.FileChunk, offset int64, size int64) error {
streamFn, err := PrepareStreamContent(masterClient, JwtForVolumeServer, chunks, offset, size)
if err != nil {
@@ -276,33 +332,6 @@ func writeZero(w io.Writer, size int64) (err error) {
return
}
func ReadAll(ctx context.Context, buffer []byte, masterClient *wdclient.MasterClient, chunks []*filer_pb.FileChunk) error {
lookupFileIdFn := func(ctx context.Context, fileId string) (targetUrls []string, err error) {
return masterClient.LookupFileId(ctx, fileId)
}
chunkViews := ViewFromChunks(ctx, lookupFileIdFn, chunks, 0, int64(len(buffer)))
idx := 0
for x := chunkViews.Front(); x != nil; x = x.Next {
chunkView := x.Value
urlStrings, err := lookupFileIdFn(ctx, chunkView.FileId)
if err != nil {
glog.V(1).InfofCtx(ctx, "operation LookupFileId %s failed, err: %v", chunkView.FileId, err)
return err
}
n, err := util_http.RetriedFetchChunkData(ctx, buffer[idx:idx+int(chunkView.ViewSize)], urlStrings, chunkView.CipherKey, chunkView.IsGzipped, chunkView.IsFullChunk(), chunkView.OffsetInChunk, chunkView.FileId)
if err != nil {
return err
}
idx += n
}
return nil
}
// ---------------- ChunkStreamReader ----------------------------------
type ChunkStreamReader struct {
head *Interval[*ChunkView]
+317
View File
@@ -0,0 +1,317 @@
package filer
import (
"bytes"
"context"
"fmt"
"io"
"math/rand"
"net/http"
"net/http/httptest"
"os"
"strings"
"sync"
"testing"
"time"
util_http "github.com/seaweedfs/seaweedfs/weed/util/http"
"github.com/seaweedfs/seaweedfs/weed/pb/filer_pb"
"github.com/seaweedfs/seaweedfs/weed/wdclient"
)
func TestMain(m *testing.M) {
// Initialize the global HTTP client required by ReadUrlAsStream
util_http.InitGlobalHttpClient()
os.Exit(m.Run())
}
// mockMasterClientForBenchmark implements HasLookupFileIdFunction and CacheInvalidator
type mockMasterClientForBenchmark struct {
urls map[string][]string
}
func (m *mockMasterClientForBenchmark) GetLookupFileIdFunction() wdclient.LookupFileIdFunctionType {
return func(ctx context.Context, fileId string) ([]string, error) {
if urls, ok := m.urls[fileId]; ok {
return urls, nil
}
return nil, fmt.Errorf("fileId %s not found", fileId)
}
}
func (m *mockMasterClientForBenchmark) InvalidateCache(fileId string) {}
// noopJwtFunc returns empty JWT for testing
func noopJwtFunc(fileId string) string {
return ""
}
// createMockVolumeServer creates an httptest server that serves chunk data
// with configurable per-request latency to simulate network conditions.
// The latency is applied once per request (simulating RTT), not per byte.
func createMockVolumeServer(chunkData map[string][]byte, latency time.Duration) *httptest.Server {
var mu sync.RWMutex
return httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
// Simulate network latency (connection setup + RTT)
if latency > 0 {
time.Sleep(latency)
}
// Extract fileId from path (e.g., "/1,abc123")
path := r.URL.Path
if strings.HasPrefix(path, "/") {
path = path[1:]
}
mu.RLock()
data, ok := chunkData[path]
mu.RUnlock()
if !ok {
http.Error(w, "not found", http.StatusNotFound)
return
}
// Handle Range header
rangeHeader := r.Header.Get("Range")
if rangeHeader != "" {
var start, end int64
fmt.Sscanf(rangeHeader, "bytes=%d-%d", &start, &end)
if start >= 0 && end < int64(len(data)) && start <= end {
w.Header().Set("Content-Length", fmt.Sprintf("%d", end-start+1))
w.WriteHeader(http.StatusPartialContent)
w.Write(data[start : end+1])
return
}
}
w.Header().Set("Content-Length", fmt.Sprintf("%d", len(data)))
w.WriteHeader(http.StatusOK)
w.Write(data)
}))
}
// benchmarkConfig holds parameters for a single benchmark scenario
type benchmarkConfig struct {
numChunks int
chunkSize int
latency time.Duration
prefetch int // 0 = sequential
}
func (c benchmarkConfig) name() string {
name := fmt.Sprintf("chunks=%d/size=%dKB/latency=%dms",
c.numChunks, c.chunkSize/1024, c.latency.Milliseconds())
if c.prefetch > 0 {
name += fmt.Sprintf("/prefetch=%d", c.prefetch)
}
return name
}
// setupBenchmark creates mock infrastructure and returns chunks, master client, and cleanup func
func setupBenchmark(b *testing.B, cfg benchmarkConfig) ([]*filer_pb.FileChunk, *mockMasterClientForBenchmark, func()) {
b.Helper()
// Generate random chunk data
chunkData := make(map[string][]byte, cfg.numChunks)
chunks := make([]*filer_pb.FileChunk, cfg.numChunks)
for i := 0; i < cfg.numChunks; i++ {
fileId := fmt.Sprintf("1,%x", i)
data := make([]byte, cfg.chunkSize)
rand.Read(data)
chunkData[fileId] = data
chunks[i] = &filer_pb.FileChunk{
FileId: fileId,
Offset: int64(i * cfg.chunkSize),
Size: uint64(cfg.chunkSize),
ModifiedTsNs: int64(i),
Fid: &filer_pb.FileId{FileKey: uint64(i)},
}
}
// Start mock volume server
server := createMockVolumeServer(chunkData, cfg.latency)
// Build URL map
urls := make(map[string][]string, cfg.numChunks)
for i := 0; i < cfg.numChunks; i++ {
fileId := fmt.Sprintf("1,%x", i)
urls[fileId] = []string{server.URL + "/" + fileId}
}
masterClient := &mockMasterClientForBenchmark{urls: urls}
cleanup := func() { server.Close() }
return chunks, masterClient, cleanup
}
// runSequentialBenchmark runs the current sequential streaming path
func runSequentialBenchmark(b *testing.B, cfg benchmarkConfig) {
chunks, masterClient, cleanup := setupBenchmark(b, cfg)
defer cleanup()
totalSize := int64(cfg.numChunks * cfg.chunkSize)
b.ResetTimer()
b.SetBytes(totalSize)
for i := 0; i < b.N; i++ {
streamFn, err := PrepareStreamContentWithThrottler(
context.Background(),
masterClient,
noopJwtFunc,
chunks,
0,
totalSize,
0, // no throttle
)
if err != nil {
b.Fatal(err)
}
if err := streamFn(io.Discard); err != nil {
b.Fatal(err)
}
}
}
// BenchmarkStreamSequential benchmarks the current sequential streaming path.
// This provides the BEFORE baseline for comparison.
func BenchmarkStreamSequential(b *testing.B) {
configs := []benchmarkConfig{
// Pure throughput (no latency)
{numChunks: 16, chunkSize: 64 * 1024, latency: 0},
{numChunks: 64, chunkSize: 64 * 1024, latency: 0},
// Moderate latency — shows RTT gap overhead
{numChunks: 16, chunkSize: 64 * 1024, latency: 5 * time.Millisecond},
{numChunks: 64, chunkSize: 64 * 1024, latency: 5 * time.Millisecond},
// High latency — significant RTT overhead
{numChunks: 16, chunkSize: 64 * 1024, latency: 20 * time.Millisecond},
{numChunks: 64, chunkSize: 64 * 1024, latency: 10 * time.Millisecond},
}
for _, cfg := range configs {
b.Run(cfg.name(), func(b *testing.B) {
runSequentialBenchmark(b, cfg)
})
}
}
// BenchmarkStreamSequentialVerify is a quick functional test that the benchmark
// infrastructure works correctly — ensures data integrity through the pipeline.
func BenchmarkStreamSequentialVerify(b *testing.B) {
cfg := benchmarkConfig{numChunks: 4, chunkSize: 1024, latency: 0}
chunks, masterClient, cleanup := setupBenchmark(b, cfg)
defer cleanup()
totalSize := int64(cfg.numChunks * cfg.chunkSize)
streamFn, err := PrepareStreamContentWithThrottler(
context.Background(),
masterClient,
noopJwtFunc,
chunks,
0,
totalSize,
0,
)
if err != nil {
b.Fatal(err)
}
var buf bytes.Buffer
if err := streamFn(&buf); err != nil {
b.Fatal(err)
}
if buf.Len() != int(totalSize) {
b.Fatalf("expected %d bytes, got %d", totalSize, buf.Len())
}
}
// runPrefetchBenchmark runs the new prefetch streaming path
func runPrefetchBenchmark(b *testing.B, cfg benchmarkConfig) {
chunks, masterClient, cleanup := setupBenchmark(b, cfg)
defer cleanup()
totalSize := int64(cfg.numChunks * cfg.chunkSize)
b.ResetTimer()
b.SetBytes(totalSize)
for i := 0; i < b.N; i++ {
streamFn, err := PrepareStreamContentWithPrefetch(
context.Background(),
masterClient,
noopJwtFunc,
chunks,
0,
totalSize,
0, // no throttle
cfg.prefetch,
)
if err != nil {
b.Fatal(err)
}
if err := streamFn(io.Discard); err != nil {
b.Fatal(err)
}
}
}
// BenchmarkStreamPrefetch benchmarks the new prefetch streaming path.
// Compare against BenchmarkStreamSequential for the AFTER measurement.
func BenchmarkStreamPrefetch(b *testing.B) {
configs := []benchmarkConfig{
// Pure throughput (no latency) — should be similar to sequential
{numChunks: 16, chunkSize: 64 * 1024, latency: 0, prefetch: 4},
{numChunks: 64, chunkSize: 64 * 1024, latency: 0, prefetch: 4},
// Moderate latency — prefetch should eliminate most RTT overhead
{numChunks: 16, chunkSize: 64 * 1024, latency: 5 * time.Millisecond, prefetch: 4},
{numChunks: 64, chunkSize: 64 * 1024, latency: 5 * time.Millisecond, prefetch: 4},
// High latency — most benefit from prefetch
{numChunks: 16, chunkSize: 64 * 1024, latency: 20 * time.Millisecond, prefetch: 4},
{numChunks: 64, chunkSize: 64 * 1024, latency: 10 * time.Millisecond, prefetch: 4},
// Vary prefetch count with moderate latency
{numChunks: 64, chunkSize: 64 * 1024, latency: 5 * time.Millisecond, prefetch: 2},
{numChunks: 64, chunkSize: 64 * 1024, latency: 5 * time.Millisecond, prefetch: 8},
}
for _, cfg := range configs {
b.Run(cfg.name(), func(b *testing.B) {
runPrefetchBenchmark(b, cfg)
})
}
}
// BenchmarkStreamPrefetchVerify verifies data integrity through the prefetch pipeline.
func BenchmarkStreamPrefetchVerify(b *testing.B) {
cfg := benchmarkConfig{numChunks: 4, chunkSize: 1024, latency: 0, prefetch: 4}
chunks, masterClient, cleanup := setupBenchmark(b, cfg)
defer cleanup()
totalSize := int64(cfg.numChunks * cfg.chunkSize)
streamFn, err := PrepareStreamContentWithPrefetch(
context.Background(),
masterClient,
noopJwtFunc,
chunks,
0,
totalSize,
0,
cfg.prefetch,
)
if err != nil {
b.Fatal(err)
}
var buf bytes.Buffer
if err := streamFn(&buf); err != nil {
b.Fatal(err)
}
if buf.Len() != int(totalSize) {
b.Fatalf("expected %d bytes, got %d", totalSize, buf.Len())
}
}
-281
View File
@@ -1,281 +0,0 @@
package filer
import (
"bytes"
"context"
"errors"
"testing"
"time"
"github.com/seaweedfs/seaweedfs/weed/pb/filer_pb"
"github.com/seaweedfs/seaweedfs/weed/wdclient"
)
// mockMasterClient implements HasLookupFileIdFunction and CacheInvalidator
type mockMasterClient struct {
lookupFunc func(ctx context.Context, fileId string) ([]string, error)
invalidatedFileIds []string
}
func (m *mockMasterClient) GetLookupFileIdFunction() wdclient.LookupFileIdFunctionType {
return m.lookupFunc
}
func (m *mockMasterClient) InvalidateCache(fileId string) {
m.invalidatedFileIds = append(m.invalidatedFileIds, fileId)
}
// Test urlSlicesEqual helper function
func TestUrlSlicesEqual(t *testing.T) {
tests := []struct {
name string
a []string
b []string
expected bool
}{
{
name: "identical slices",
a: []string{"http://server1", "http://server2"},
b: []string{"http://server1", "http://server2"},
expected: true,
},
{
name: "same URLs different order",
a: []string{"http://server1", "http://server2"},
b: []string{"http://server2", "http://server1"},
expected: true,
},
{
name: "different URLs",
a: []string{"http://server1", "http://server2"},
b: []string{"http://server1", "http://server3"},
expected: false,
},
{
name: "different lengths",
a: []string{"http://server1"},
b: []string{"http://server1", "http://server2"},
expected: false,
},
{
name: "empty slices",
a: []string{},
b: []string{},
expected: true,
},
{
name: "duplicates in both",
a: []string{"http://server1", "http://server1"},
b: []string{"http://server1", "http://server1"},
expected: true,
},
{
name: "different duplicate counts",
a: []string{"http://server1", "http://server1"},
b: []string{"http://server1", "http://server2"},
expected: false,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
result := urlSlicesEqual(tt.a, tt.b)
if result != tt.expected {
t.Errorf("urlSlicesEqual(%v, %v) = %v; want %v", tt.a, tt.b, result, tt.expected)
}
})
}
}
// Test cache invalidation when read fails
func TestStreamContentWithCacheInvalidation(t *testing.T) {
ctx := context.Background()
fileId := "3,01234567890"
callCount := 0
oldUrls := []string{"http://failed-server:8080"}
newUrls := []string{"http://working-server:8080"}
mock := &mockMasterClient{
lookupFunc: func(ctx context.Context, fid string) ([]string, error) {
callCount++
if callCount == 1 {
// First call returns failing server
return oldUrls, nil
}
// After invalidation, return working server
return newUrls, nil
},
}
// Create a simple chunk
chunks := []*filer_pb.FileChunk{
{
FileId: fileId,
Offset: 0,
Size: 10,
},
}
streamFn, err := PrepareStreamContentWithThrottler(ctx, mock, noJwtFunc, chunks, 0, 10, 0)
if err != nil {
t.Fatalf("PrepareStreamContentWithThrottler failed: %v", err)
}
// Note: This test can't fully execute streamFn because it would require actual HTTP servers
// However, we can verify the setup was created correctly
if streamFn == nil {
t.Fatal("Expected non-nil stream function")
}
// Verify the lookup was called
if callCount != 1 {
t.Errorf("Expected 1 lookup call, got %d", callCount)
}
}
// Test that InvalidateCache is called on read failure
func TestCacheInvalidationInterface(t *testing.T) {
mock := &mockMasterClient{
lookupFunc: func(ctx context.Context, fileId string) ([]string, error) {
return []string{"http://server:8080"}, nil
},
}
fileId := "3,test123"
// Simulate invalidation
if invalidator, ok := interface{}(mock).(CacheInvalidator); ok {
invalidator.InvalidateCache(fileId)
} else {
t.Fatal("mockMasterClient should implement CacheInvalidator")
}
// Check that the file ID was recorded as invalidated
if len(mock.invalidatedFileIds) != 1 {
t.Fatalf("Expected 1 invalidated file ID, got %d", len(mock.invalidatedFileIds))
}
if mock.invalidatedFileIds[0] != fileId {
t.Errorf("Expected invalidated file ID %s, got %s", fileId, mock.invalidatedFileIds[0])
}
}
// Test retry logic doesn't retry with same URLs
func TestRetryLogicSkipsSameUrls(t *testing.T) {
// This test verifies that the urlSlicesEqual check prevents infinite retries
sameUrls := []string{"http://server1:8080", "http://server2:8080"}
differentUrls := []string{"http://server3:8080", "http://server4:8080"}
// Same URLs should return true (and thus skip retry)
if !urlSlicesEqual(sameUrls, sameUrls) {
t.Error("Expected same URLs to be equal")
}
// Different URLs should return false (and thus allow retry)
if urlSlicesEqual(sameUrls, differentUrls) {
t.Error("Expected different URLs to not be equal")
}
}
func TestCanceledStreamSkipsCacheInvalidation(t *testing.T) {
ctx, cancel := context.WithCancel(context.Background())
fileId := "3,canceled"
mock := &mockMasterClient{
lookupFunc: func(ctx context.Context, fid string) ([]string, error) {
return []string{"http://server:8080"}, nil
},
}
chunks := []*filer_pb.FileChunk{
{
FileId: fileId,
Offset: 0,
Size: 10,
},
}
streamFn, err := PrepareStreamContentWithThrottler(ctx, mock, noJwtFunc, chunks, 0, 10, 0)
if err != nil {
t.Fatalf("PrepareStreamContentWithThrottler failed: %v", err)
}
cancel()
err = streamFn(&bytes.Buffer{})
if err != context.Canceled {
t.Fatalf("expected context.Canceled, got %v", err)
}
if len(mock.invalidatedFileIds) != 0 {
t.Fatalf("expected no cache invalidation on cancellation, got %v", mock.invalidatedFileIds)
}
}
func TestPrepareStreamContentSkipsLookupWhenContextAlreadyCanceled(t *testing.T) {
oldSchedule := getLookupFileIdBackoffSchedule
getLookupFileIdBackoffSchedule = []time.Duration{time.Millisecond}
t.Cleanup(func() {
getLookupFileIdBackoffSchedule = oldSchedule
})
ctx, cancel := context.WithCancel(context.Background())
cancel()
lookupCalls := 0
mock := &mockMasterClient{
lookupFunc: func(ctx context.Context, fileId string) ([]string, error) {
lookupCalls++
return nil, errors.New("lookup should not run")
},
}
chunks := []*filer_pb.FileChunk{
{
FileId: "3,precanceled",
Offset: 0,
Size: 10,
},
}
_, err := PrepareStreamContentWithThrottler(ctx, mock, noJwtFunc, chunks, 0, 10, 0)
if !errors.Is(err, context.Canceled) {
t.Fatalf("expected context.Canceled, got %v", err)
}
if lookupCalls != 0 {
t.Fatalf("expected no lookup calls after cancellation, got %d", lookupCalls)
}
}
func TestPrepareStreamContentStopsLookupRetriesAfterContextCancellation(t *testing.T) {
oldSchedule := getLookupFileIdBackoffSchedule
getLookupFileIdBackoffSchedule = []time.Duration{time.Millisecond, time.Millisecond, time.Millisecond}
t.Cleanup(func() {
getLookupFileIdBackoffSchedule = oldSchedule
})
ctx, cancel := context.WithCancel(context.Background())
lookupCalls := 0
mock := &mockMasterClient{
lookupFunc: func(ctx context.Context, fileId string) ([]string, error) {
lookupCalls++
cancel()
return nil, context.Canceled
},
}
chunks := []*filer_pb.FileChunk{
{
FileId: "3,cancel-during-lookup",
Offset: 0,
Size: 10,
},
}
_, err := PrepareStreamContentWithThrottler(ctx, mock, noJwtFunc, chunks, 0, 10, 0)
if !errors.Is(err, context.Canceled) {
t.Fatalf("expected context.Canceled, got %v", err)
}
if lookupCalls != 1 {
t.Fatalf("expected lookup retries to stop after cancellation, got %d calls", lookupCalls)
}
}
+274
View File
@@ -0,0 +1,274 @@
package filer
import (
"context"
"fmt"
"io"
"sync"
"time"
"github.com/seaweedfs/seaweedfs/weed/glog"
"github.com/seaweedfs/seaweedfs/weed/stats"
"github.com/seaweedfs/seaweedfs/weed/util"
"github.com/seaweedfs/seaweedfs/weed/util/mem"
"github.com/seaweedfs/seaweedfs/weed/wdclient"
)
// chunkPipeResult represents a prefetched chunk streaming through a pipe.
// The fetch goroutine writes data into the pipeWriter; the consumer reads from pipeReader.
type chunkPipeResult struct {
chunkView *ChunkView
reader *io.PipeReader
fetchErr error // final error from fetch goroutine
written int64 // bytes written by fetch goroutine
done chan struct{} // closed when fetch goroutine finishes
urlStrings []string // snapshot of URLs at dispatch time (for retry logic)
}
// streamChunksPrefetched streams chunks with concurrent prefetch using io.Pipe.
//
// For each chunk in file order, a goroutine is launched (bounded by a semaphore)
// that establishes an HTTP connection to the volume server and streams data through
// an io.Pipe. The consumer reads from pipes in order, writing to the response.
//
// Memory usage is minimal: pipes are synchronous (no buffering), and only one
// reusable copy buffer is allocated for the consumer.
func streamChunksPrefetched(
ctx context.Context,
writer io.Writer,
chunkViews *IntervalList[*ChunkView],
fileId2Url map[string][]string,
jwtFunc VolumeServerJwtFunction,
masterClient wdclient.HasLookupFileIdFunction,
offset int64,
size int64,
downloadMaxBytesPs int64,
prefetchAhead int,
) error {
downloadThrottler := util.NewWriteThrottler(downloadMaxBytesPs)
// Create a local cancellable context so the consumer can stop the producer
// and all in-flight fetch goroutines on error (e.g., client disconnect).
localCtx, localCancel := context.WithCancel(ctx)
defer localCancel()
// Ordered channel: one entry per chunk, in file order.
// Capacity = prefetchAhead so the producer can run ahead.
// Uses pointer to avoid copying the struct while fetch goroutines write to it.
results := make(chan *chunkPipeResult, prefetchAhead)
// Semaphore to limit concurrent fetch goroutines (and thus HTTP connections).
sem := make(chan struct{}, prefetchAhead)
// Producer: walks chunk list, launches fetch goroutines, sends results in order.
// The producer only reads from fileId2Url (populated before streaming starts),
// so there is no concurrent map access — the consumer never writes to it.
var producerWg sync.WaitGroup
producerWg.Add(1)
go func() {
defer producerWg.Done()
defer close(results)
for x := chunkViews.Front(); x != nil; x = x.Next {
chunkView := x.Value
// Check context before starting new fetch
select {
case <-localCtx.Done():
return
default:
}
// Acquire semaphore slot (bounds concurrent HTTP connections)
select {
case sem <- struct{}{}:
case <-localCtx.Done():
return
}
pr, pw := io.Pipe()
urlStrings := fileId2Url[chunkView.FileId]
jwt := jwtFunc(chunkView.FileId)
result := &chunkPipeResult{
chunkView: chunkView,
reader: pr,
done: make(chan struct{}),
urlStrings: urlStrings,
}
// Launch fetch goroutine
go func(cv *ChunkView, urls []string, jwt string, pw *io.PipeWriter, res *chunkPipeResult) {
defer func() { <-sem }() // release semaphore
defer close(res.done)
written, err := retriedStreamFetchChunkData(
localCtx, pw, urls, jwt,
cv.CipherKey, cv.IsGzipped, cv.IsFullChunk(),
cv.OffsetInChunk, int(cv.ViewSize),
)
res.written = written
res.fetchErr = err
if err != nil {
pw.CloseWithError(err)
} else {
pw.Close()
}
}(chunkView, urlStrings, jwt, pw, result)
// Send result to consumer (blocks if channel full, back-pressuring producer)
select {
case results <- result:
case <-localCtx.Done():
// Consumer gone; close the pipe and wait for the fetch goroutine
// to finish so we don't leak it (this result was never sent to
// the channel, so the drain loop won't handle it).
pr.Close()
<-result.done
return
}
}
}()
// Consumer: reads from results channel in order, writes to response writer.
// Use the SeaweedFS memory pool for the copy buffer to reduce GC pressure.
copyBuf := mem.Allocate(256 * 1024)
defer mem.Free(copyBuf)
remaining := size
var consumeErr error
for result := range results {
chunkView := result.chunkView
// Handle gap before this chunk (zero-fill)
if offset < chunkView.ViewOffset {
gap := chunkView.ViewOffset - offset
remaining -= gap
glog.V(4).InfofCtx(ctx, "prefetch zero [%d,%d)", offset, chunkView.ViewOffset)
if err := writeZero(writer, gap); err != nil {
consumeErr = fmt.Errorf("write zero [%d,%d): %w", offset, chunkView.ViewOffset, err)
result.reader.Close()
break
}
offset = chunkView.ViewOffset
}
// Stream chunk data from pipe to response
start := time.Now()
_, copyErr := io.CopyBuffer(writer, result.reader, copyBuf)
result.reader.Close()
// Wait for fetch goroutine to finish to get final error
<-result.done
// Determine the effective error
err := copyErr
if err == nil && result.fetchErr != nil && result.written == 0 {
err = result.fetchErr
}
// If the fetcher itself failed before writing any data, try cache invalidation
// + re-fetch (same as sequential path stream.go:197). We check result.fetchErr
// and result.written (not copied) to avoid wrongly retrying when the fetch
// succeeded but the response writer failed on the first write.
if result.fetchErr != nil && result.written == 0 {
if err := localCtx.Err(); err != nil {
consumeErr = err
break
}
retryErr := retryWithCacheInvalidation(localCtx, writer, chunkView, result.urlStrings, jwtFunc, masterClient)
if retryErr != nil {
stats.FilerHandlerCounter.WithLabelValues("chunkDownloadError").Inc()
consumeErr = fmt.Errorf("read chunk: %w", retryErr)
break
}
// Retry succeeded
err = nil
} else if err != nil {
if localCtx.Err() != nil {
consumeErr = localCtx.Err()
} else {
stats.FilerHandlerCounter.WithLabelValues("chunkDownloadError").Inc()
consumeErr = fmt.Errorf("read chunk: %w", err)
}
break
}
offset += int64(chunkView.ViewSize)
remaining -= int64(chunkView.ViewSize)
stats.FilerRequestHistogram.WithLabelValues("chunkDownload").Observe(time.Since(start).Seconds())
stats.FilerHandlerCounter.WithLabelValues("chunkDownload").Inc()
downloadThrottler.MaybeSlowdown(int64(chunkView.ViewSize))
}
// Cancel the local context to stop the producer and any in-flight fetchers early.
// This ensures goroutines don't linger after the consumer exits (e.g., on write error).
localCancel()
// Drain remaining results to close pipes and unblock fetch goroutines
for result := range results {
result.reader.Close()
<-result.done
}
// Wait for producer to finish
producerWg.Wait()
if consumeErr != nil {
return consumeErr
}
// Handle trailing zero-fill
if remaining > 0 {
glog.V(4).InfofCtx(ctx, "prefetch zero [%d,%d)", offset, offset+remaining)
if err := writeZero(writer, remaining); err != nil {
return fmt.Errorf("write zero [%d,%d): %w", offset, offset+remaining, err)
}
}
return nil
}
// retryWithCacheInvalidation attempts to re-fetch a chunk after invalidating the URL cache.
// This mirrors the retry logic in PrepareStreamContentWithThrottler's sequential path.
func retryWithCacheInvalidation(
ctx context.Context,
writer io.Writer,
chunkView *ChunkView,
oldUrlStrings []string,
jwtFunc VolumeServerJwtFunction,
masterClient wdclient.HasLookupFileIdFunction,
) error {
invalidator, ok := masterClient.(CacheInvalidator)
if !ok {
return fmt.Errorf("read chunk %s failed and no cache invalidator available", chunkView.FileId)
}
glog.V(0).InfofCtx(ctx, "prefetch read chunk %s failed, invalidating cache and retrying", chunkView.FileId)
invalidator.InvalidateCache(chunkView.FileId)
newUrlStrings, lookupErr := masterClient.GetLookupFileIdFunction()(ctx, chunkView.FileId)
if lookupErr != nil {
glog.WarningfCtx(ctx, "failed to re-lookup chunk %s after cache invalidation: %v", chunkView.FileId, lookupErr)
return fmt.Errorf("re-lookup chunk %s: %w", chunkView.FileId, lookupErr)
}
if len(newUrlStrings) == 0 {
glog.WarningfCtx(ctx, "re-lookup for chunk %s returned no locations, skipping retry", chunkView.FileId)
return fmt.Errorf("re-lookup chunk %s: no locations", chunkView.FileId)
}
if urlSlicesEqual(oldUrlStrings, newUrlStrings) {
glog.V(0).InfofCtx(ctx, "re-lookup returned same locations for chunk %s, skipping retry", chunkView.FileId)
return fmt.Errorf("read chunk %s failed, same locations after cache invalidation", chunkView.FileId)
}
glog.V(0).InfofCtx(ctx, "retrying read chunk %s with new locations: %v", chunkView.FileId, newUrlStrings)
jwt := jwtFunc(chunkView.FileId)
_, err := retriedStreamFetchChunkData(
ctx, writer, newUrlStrings, jwt,
chunkView.CipherKey, chunkView.IsGzipped, chunkView.IsFullChunk(),
chunkView.OffsetInChunk, int(chunkView.ViewSize),
)
return err
}
+365
View File
@@ -0,0 +1,365 @@
package filer
import (
"bytes"
"context"
"fmt"
"io"
"math/rand"
"net/http"
"net/http/httptest"
"strings"
"sync"
"sync/atomic"
"testing"
"time"
"github.com/seaweedfs/seaweedfs/weed/pb/filer_pb"
"github.com/seaweedfs/seaweedfs/weed/wdclient"
)
// testMasterClient implements HasLookupFileIdFunction and CacheInvalidator for tests
type testMasterClient struct {
urls map[string][]string
invalidatedCount int32
}
func (m *testMasterClient) GetLookupFileIdFunction() wdclient.LookupFileIdFunctionType {
return func(ctx context.Context, fileId string) ([]string, error) {
if urls, ok := m.urls[fileId]; ok {
return urls, nil
}
return nil, fmt.Errorf("fileId %s not found", fileId)
}
}
func (m *testMasterClient) InvalidateCache(fileId string) {
atomic.AddInt32(&m.invalidatedCount, 1)
}
func noopJwt(fileId string) string { return "" }
// createTestServer creates a mock volume server that serves chunk data.
// Supports Range header for partial chunk reads (exercising OffsetInChunk paths).
func createTestServer(chunkData map[string][]byte) *httptest.Server {
var mu sync.RWMutex
return httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
path := r.URL.Path
if strings.HasPrefix(path, "/") {
path = path[1:]
}
mu.RLock()
data, ok := chunkData[path]
mu.RUnlock()
if !ok {
http.Error(w, "not found", http.StatusNotFound)
return
}
// Handle Range header for partial chunk reads
rangeHeader := r.Header.Get("Range")
if rangeHeader != "" {
var start, end int64
if _, err := fmt.Sscanf(rangeHeader, "bytes=%d-%d", &start, &end); err == nil {
if start < 0 || end >= int64(len(data)) || start > end {
w.Header().Set("Content-Range", fmt.Sprintf("bytes */%d", len(data)))
w.WriteHeader(http.StatusRequestedRangeNotSatisfiable)
return
}
rangeData := data[start : end+1]
w.Header().Set("Content-Range", fmt.Sprintf("bytes %d-%d/%d", start, end, len(data)))
w.Header().Set("Content-Length", fmt.Sprintf("%d", len(rangeData)))
w.WriteHeader(http.StatusPartialContent)
w.Write(rangeData)
return
}
}
w.Header().Set("Content-Length", fmt.Sprintf("%d", len(data)))
w.WriteHeader(http.StatusOK)
w.Write(data)
}))
}
// makeChunksAndServer creates N chunks of given size, a mock server, and a master client
func makeChunksAndServer(t *testing.T, numChunks, chunkSize int) ([]*filer_pb.FileChunk, *testMasterClient, map[string][]byte, func()) {
t.Helper()
chunkData := make(map[string][]byte, numChunks)
chunks := make([]*filer_pb.FileChunk, numChunks)
for i := 0; i < numChunks; i++ {
fileId := fmt.Sprintf("1,%x", i)
data := make([]byte, chunkSize)
rand.Read(data)
chunkData[fileId] = data
chunks[i] = &filer_pb.FileChunk{
FileId: fileId,
Offset: int64(i * chunkSize),
Size: uint64(chunkSize),
ModifiedTsNs: int64(i),
Fid: &filer_pb.FileId{FileKey: uint64(i)},
}
}
server := createTestServer(chunkData)
urls := make(map[string][]string, numChunks)
for i := 0; i < numChunks; i++ {
fileId := fmt.Sprintf("1,%x", i)
urls[fileId] = []string{server.URL + "/" + fileId}
}
masterClient := &testMasterClient{urls: urls}
return chunks, masterClient, chunkData, func() { server.Close() }
}
// TestPrefetchInOrderDelivery verifies chunks are written to the output in correct file order
func TestPrefetchInOrderDelivery(t *testing.T) {
chunks, masterClient, chunkData, cleanup := makeChunksAndServer(t, 8, 4096)
defer cleanup()
totalSize := int64(8 * 4096)
streamFn, err := PrepareStreamContentWithPrefetch(
context.Background(), masterClient, noopJwt,
chunks, 0, totalSize, 0, 4,
)
if err != nil {
t.Fatal(err)
}
var buf bytes.Buffer
if err := streamFn(&buf); err != nil {
t.Fatal(err)
}
// Verify total size
if buf.Len() != int(totalSize) {
t.Fatalf("expected %d bytes, got %d", totalSize, buf.Len())
}
// Verify data matches chunk-by-chunk in order
result := buf.Bytes()
for i := 0; i < 8; i++ {
fileId := fmt.Sprintf("1,%x", i)
expected := chunkData[fileId]
got := result[i*4096 : (i+1)*4096]
if !bytes.Equal(expected, got) {
t.Fatalf("chunk %d (%s) data mismatch at offset %d", i, fileId, i*4096)
}
}
}
// TestPrefetchSingleChunk verifies the pipeline works with just one chunk
func TestPrefetchSingleChunk(t *testing.T) {
chunks, masterClient, chunkData, cleanup := makeChunksAndServer(t, 1, 8192)
defer cleanup()
streamFn, err := PrepareStreamContentWithPrefetch(
context.Background(), masterClient, noopJwt,
chunks, 0, 8192, 0, 4,
)
if err != nil {
t.Fatal(err)
}
var buf bytes.Buffer
if err := streamFn(&buf); err != nil {
t.Fatal(err)
}
expected := chunkData["1,0"]
if !bytes.Equal(expected, buf.Bytes()) {
t.Fatal("single chunk data mismatch")
}
}
// TestPrefetchFallbackToSequential verifies prefetch=1 falls back to sequential path
func TestPrefetchFallbackToSequential(t *testing.T) {
chunks, masterClient, chunkData, cleanup := makeChunksAndServer(t, 4, 1024)
defer cleanup()
totalSize := int64(4 * 1024)
streamFn, err := PrepareStreamContentWithPrefetch(
context.Background(), masterClient, noopJwt,
chunks, 0, totalSize, 0, 1, // prefetch=1 -> sequential
)
if err != nil {
t.Fatal(err)
}
var buf bytes.Buffer
if err := streamFn(&buf); err != nil {
t.Fatal(err)
}
if buf.Len() != int(totalSize) {
t.Fatalf("expected %d bytes, got %d", totalSize, buf.Len())
}
// Verify data order
result := buf.Bytes()
for i := 0; i < 4; i++ {
fileId := fmt.Sprintf("1,%x", i)
expected := chunkData[fileId]
got := result[i*1024 : (i+1)*1024]
if !bytes.Equal(expected, got) {
t.Fatalf("chunk %d data mismatch", i)
}
}
}
// TestPrefetchContextCancellation verifies all goroutines clean up on cancellation
func TestPrefetchContextCancellation(t *testing.T) {
// Use a slow server so cancellation happens mid-stream
var requestCount int32
server := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
atomic.AddInt32(&requestCount, 1)
// Slow response
time.Sleep(100 * time.Millisecond)
w.Header().Set("Content-Length", "1024")
w.WriteHeader(http.StatusOK)
w.Write(make([]byte, 1024))
}))
defer server.Close()
numChunks := 16
chunks := make([]*filer_pb.FileChunk, numChunks)
urls := make(map[string][]string, numChunks)
for i := 0; i < numChunks; i++ {
fileId := fmt.Sprintf("1,%x", i)
chunks[i] = &filer_pb.FileChunk{
FileId: fileId, Offset: int64(i * 1024), Size: 1024,
ModifiedTsNs: int64(i), Fid: &filer_pb.FileId{FileKey: uint64(i)},
}
urls[fileId] = []string{server.URL + "/" + fileId}
}
masterClient := &testMasterClient{urls: urls}
// Cancel after a short time
ctx, cancel := context.WithTimeout(context.Background(), 50*time.Millisecond)
defer cancel()
streamFn, err := PrepareStreamContentWithPrefetch(
ctx, masterClient, noopJwt,
chunks, 0, int64(numChunks*1024), 0, 4,
)
if err != nil {
// URL resolution may fail due to cancellation — that's expected
return
}
err = streamFn(io.Discard)
if err == nil {
t.Fatal("expected error from cancelled context")
}
// Verify not all chunks were requested (cancellation stopped early)
reqs := atomic.LoadInt32(&requestCount)
if reqs >= int32(numChunks) {
t.Logf("warning: all %d chunks were requested despite cancellation (got %d)", numChunks, reqs)
}
}
// TestPrefetchRangeRequest verifies prefetch works with offset/size subset
func TestPrefetchRangeRequest(t *testing.T) {
chunks, masterClient, chunkData, cleanup := makeChunksAndServer(t, 8, 4096)
defer cleanup()
// Request only chunks 2-5 (offset=8192, size=16384)
offset := int64(2 * 4096)
size := int64(4 * 4096)
streamFn, err := PrepareStreamContentWithPrefetch(
context.Background(), masterClient, noopJwt,
chunks, offset, size, 0, 4,
)
if err != nil {
t.Fatal(err)
}
var buf bytes.Buffer
if err := streamFn(&buf); err != nil {
t.Fatal(err)
}
if buf.Len() != int(size) {
t.Fatalf("expected %d bytes, got %d", size, buf.Len())
}
// Verify data matches chunks 2-5
result := buf.Bytes()
for i := 2; i < 6; i++ {
fileId := fmt.Sprintf("1,%x", i)
expected := chunkData[fileId]
start := (i - 2) * 4096
got := result[start : start+4096]
if !bytes.Equal(expected, got) {
t.Fatalf("chunk %d data mismatch in range request", i)
}
}
}
// TestPrefetchLargePrefetchCount verifies prefetch > numChunks is handled gracefully
func TestPrefetchLargePrefetchCount(t *testing.T) {
chunks, masterClient, _, cleanup := makeChunksAndServer(t, 3, 1024)
defer cleanup()
totalSize := int64(3 * 1024)
// prefetch=10 but only 3 chunks — should work fine
streamFn, err := PrepareStreamContentWithPrefetch(
context.Background(), masterClient, noopJwt,
chunks, 0, totalSize, 0, 10,
)
if err != nil {
t.Fatal(err)
}
var buf bytes.Buffer
if err := streamFn(&buf); err != nil {
t.Fatal(err)
}
if buf.Len() != int(totalSize) {
t.Fatalf("expected %d bytes, got %d", totalSize, buf.Len())
}
}
// TestPrefetchConcurrentDownloads verifies multiple concurrent prefetch streams
func TestPrefetchConcurrentDownloads(t *testing.T) {
chunks, masterClient, _, cleanup := makeChunksAndServer(t, 8, 2048)
defer cleanup()
totalSize := int64(8 * 2048)
var wg sync.WaitGroup
errors := make(chan error, 4)
for i := 0; i < 4; i++ {
wg.Add(1)
go func() {
defer wg.Done()
streamFn, err := PrepareStreamContentWithPrefetch(
context.Background(), masterClient, noopJwt,
chunks, 0, totalSize, 0, 4,
)
if err != nil {
errors <- err
return
}
if err := streamFn(io.Discard); err != nil {
errors <- err
}
}()
}
wg.Wait()
close(errors)
for err := range errors {
t.Fatalf("concurrent download error: %v", err)
}
}
-13
View File
@@ -37,11 +37,6 @@ func GenerateRandomString(length int, charset string) (string, error) {
return string(b), nil
}
// GenerateAccessKeyId generates a new access key ID.
func GenerateAccessKeyId() (string, error) {
return GenerateRandomString(AccessKeyIdLength, CharsetUpper)
}
// GenerateSecretAccessKey generates a new secret access key.
func GenerateSecretAccessKey() (string, error) {
return GenerateRandomString(SecretAccessKeyLength, Charset)
@@ -179,11 +174,3 @@ func MapToIdentitiesAction(action string) string {
return ""
}
}
// MaskAccessKey masks an access key for logging, showing only the first 4 characters.
func MaskAccessKey(accessKeyId string) string {
if len(accessKeyId) > 4 {
return accessKeyId[:4] + "***"
}
return accessKeyId
}
-164
View File
@@ -1,164 +0,0 @@
package iam
import (
"testing"
"github.com/seaweedfs/seaweedfs/weed/s3api/s3_constants"
"github.com/stretchr/testify/assert"
)
func TestHash(t *testing.T) {
input := "test"
result := Hash(&input)
assert.NotEmpty(t, result)
assert.Len(t, result, 40) // SHA1 hex is 40 chars
// Same input should produce same hash
result2 := Hash(&input)
assert.Equal(t, result, result2)
// Different input should produce different hash
different := "different"
result3 := Hash(&different)
assert.NotEqual(t, result, result3)
}
func TestGenerateRandomString(t *testing.T) {
// Valid generation
result, err := GenerateRandomString(10, CharsetUpper)
assert.NoError(t, err)
assert.Len(t, result, 10)
// Different calls should produce different results (with high probability)
result2, err := GenerateRandomString(10, CharsetUpper)
assert.NoError(t, err)
assert.NotEqual(t, result, result2)
// Invalid length
_, err = GenerateRandomString(0, CharsetUpper)
assert.Error(t, err)
_, err = GenerateRandomString(-1, CharsetUpper)
assert.Error(t, err)
// Empty charset
_, err = GenerateRandomString(10, "")
assert.Error(t, err)
}
func TestGenerateAccessKeyId(t *testing.T) {
keyId, err := GenerateAccessKeyId()
assert.NoError(t, err)
assert.Len(t, keyId, AccessKeyIdLength)
}
func TestGenerateSecretAccessKey(t *testing.T) {
secretKey, err := GenerateSecretAccessKey()
assert.NoError(t, err)
assert.Len(t, secretKey, SecretAccessKeyLength)
}
func TestGenerateSecretAccessKey_URLSafe(t *testing.T) {
// Generate multiple keys to increase probability of catching unsafe chars
for i := 0; i < 100; i++ {
secretKey, err := GenerateSecretAccessKey()
assert.NoError(t, err)
// Verify no URL-unsafe characters that would cause authentication issues
assert.NotContains(t, secretKey, "/", "Secret key should not contain /")
assert.NotContains(t, secretKey, "+", "Secret key should not contain +")
// Verify only expected characters are present
for _, char := range secretKey {
assert.Contains(t, Charset, string(char), "Secret key contains unexpected character: %c", char)
}
}
}
func TestStringSlicesEqual(t *testing.T) {
tests := []struct {
a []string
b []string
expected bool
}{
{[]string{"a", "b", "c"}, []string{"a", "b", "c"}, true},
{[]string{"c", "b", "a"}, []string{"a", "b", "c"}, true}, // Order independent
{[]string{"a", "b"}, []string{"a", "b", "c"}, false},
{[]string{}, []string{}, true},
{nil, nil, true},
{[]string{"a"}, []string{"b"}, false},
}
for _, test := range tests {
result := StringSlicesEqual(test.a, test.b)
assert.Equal(t, test.expected, result)
}
}
func TestMapToStatementAction(t *testing.T) {
tests := []struct {
input string
expected string
}{
{StatementActionAdmin, s3_constants.ACTION_ADMIN},
{StatementActionWrite, s3_constants.ACTION_WRITE},
{StatementActionRead, s3_constants.ACTION_READ},
{StatementActionList, s3_constants.ACTION_LIST},
{StatementActionDelete, s3_constants.ACTION_DELETE_BUCKET},
// Test fine-grained S3 action mappings (Issue #7864)
{"DeleteObject", s3_constants.ACTION_WRITE},
{"s3:DeleteObject", s3_constants.ACTION_WRITE},
{"PutObject", s3_constants.ACTION_WRITE},
{"s3:PutObject", s3_constants.ACTION_WRITE},
{"GetObject", s3_constants.ACTION_READ},
{"s3:GetObject", s3_constants.ACTION_READ},
{"ListBucket", s3_constants.ACTION_LIST},
{"s3:ListBucket", s3_constants.ACTION_LIST},
{"PutObjectAcl", s3_constants.ACTION_WRITE_ACP},
{"s3:PutObjectAcl", s3_constants.ACTION_WRITE_ACP},
{"GetObjectAcl", s3_constants.ACTION_READ_ACP},
{"s3:GetObjectAcl", s3_constants.ACTION_READ_ACP},
{"unknown", ""},
}
for _, test := range tests {
result := MapToStatementAction(test.input)
assert.Equal(t, test.expected, result, "Failed for input: %s", test.input)
}
}
func TestMapToIdentitiesAction(t *testing.T) {
tests := []struct {
input string
expected string
}{
{s3_constants.ACTION_ADMIN, StatementActionAdmin},
{s3_constants.ACTION_WRITE, StatementActionWrite},
{s3_constants.ACTION_READ, StatementActionRead},
{s3_constants.ACTION_LIST, StatementActionList},
{s3_constants.ACTION_DELETE_BUCKET, StatementActionDelete},
{"unknown", ""},
}
for _, test := range tests {
result := MapToIdentitiesAction(test.input)
assert.Equal(t, test.expected, result)
}
}
func TestMaskAccessKey(t *testing.T) {
tests := []struct {
input string
expected string
}{
{"AKIAIOSFODNN7EXAMPLE", "AKIA***"},
{"AKIA", "AKIA"},
{"AKI", "AKI"},
{"", ""},
}
for _, test := range tests {
result := MaskAccessKey(test.input)
assert.Equal(t, test.expected, result)
}
}
+17 -26
View File
@@ -202,32 +202,6 @@ func (m *IAMManager) getFilerAddress() string {
return "" // Fallback to empty string if no provider is set
}
// createRoleStore creates a role store based on configuration
func (m *IAMManager) createRoleStore(config *RoleStoreConfig) (RoleStore, error) {
if config == nil {
// Default to generic cached filer role store when no config provided
return NewGenericCachedRoleStore(nil, nil)
}
switch config.StoreType {
case "", "filer":
// Check if caching is explicitly disabled
if config.StoreConfig != nil {
if noCache, ok := config.StoreConfig["noCache"].(bool); ok && noCache {
return NewFilerRoleStore(config.StoreConfig, nil)
}
}
// Default to generic cached filer store for better performance
return NewGenericCachedRoleStore(config.StoreConfig, nil)
case "cached-filer", "generic-cached":
return NewGenericCachedRoleStore(config.StoreConfig, nil)
case "memory":
return NewMemoryRoleStore(), nil
default:
return nil, fmt.Errorf("unsupported role store type: %s", config.StoreType)
}
}
// createRoleStoreWithProvider creates a role store with a filer address provider function
func (m *IAMManager) createRoleStoreWithProvider(config *RoleStoreConfig, filerAddressProvider func() string) (RoleStore, error) {
if config == nil {
@@ -725,6 +699,23 @@ func (m *IAMManager) ExpireSessionForTesting(ctx context.Context, sessionToken s
return m.stsService.ExpireSessionForTesting(ctx, sessionToken)
}
// GetPoliciesForUser returns the policy names attached to an IAM user.
// Returns an error if the user store is not configured or the lookup fails,
// so callers can fail closed on policy-resolution failures.
func (m *IAMManager) GetPoliciesForUser(ctx context.Context, username string) ([]string, error) {
if m.userStore == nil {
return nil, fmt.Errorf("user store not configured")
}
user, err := m.userStore.GetUser(ctx, username)
if err != nil {
return nil, fmt.Errorf("failed to look up user %q: %w", username, err)
}
if user == nil {
return nil, nil
}
return user.PolicyNames, nil
}
// GetSTSService returns the STS service instance
func (m *IAMManager) GetSTSService() *sts.STSService {
return m.stsService
-154
View File
@@ -388,157 +388,3 @@ type CachedFilerRoleStoreConfig struct {
ListTTL string `json:"listTtl,omitempty"` // e.g., "1m", "30s"
MaxCacheSize int `json:"maxCacheSize,omitempty"` // Maximum number of cached roles
}
// NewCachedFilerRoleStore creates a new cached filer-based role store
func NewCachedFilerRoleStore(config map[string]interface{}) (*CachedFilerRoleStore, error) {
// Create underlying filer store
filerStore, err := NewFilerRoleStore(config, nil)
if err != nil {
return nil, fmt.Errorf("failed to create filer role store: %w", err)
}
// Parse cache configuration with defaults
cacheTTL := 5 * time.Minute // Default 5 minutes for role cache
listTTL := 1 * time.Minute // Default 1 minute for list cache
maxCacheSize := 1000 // Default max 1000 cached roles
if config != nil {
if ttlStr, ok := config["ttl"].(string); ok && ttlStr != "" {
if parsed, err := time.ParseDuration(ttlStr); err == nil {
cacheTTL = parsed
}
}
if listTTLStr, ok := config["listTtl"].(string); ok && listTTLStr != "" {
if parsed, err := time.ParseDuration(listTTLStr); err == nil {
listTTL = parsed
}
}
if maxSize, ok := config["maxCacheSize"].(int); ok && maxSize > 0 {
maxCacheSize = maxSize
}
}
// Create ccache instances with appropriate configurations
pruneCount := int64(maxCacheSize) >> 3
if pruneCount <= 0 {
pruneCount = 100
}
store := &CachedFilerRoleStore{
filerStore: filerStore,
cache: ccache.New(ccache.Configure().MaxSize(int64(maxCacheSize)).ItemsToPrune(uint32(pruneCount))),
listCache: ccache.New(ccache.Configure().MaxSize(100).ItemsToPrune(10)), // Smaller cache for lists
ttl: cacheTTL,
listTTL: listTTL,
}
glog.V(2).Infof("Initialized CachedFilerRoleStore with TTL %v, List TTL %v, Max Cache Size %d",
cacheTTL, listTTL, maxCacheSize)
return store, nil
}
// StoreRole stores a role definition and invalidates the cache
func (c *CachedFilerRoleStore) StoreRole(ctx context.Context, filerAddress string, roleName string, role *RoleDefinition) error {
// Store in filer
err := c.filerStore.StoreRole(ctx, filerAddress, roleName, role)
if err != nil {
return err
}
// Invalidate cache entries
c.cache.Delete(roleName)
c.listCache.Clear() // Invalidate list cache
glog.V(3).Infof("Stored and invalidated cache for role %s", roleName)
return nil
}
// GetRole retrieves a role definition with caching
func (c *CachedFilerRoleStore) GetRole(ctx context.Context, filerAddress string, roleName string) (*RoleDefinition, error) {
// Try to get from cache first
item := c.cache.Get(roleName)
if item != nil {
// Cache hit - return cached role (DO NOT extend TTL)
role := item.Value().(*RoleDefinition)
glog.V(4).Infof("Cache hit for role %s", roleName)
return copyRoleDefinition(role), nil
}
// Cache miss - fetch from filer
glog.V(4).Infof("Cache miss for role %s, fetching from filer", roleName)
role, err := c.filerStore.GetRole(ctx, filerAddress, roleName)
if err != nil {
return nil, err
}
// Cache the result with TTL
c.cache.Set(roleName, copyRoleDefinition(role), c.ttl)
glog.V(3).Infof("Cached role %s with TTL %v", roleName, c.ttl)
return role, nil
}
// ListRoles lists all role names with caching
func (c *CachedFilerRoleStore) ListRoles(ctx context.Context, filerAddress string) ([]string, error) {
// Use a constant key for the role list cache
const listCacheKey = "role_list"
// Try to get from list cache first
item := c.listCache.Get(listCacheKey)
if item != nil {
// Cache hit - return cached list (DO NOT extend TTL)
roles := item.Value().([]string)
glog.V(4).Infof("List cache hit, returning %d roles", len(roles))
return append([]string(nil), roles...), nil // Return a copy
}
// Cache miss - fetch from filer
glog.V(4).Infof("List cache miss, fetching from filer")
roles, err := c.filerStore.ListRoles(ctx, filerAddress)
if err != nil {
return nil, err
}
// Cache the result with TTL (store a copy)
rolesCopy := append([]string(nil), roles...)
c.listCache.Set(listCacheKey, rolesCopy, c.listTTL)
glog.V(3).Infof("Cached role list with %d entries, TTL %v", len(roles), c.listTTL)
return roles, nil
}
// DeleteRole deletes a role definition and invalidates the cache
func (c *CachedFilerRoleStore) DeleteRole(ctx context.Context, filerAddress string, roleName string) error {
// Delete from filer
err := c.filerStore.DeleteRole(ctx, filerAddress, roleName)
if err != nil {
return err
}
// Invalidate cache entries
c.cache.Delete(roleName)
c.listCache.Clear() // Invalidate list cache
glog.V(3).Infof("Deleted and invalidated cache for role %s", roleName)
return nil
}
// ClearCache clears all cached entries (for testing or manual cache invalidation)
func (c *CachedFilerRoleStore) ClearCache() {
c.cache.Clear()
c.listCache.Clear()
glog.V(2).Infof("Cleared all role cache entries")
}
// GetCacheStats returns cache statistics
func (c *CachedFilerRoleStore) GetCacheStats() map[string]interface{} {
return map[string]interface{}{
"roleCache": map[string]interface{}{
"size": c.cache.ItemCount(),
"ttl": c.ttl.String(),
},
"listCache": map[string]interface{}{
"size": c.listCache.ItemCount(),
"ttl": c.listTTL.String(),
},
}
}
-687
View File
@@ -1,687 +0,0 @@
package policy
import (
"context"
"testing"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
func TestConditionSetOperators(t *testing.T) {
engine := setupTestPolicyEngine(t)
t.Run("ForAnyValue:StringEquals", func(t *testing.T) {
trustPolicy := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "AllowOIDC",
Effect: "Allow",
Action: []string{"sts:AssumeRoleWithWebIdentity"},
Condition: map[string]map[string]interface{}{
"ForAnyValue:StringEquals": {
"oidc:roles": []string{"Dev.SeaweedFS.TestBucket.ReadWrite", "Dev.SeaweedFS.Admin"},
},
},
},
},
}
// Match: Admin is in the requested roles
evalCtxMatch := &EvaluationContext{
Principal: "web-identity-user",
Action: "sts:AssumeRoleWithWebIdentity",
Resource: "arn:aws:iam::role/test-role",
RequestContext: map[string]interface{}{
"oidc:roles": []string{"Dev.SeaweedFS.Admin", "OtherRole"},
},
}
resultMatch, err := engine.EvaluateTrustPolicy(context.Background(), trustPolicy, evalCtxMatch)
require.NoError(t, err)
assert.Equal(t, EffectAllow, resultMatch.Effect)
// No Match
evalCtxNoMatch := &EvaluationContext{
Principal: "web-identity-user",
Action: "sts:AssumeRoleWithWebIdentity",
Resource: "arn:aws:iam::role/test-role",
RequestContext: map[string]interface{}{
"oidc:roles": []string{"OtherRole1", "OtherRole2"},
},
}
resultNoMatch, err := engine.EvaluateTrustPolicy(context.Background(), trustPolicy, evalCtxNoMatch)
require.NoError(t, err)
assert.Equal(t, EffectDeny, resultNoMatch.Effect)
// No Match: Empty context for ForAnyValue (should deny)
evalCtxEmpty := &EvaluationContext{
Principal: "web-identity-user",
Action: "sts:AssumeRoleWithWebIdentity",
Resource: "arn:aws:iam::role/test-role",
RequestContext: map[string]interface{}{
"oidc:roles": []string{},
},
}
resultEmpty, err := engine.EvaluateTrustPolicy(context.Background(), trustPolicy, evalCtxEmpty)
require.NoError(t, err)
assert.Equal(t, EffectDeny, resultEmpty.Effect, "ForAnyValue should deny when context is empty")
})
t.Run("ForAllValues:StringEquals", func(t *testing.T) {
trustPolicyAll := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "AllowOIDCAll",
Effect: "Allow",
Action: []string{"sts:AssumeRoleWithWebIdentity"},
Condition: map[string]map[string]interface{}{
"ForAllValues:StringEquals": {
"oidc:roles": []string{"RoleA", "RoleB", "RoleC"},
},
},
},
},
}
// Match: All requested roles ARE in the allowed set
evalCtxAllMatch := &EvaluationContext{
Principal: "web-identity-user",
Action: "sts:AssumeRoleWithWebIdentity",
Resource: "arn:aws:iam::role/test-role",
RequestContext: map[string]interface{}{
"oidc:roles": []string{"RoleA", "RoleB"},
},
}
resultAllMatch, err := engine.EvaluateTrustPolicy(context.Background(), trustPolicyAll, evalCtxAllMatch)
require.NoError(t, err)
assert.Equal(t, EffectAllow, resultAllMatch.Effect)
// Fail: RoleD is NOT in the allowed set
evalCtxAllFail := &EvaluationContext{
Principal: "web-identity-user",
Action: "sts:AssumeRoleWithWebIdentity",
Resource: "arn:aws:iam::role/test-role",
RequestContext: map[string]interface{}{
"oidc:roles": []string{"RoleA", "RoleD"},
},
}
resultAllFail, err := engine.EvaluateTrustPolicy(context.Background(), trustPolicyAll, evalCtxAllFail)
require.NoError(t, err)
assert.Equal(t, EffectDeny, resultAllFail.Effect)
// Vacuously true: Request has NO roles
evalCtxEmpty := &EvaluationContext{
Principal: "web-identity-user",
Action: "sts:AssumeRoleWithWebIdentity",
Resource: "arn:aws:iam::role/test-role",
RequestContext: map[string]interface{}{
"oidc:roles": []string{},
},
}
resultEmpty, err := engine.EvaluateTrustPolicy(context.Background(), trustPolicyAll, evalCtxEmpty)
require.NoError(t, err)
assert.Equal(t, EffectAllow, resultEmpty.Effect)
})
t.Run("ForAllValues:NumericEqualsVacuouslyTrue", func(t *testing.T) {
policy := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "AllowNumericAll",
Effect: "Allow",
Action: []string{"sts:AssumeRole"},
Condition: map[string]map[string]interface{}{
"ForAllValues:NumericEquals": {
"aws:MultiFactorAuthAge": []string{"3600", "7200"},
},
},
},
},
}
// Vacuously true: Request has NO MFA age info
evalCtxEmpty := &EvaluationContext{
Principal: "user",
Action: "sts:AssumeRole",
Resource: "arn:aws:iam::role/test-role",
RequestContext: map[string]interface{}{
"aws:MultiFactorAuthAge": []string{},
},
}
resultEmpty, err := engine.EvaluateTrustPolicy(context.Background(), policy, evalCtxEmpty)
require.NoError(t, err)
assert.Equal(t, EffectAllow, resultEmpty.Effect, "Should allow when numeric context is empty for ForAllValues")
})
t.Run("ForAllValues:BoolVacuouslyTrue", func(t *testing.T) {
policy := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "AllowBoolAll",
Effect: "Allow",
Action: []string{"sts:AssumeRole"},
Condition: map[string]map[string]interface{}{
"ForAllValues:Bool": {
"aws:SecureTransport": "true",
},
},
},
},
}
// Vacuously true
evalCtxEmpty := &EvaluationContext{
Principal: "user",
Action: "sts:AssumeRole",
Resource: "arn:aws:iam::role/test-role",
RequestContext: map[string]interface{}{
"aws:SecureTransport": []interface{}{},
},
}
resultEmpty, err := engine.EvaluateTrustPolicy(context.Background(), policy, evalCtxEmpty)
require.NoError(t, err)
assert.Equal(t, EffectAllow, resultEmpty.Effect, "Should allow when bool context is empty for ForAllValues")
})
t.Run("ForAllValues:DateVacuouslyTrue", func(t *testing.T) {
policy := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "AllowDateAll",
Effect: "Allow",
Action: []string{"sts:AssumeRole"},
Condition: map[string]map[string]interface{}{
"ForAllValues:DateGreaterThan": {
"aws:CurrentTime": "2020-01-01T00:00:00Z",
},
},
},
},
}
// Vacuously true
evalCtxEmpty := &EvaluationContext{
Principal: "user",
Action: "sts:AssumeRole",
Resource: "arn:aws:iam::role/test-role",
RequestContext: map[string]interface{}{
"aws:CurrentTime": []interface{}{},
},
}
resultEmpty, err := engine.EvaluateTrustPolicy(context.Background(), policy, evalCtxEmpty)
require.NoError(t, err)
assert.Equal(t, EffectAllow, resultEmpty.Effect, "Should allow when date context is empty for ForAllValues")
})
t.Run("ForAllValues:DateWithLabelsAsStrings", func(t *testing.T) {
policy := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "AllowDateStrings",
Effect: "Allow",
Action: []string{"sts:AssumeRole"},
Condition: map[string]map[string]interface{}{
"ForAllValues:DateGreaterThan": {
"aws:CurrentTime": "2020-01-01T00:00:00Z",
},
},
},
},
}
evalCtx := &EvaluationContext{
Principal: "user",
Action: "sts:AssumeRole",
Resource: "arn:aws:iam::role/test-role",
RequestContext: map[string]interface{}{
"aws:CurrentTime": []string{"2021-01-01T00:00:00Z", "2022-01-01T00:00:00Z"},
},
}
result, err := engine.EvaluateTrustPolicy(context.Background(), policy, evalCtx)
require.NoError(t, err)
assert.Equal(t, EffectAllow, result.Effect, "Should allow when date context is a slice of strings")
})
t.Run("ForAllValues:BoolWithLabelsAsStrings", func(t *testing.T) {
policy := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "AllowBoolStrings",
Effect: "Allow",
Action: []string{"sts:AssumeRole"},
Condition: map[string]map[string]interface{}{
"ForAllValues:Bool": {
"aws:SecureTransport": "true",
},
},
},
},
}
evalCtx := &EvaluationContext{
Principal: "user",
Action: "sts:AssumeRole",
Resource: "arn:aws:iam::role/test-role",
RequestContext: map[string]interface{}{
"aws:SecureTransport": []string{"true", "true"},
},
}
result, err := engine.EvaluateTrustPolicy(context.Background(), policy, evalCtx)
require.NoError(t, err)
assert.Equal(t, EffectAllow, result.Effect, "Should allow when bool context is a slice of strings")
})
t.Run("StringEqualsIgnoreCaseWithVariable", func(t *testing.T) {
policyDoc := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "AllowVar",
Effect: "Allow",
Action: []string{"s3:GetObject"},
Resource: []string{"arn:aws:s3:::bucket/*"},
Condition: map[string]map[string]interface{}{
"StringEqualsIgnoreCase": {
"s3:prefix": "${aws:username}/",
},
},
},
},
}
err := engine.AddPolicy("", "var-policy", policyDoc)
require.NoError(t, err)
evalCtx := &EvaluationContext{
Principal: "user",
Action: "s3:GetObject",
Resource: "arn:aws:s3:::bucket/ALICE/file.txt",
RequestContext: map[string]interface{}{
"s3:prefix": "ALICE/",
"aws:username": "alice",
},
}
result, err := engine.Evaluate(context.Background(), "", evalCtx, []string{"var-policy"})
require.NoError(t, err)
assert.Equal(t, EffectAllow, result.Effect, "Should allow when variable expands and matches case-insensitively")
})
t.Run("StringLike:CaseSensitivity", func(t *testing.T) {
policyDoc := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "AllowCaseSensitiveLike",
Effect: "Allow",
Action: []string{"s3:GetObject"},
Resource: []string{"arn:aws:s3:::bucket/*"},
Condition: map[string]map[string]interface{}{
"StringLike": {
"s3:prefix": "Project/*",
},
},
},
},
}
err := engine.AddPolicy("", "like-policy", policyDoc)
require.NoError(t, err)
// Match: Case sensitive match
evalCtxMatch := &EvaluationContext{
Principal: "user",
Action: "s3:GetObject",
Resource: "arn:aws:s3:::bucket/Project/file.txt",
RequestContext: map[string]interface{}{
"s3:prefix": "Project/data",
},
}
resultMatch, err := engine.Evaluate(context.Background(), "", evalCtxMatch, []string{"like-policy"})
require.NoError(t, err)
assert.Equal(t, EffectAllow, resultMatch.Effect, "Should allow when case matches exactly")
// Fail: Case insensitive match (should fail for StringLike)
evalCtxFail := &EvaluationContext{
Principal: "user",
Action: "s3:GetObject",
Resource: "arn:aws:s3:::bucket/project/file.txt",
RequestContext: map[string]interface{}{
"s3:prefix": "project/data", // lowercase 'p'
},
}
resultFail, err := engine.Evaluate(context.Background(), "", evalCtxFail, []string{"like-policy"})
require.NoError(t, err)
assert.Equal(t, EffectDeny, resultFail.Effect, "Should deny when case does not match for StringLike")
})
t.Run("NumericNotEquals:Logic", func(t *testing.T) {
policy := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "DenySpecificAges",
Effect: "Allow",
Action: []string{"sts:AssumeRole"},
Resource: []string{"*"},
Condition: map[string]map[string]interface{}{
"ForAllValues:NumericNotEquals": {
"aws:MultiFactorAuthAge": []string{"3600", "7200"},
},
},
},
},
}
err := engine.AddPolicy("", "numeric-not-equals-policy", policy)
require.NoError(t, err)
// Fail: One age matches an excluded value (3600)
evalCtxFail := &EvaluationContext{
Principal: "user",
Action: "sts:AssumeRole",
Resource: "arn:aws:iam::role/test-role",
RequestContext: map[string]interface{}{
"aws:MultiFactorAuthAge": []string{"3600", "1800"},
},
}
resultFail, err := engine.Evaluate(context.Background(), "", evalCtxFail, []string{"numeric-not-equals-policy"})
require.NoError(t, err)
assert.Equal(t, EffectDeny, resultFail.Effect, "Should deny when one age matches an excluded value")
// Pass: No age matches any excluded value
evalCtxPass := &EvaluationContext{
Principal: "user",
Action: "sts:AssumeRole",
Resource: "arn:aws:iam::role/test-role",
RequestContext: map[string]interface{}{
"aws:MultiFactorAuthAge": []string{"1800", "900"},
},
}
resultPass, err := engine.Evaluate(context.Background(), "", evalCtxPass, []string{"numeric-not-equals-policy"})
require.NoError(t, err)
assert.Equal(t, EffectAllow, resultPass.Effect, "Should allow when no age matches excluded values")
})
t.Run("DateNotEquals:Logic", func(t *testing.T) {
policy := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "DenySpecificTimes",
Effect: "Allow",
Action: []string{"sts:AssumeRole"},
Resource: []string{"*"},
Condition: map[string]map[string]interface{}{
"ForAllValues:DateNotEquals": {
"aws:CurrentTime": []string{"2024-01-01T00:00:00Z", "2024-01-02T00:00:00Z"},
},
},
},
},
}
err := engine.AddPolicy("", "date-not-equals-policy", policy)
require.NoError(t, err)
// Fail: One time matches an excluded value
evalCtxFail := &EvaluationContext{
Principal: "user",
Action: "sts:AssumeRole",
Resource: "arn:aws:iam::role/test-role",
RequestContext: map[string]interface{}{
"aws:CurrentTime": []string{"2024-01-01T00:00:00Z", "2024-01-03T00:00:00Z"},
},
}
resultFail, err := engine.Evaluate(context.Background(), "", evalCtxFail, []string{"date-not-equals-policy"})
require.NoError(t, err)
assert.Equal(t, EffectDeny, resultFail.Effect, "Should deny when one date matches an excluded value")
})
t.Run("IpAddress:SetOperators", func(t *testing.T) {
policy := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "AllowSpecificIPs",
Effect: "Allow",
Action: []string{"s3:GetObject"},
Resource: []string{"*"},
Condition: map[string]map[string]interface{}{
"ForAllValues:IpAddress": {
"aws:SourceIp": []string{"192.168.1.0/24", "10.0.0.1"},
},
},
},
},
}
err := engine.AddPolicy("", "ip-set-policy", policy)
require.NoError(t, err)
// Match: All source IPs are in allowed ranges
evalCtxMatch := &EvaluationContext{
Principal: "user",
Action: "s3:GetObject",
Resource: "arn:aws:s3:::bucket/file.txt",
RequestContext: map[string]interface{}{
"aws:SourceIp": []string{"192.168.1.10", "10.0.0.1"},
},
}
resultMatch, err := engine.Evaluate(context.Background(), "", evalCtxMatch, []string{"ip-set-policy"})
require.NoError(t, err)
assert.Equal(t, EffectAllow, resultMatch.Effect)
// Fail: One source IP is NOT in allowed ranges
evalCtxFail := &EvaluationContext{
Principal: "user",
Action: "s3:GetObject",
Resource: "arn:aws:s3:::bucket/file.txt",
RequestContext: map[string]interface{}{
"aws:SourceIp": []string{"192.168.1.10", "172.16.0.1"},
},
}
resultFail, err := engine.Evaluate(context.Background(), "", evalCtxFail, []string{"ip-set-policy"})
require.NoError(t, err)
assert.Equal(t, EffectDeny, resultFail.Effect)
// ForAnyValue: IPAddress
policyAny := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "AllowAnySpecificIPs",
Effect: "Allow",
Action: []string{"s3:GetObject"},
Resource: []string{"*"},
Condition: map[string]map[string]interface{}{
"ForAnyValue:IpAddress": {
"aws:SourceIp": []string{"192.168.1.0/24"},
},
},
},
},
}
err = engine.AddPolicy("", "ip-any-policy", policyAny)
require.NoError(t, err)
evalCtxAnyMatch := &EvaluationContext{
Principal: "user",
Action: "s3:GetObject",
Resource: "arn:aws:s3:::bucket/file.txt",
RequestContext: map[string]interface{}{
"aws:SourceIp": []string{"192.168.1.10", "172.16.0.1"},
},
}
resultAnyMatch, err := engine.Evaluate(context.Background(), "", evalCtxAnyMatch, []string{"ip-any-policy"})
require.NoError(t, err)
assert.Equal(t, EffectAllow, resultAnyMatch.Effect)
})
t.Run("IpAddress:SingleStringValue", func(t *testing.T) {
policy := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "AllowSingleIP",
Effect: "Allow",
Action: []string{"s3:GetObject"},
Resource: []string{"*"},
Condition: map[string]map[string]interface{}{
"IpAddress": {
"aws:SourceIp": "192.168.1.1",
},
},
},
},
}
err := engine.AddPolicy("", "ip-single-policy", policy)
require.NoError(t, err)
evalCtxMatch := &EvaluationContext{
Principal: "user",
Action: "s3:GetObject",
Resource: "arn:aws:s3:::bucket/file.txt",
RequestContext: map[string]interface{}{
"aws:SourceIp": "192.168.1.1",
},
}
resultMatch, err := engine.Evaluate(context.Background(), "", evalCtxMatch, []string{"ip-single-policy"})
require.NoError(t, err)
assert.Equal(t, EffectAllow, resultMatch.Effect)
evalCtxNoMatch := &EvaluationContext{
Principal: "user",
Action: "s3:GetObject",
Resource: "arn:aws:s3:::bucket/file.txt",
RequestContext: map[string]interface{}{
"aws:SourceIp": "10.0.0.1",
},
}
resultNoMatch, err := engine.Evaluate(context.Background(), "", evalCtxNoMatch, []string{"ip-single-policy"})
require.NoError(t, err)
assert.Equal(t, EffectDeny, resultNoMatch.Effect)
})
t.Run("Bool:StringSlicePolicyValues", func(t *testing.T) {
policy := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "AllowWithBoolStrings",
Effect: "Allow",
Action: []string{"s3:GetObject"},
Resource: []string{"*"},
Condition: map[string]map[string]interface{}{
"Bool": {
"aws:SecureTransport": []string{"true", "false"},
},
},
},
},
}
err := engine.AddPolicy("", "bool-string-slice-policy", policy)
require.NoError(t, err)
evalCtx := &EvaluationContext{
Principal: "user",
Action: "s3:GetObject",
Resource: "arn:aws:s3:::bucket/file.txt",
RequestContext: map[string]interface{}{
"aws:SecureTransport": "true",
},
}
result, err := engine.Evaluate(context.Background(), "", evalCtx, []string{"bool-string-slice-policy"})
require.NoError(t, err)
assert.Equal(t, EffectAllow, result.Effect)
})
t.Run("StringEqualsIgnoreCase:StringSlicePolicyValues", func(t *testing.T) {
policy := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "AllowWithIgnoreCaseStrings",
Effect: "Allow",
Action: []string{"s3:GetObject"},
Resource: []string{"*"},
Condition: map[string]map[string]interface{}{
"StringEqualsIgnoreCase": {
"s3:x-amz-server-side-encryption": []string{"AES256", "aws:kms"},
},
},
},
},
}
err := engine.AddPolicy("", "string-ignorecase-slice-policy", policy)
require.NoError(t, err)
evalCtx := &EvaluationContext{
Principal: "user",
Action: "s3:GetObject",
Resource: "arn:aws:s3:::bucket/file.txt",
RequestContext: map[string]interface{}{
"s3:x-amz-server-side-encryption": "aes256",
},
}
result, err := engine.Evaluate(context.Background(), "", evalCtx, []string{"string-ignorecase-slice-policy"})
require.NoError(t, err)
assert.Equal(t, EffectAllow, result.Effect)
})
t.Run("IpAddress:CustomContextKey", func(t *testing.T) {
policy := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "AllowCustomIPKey",
Effect: "Allow",
Action: []string{"s3:GetObject"},
Resource: []string{"*"},
Condition: map[string]map[string]interface{}{
"IpAddress": {
"custom:VpcIp": "10.0.0.0/16",
},
},
},
},
}
err := engine.AddPolicy("", "ip-custom-key-policy", policy)
require.NoError(t, err)
evalCtxMatch := &EvaluationContext{
Principal: "user",
Action: "s3:GetObject",
Resource: "arn:aws:s3:::bucket/file.txt",
RequestContext: map[string]interface{}{
"custom:VpcIp": "10.0.5.1",
},
}
resultMatch, err := engine.Evaluate(context.Background(), "", evalCtxMatch, []string{"ip-custom-key-policy"})
require.NoError(t, err)
assert.Equal(t, EffectAllow, resultMatch.Effect)
evalCtxNoMatch := &EvaluationContext{
Principal: "user",
Action: "s3:GetObject",
Resource: "arn:aws:s3:::bucket/file.txt",
RequestContext: map[string]interface{}{
"custom:VpcIp": "192.168.1.1",
},
}
resultNoMatch, err := engine.Evaluate(context.Background(), "", evalCtxNoMatch, []string{"ip-custom-key-policy"})
require.NoError(t, err)
assert.Equal(t, EffectDeny, resultNoMatch.Effect)
})
}
-101
View File
@@ -1,101 +0,0 @@
package policy
import (
"context"
"testing"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
func TestNegationSetOperators(t *testing.T) {
engine := setupTestPolicyEngine(t)
t.Run("ForAllValues:StringNotEquals", func(t *testing.T) {
policy := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "DenyAdmin",
Effect: "Allow",
Action: []string{"sts:AssumeRole"},
Condition: map[string]map[string]interface{}{
"ForAllValues:StringNotEquals": {
"oidc:roles": []string{"Admin"},
},
},
},
},
}
// All roles are NOT "Admin" -> Should Allow
evalCtxAllow := &EvaluationContext{
Principal: "user",
Action: "sts:AssumeRole",
Resource: "arn:aws:iam::role/test-role",
RequestContext: map[string]interface{}{
"oidc:roles": []string{"User", "Developer"},
},
}
resultAllow, err := engine.EvaluateTrustPolicy(context.Background(), policy, evalCtxAllow)
require.NoError(t, err)
assert.Equal(t, EffectAllow, resultAllow.Effect, "Should allow when ALL roles satisfy StringNotEquals Admin")
// One role is "Admin" -> Should Deny
evalCtxDeny := &EvaluationContext{
Principal: "user",
Action: "sts:AssumeRole",
Resource: "arn:aws:iam::role/test-role",
RequestContext: map[string]interface{}{
"oidc:roles": []string{"Admin", "User"},
},
}
resultDeny, err := engine.EvaluateTrustPolicy(context.Background(), policy, evalCtxDeny)
require.NoError(t, err)
assert.Equal(t, EffectDeny, resultDeny.Effect, "Should deny when one role is Admin and fails StringNotEquals")
})
t.Run("ForAnyValue:StringNotEquals", func(t *testing.T) {
policy := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "Requirement",
Effect: "Allow",
Action: []string{"sts:AssumeRole"},
Condition: map[string]map[string]interface{}{
"ForAnyValue:StringNotEquals": {
"oidc:roles": []string{"Prohibited"},
},
},
},
},
}
// At least one role is NOT prohibited -> Should Allow
evalCtxAllow := &EvaluationContext{
Principal: "user",
Action: "sts:AssumeRole",
Resource: "arn:aws:iam::role/test-role",
RequestContext: map[string]interface{}{
"oidc:roles": []string{"Prohibited", "Allowed"},
},
}
resultAllow, err := engine.EvaluateTrustPolicy(context.Background(), policy, evalCtxAllow)
require.NoError(t, err)
assert.Equal(t, EffectAllow, resultAllow.Effect, "Should allow when at least one role is NOT Prohibited")
// All roles are Prohibited -> Should Deny
evalCtxDeny := &EvaluationContext{
Principal: "user",
Action: "sts:AssumeRole",
Resource: "arn:aws:iam::role/test-role",
RequestContext: map[string]interface{}{
"oidc:roles": []string{"Prohibited", "Prohibited"},
},
}
resultDeny, err := engine.EvaluateTrustPolicy(context.Background(), policy, evalCtxDeny)
require.NoError(t, err)
assert.Equal(t, EffectDeny, resultDeny.Effect, "Should deny when ALL roles are Prohibited")
})
}
-61
View File
@@ -1155,11 +1155,6 @@ func ValidatePolicyDocumentWithType(policy *PolicyDocument, policyType string) e
return nil
}
// validateStatement validates a single statement (for backward compatibility)
func validateStatement(statement *Statement) error {
return validateStatementWithType(statement, "resource")
}
// validateStatementWithType validates a single statement based on policy type
func validateStatementWithType(statement *Statement, policyType string) error {
if statement.Effect != "Allow" && statement.Effect != "Deny" {
@@ -1198,29 +1193,6 @@ func validateStatementWithType(statement *Statement, policyType string) error {
return nil
}
// matchResource checks if a resource pattern matches a requested resource
// Uses hybrid approach: simple suffix wildcards for compatibility, filepath.Match for complex patterns
func matchResource(pattern, resource string) bool {
if pattern == resource {
return true
}
// Handle simple suffix wildcard (backward compatibility)
if strings.HasSuffix(pattern, "*") {
prefix := pattern[:len(pattern)-1]
return strings.HasPrefix(resource, prefix)
}
// For complex patterns, use filepath.Match for advanced wildcard support (*, ?, [])
matched, err := filepath.Match(pattern, resource)
if err != nil {
// Fallback to exact match if pattern is malformed
return pattern == resource
}
return matched
}
// awsIAMMatch performs AWS IAM-compliant pattern matching with case-insensitivity and policy variable support
func awsIAMMatch(pattern, value string, evalCtx *EvaluationContext) bool {
// Step 1: Substitute policy variables (e.g., ${aws:username}, ${saml:username})
@@ -1274,16 +1246,6 @@ func expandPolicyVariables(pattern string, evalCtx *EvaluationContext) string {
return result
}
// getContextValue safely gets a value from the evaluation context
func getContextValue(evalCtx *EvaluationContext, key, defaultValue string) string {
if value, exists := evalCtx.RequestContext[key]; exists {
if str, ok := value.(string); ok {
return str
}
}
return defaultValue
}
// AwsWildcardMatch performs case-insensitive wildcard matching like AWS IAM
func AwsWildcardMatch(pattern, value string) bool {
// Create regex pattern key for caching
@@ -1322,29 +1284,6 @@ func AwsWildcardMatch(pattern, value string) bool {
return regex.MatchString(value)
}
// matchAction checks if an action pattern matches a requested action
// Uses hybrid approach: simple suffix wildcards for compatibility, filepath.Match for complex patterns
func matchAction(pattern, action string) bool {
if pattern == action {
return true
}
// Handle simple suffix wildcard (backward compatibility)
if strings.HasSuffix(pattern, "*") {
prefix := pattern[:len(pattern)-1]
return strings.HasPrefix(action, prefix)
}
// For complex patterns, use filepath.Match for advanced wildcard support (*, ?, [])
matched, err := filepath.Match(pattern, action)
if err != nil {
// Fallback to exact match if pattern is malformed
return pattern == action
}
return matched
}
// evaluateStringConditionIgnoreCase evaluates string conditions with case insensitivity
func (e *PolicyEngine) evaluateStringConditionIgnoreCase(block map[string]interface{}, evalCtx *EvaluationContext, shouldMatch bool, useWildcard bool, forAllValues bool) bool {
for key, expectedValues := range block {
@@ -1,421 +0,0 @@
package policy
import (
"context"
"testing"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
// TestPrincipalMatching tests the matchesPrincipal method
func TestPrincipalMatching(t *testing.T) {
engine := setupTestPolicyEngine(t)
tests := []struct {
name string
principal interface{}
evalCtx *EvaluationContext
want bool
}{
{
name: "plain wildcard principal",
principal: "*",
evalCtx: &EvaluationContext{
RequestContext: map[string]interface{}{},
},
want: true,
},
{
name: "structured wildcard federated principal",
principal: map[string]interface{}{
"Federated": "*",
},
evalCtx: &EvaluationContext{
RequestContext: map[string]interface{}{},
},
want: true,
},
{
name: "wildcard in array",
principal: map[string]interface{}{
"Federated": []interface{}{"specific-provider", "*"},
},
evalCtx: &EvaluationContext{
RequestContext: map[string]interface{}{},
},
want: true,
},
{
name: "specific federated provider match",
principal: map[string]interface{}{
"Federated": "https://example.com/oidc",
},
evalCtx: &EvaluationContext{
RequestContext: map[string]interface{}{
"aws:FederatedProvider": "https://example.com/oidc",
},
},
want: true,
},
{
name: "specific federated provider no match",
principal: map[string]interface{}{
"Federated": "https://example.com/oidc",
},
evalCtx: &EvaluationContext{
RequestContext: map[string]interface{}{
"aws:FederatedProvider": "https://other.com/oidc",
},
},
want: false,
},
{
name: "array with specific provider match",
principal: map[string]interface{}{
"Federated": []string{"https://provider1.com", "https://provider2.com"},
},
evalCtx: &EvaluationContext{
RequestContext: map[string]interface{}{
"aws:FederatedProvider": "https://provider2.com",
},
},
want: true,
},
{
name: "AWS principal match",
principal: map[string]interface{}{
"AWS": "arn:aws:iam::123456789012:user/alice",
},
evalCtx: &EvaluationContext{
RequestContext: map[string]interface{}{
"aws:PrincipalArn": "arn:aws:iam::123456789012:user/alice",
},
},
want: true,
},
{
name: "Service principal match",
principal: map[string]interface{}{
"Service": "s3.amazonaws.com",
},
evalCtx: &EvaluationContext{
RequestContext: map[string]interface{}{
"aws:PrincipalServiceName": "s3.amazonaws.com",
},
},
want: true,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
result := engine.matchesPrincipal(tt.principal, tt.evalCtx)
assert.Equal(t, tt.want, result, "Principal matching failed for: %s", tt.name)
})
}
}
// TestEvaluatePrincipalValue tests the evaluatePrincipalValue method
func TestEvaluatePrincipalValue(t *testing.T) {
engine := setupTestPolicyEngine(t)
tests := []struct {
name string
principalValue interface{}
contextKey string
evalCtx *EvaluationContext
want bool
}{
{
name: "wildcard string",
principalValue: "*",
contextKey: "aws:FederatedProvider",
evalCtx: &EvaluationContext{
RequestContext: map[string]interface{}{},
},
want: true,
},
{
name: "specific string match",
principalValue: "https://example.com",
contextKey: "aws:FederatedProvider",
evalCtx: &EvaluationContext{
RequestContext: map[string]interface{}{
"aws:FederatedProvider": "https://example.com",
},
},
want: true,
},
{
name: "specific string no match",
principalValue: "https://example.com",
contextKey: "aws:FederatedProvider",
evalCtx: &EvaluationContext{
RequestContext: map[string]interface{}{
"aws:FederatedProvider": "https://other.com",
},
},
want: false,
},
{
name: "wildcard in array",
principalValue: []interface{}{"provider1", "*"},
contextKey: "aws:FederatedProvider",
evalCtx: &EvaluationContext{
RequestContext: map[string]interface{}{},
},
want: true,
},
{
name: "array match",
principalValue: []string{"provider1", "provider2", "provider3"},
contextKey: "aws:FederatedProvider",
evalCtx: &EvaluationContext{
RequestContext: map[string]interface{}{
"aws:FederatedProvider": "provider2",
},
},
want: true,
},
{
name: "array no match",
principalValue: []string{"provider1", "provider2"},
contextKey: "aws:FederatedProvider",
evalCtx: &EvaluationContext{
RequestContext: map[string]interface{}{
"aws:FederatedProvider": "provider3",
},
},
want: false,
},
{
name: "missing context key",
principalValue: "specific-value",
contextKey: "aws:FederatedProvider",
evalCtx: &EvaluationContext{
RequestContext: map[string]interface{}{},
},
want: false,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
result := engine.evaluatePrincipalValue(tt.principalValue, tt.evalCtx, tt.contextKey)
assert.Equal(t, tt.want, result, "Principal value evaluation failed for: %s", tt.name)
})
}
}
// TestTrustPolicyEvaluation tests the EvaluateTrustPolicy method
func TestTrustPolicyEvaluation(t *testing.T) {
engine := setupTestPolicyEngine(t)
tests := []struct {
name string
trustPolicy *PolicyDocument
evalCtx *EvaluationContext
wantEffect Effect
wantErr bool
}{
{
name: "wildcard federated principal allows any provider",
trustPolicy: &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Effect: "Allow",
Principal: map[string]interface{}{
"Federated": "*",
},
Action: []string{"sts:AssumeRoleWithWebIdentity"},
},
},
},
evalCtx: &EvaluationContext{
Action: "sts:AssumeRoleWithWebIdentity",
RequestContext: map[string]interface{}{
"aws:FederatedProvider": "https://any-provider.com",
},
},
wantEffect: EffectAllow,
wantErr: false,
},
{
name: "specific federated principal matches",
trustPolicy: &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Effect: "Allow",
Principal: map[string]interface{}{
"Federated": "https://example.com/oidc",
},
Action: []string{"sts:AssumeRoleWithWebIdentity"},
},
},
},
evalCtx: &EvaluationContext{
Action: "sts:AssumeRoleWithWebIdentity",
RequestContext: map[string]interface{}{
"aws:FederatedProvider": "https://example.com/oidc",
},
},
wantEffect: EffectAllow,
wantErr: false,
},
{
name: "specific federated principal does not match",
trustPolicy: &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Effect: "Allow",
Principal: map[string]interface{}{
"Federated": "https://example.com/oidc",
},
Action: []string{"sts:AssumeRoleWithWebIdentity"},
},
},
},
evalCtx: &EvaluationContext{
Action: "sts:AssumeRoleWithWebIdentity",
RequestContext: map[string]interface{}{
"aws:FederatedProvider": "https://other.com/oidc",
},
},
wantEffect: EffectDeny,
wantErr: false,
},
{
name: "plain wildcard principal",
trustPolicy: &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Effect: "Allow",
Principal: "*",
Action: []string{"sts:AssumeRoleWithWebIdentity"},
},
},
},
evalCtx: &EvaluationContext{
Action: "sts:AssumeRoleWithWebIdentity",
RequestContext: map[string]interface{}{
"aws:FederatedProvider": "https://any-provider.com",
},
},
wantEffect: EffectAllow,
wantErr: false,
},
{
name: "trust policy with conditions",
trustPolicy: &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Effect: "Allow",
Principal: map[string]interface{}{
"Federated": "*",
},
Action: []string{"sts:AssumeRoleWithWebIdentity"},
Condition: map[string]map[string]interface{}{
"StringEquals": {
"oidc:aud": "my-app-id",
},
},
},
},
},
evalCtx: &EvaluationContext{
Action: "sts:AssumeRoleWithWebIdentity",
RequestContext: map[string]interface{}{
"aws:FederatedProvider": "https://provider.com",
"oidc:aud": "my-app-id",
},
},
wantEffect: EffectAllow,
wantErr: false,
},
{
name: "trust policy condition not met",
trustPolicy: &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Effect: "Allow",
Principal: map[string]interface{}{
"Federated": "*",
},
Action: []string{"sts:AssumeRoleWithWebIdentity"},
Condition: map[string]map[string]interface{}{
"StringEquals": {
"oidc:aud": "my-app-id",
},
},
},
},
},
evalCtx: &EvaluationContext{
Action: "sts:AssumeRoleWithWebIdentity",
RequestContext: map[string]interface{}{
"aws:FederatedProvider": "https://provider.com",
"oidc:aud": "wrong-app-id",
},
},
wantEffect: EffectDeny,
wantErr: false,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
result, err := engine.EvaluateTrustPolicy(context.Background(), tt.trustPolicy, tt.evalCtx)
if tt.wantErr {
assert.Error(t, err)
} else {
require.NoError(t, err)
assert.Equal(t, tt.wantEffect, result.Effect, "Trust policy evaluation failed for: %s", tt.name)
}
})
}
}
// TestGetPrincipalContextKey tests the context key mapping
func TestGetPrincipalContextKey(t *testing.T) {
tests := []struct {
name string
principalType string
want string
}{
{
name: "Federated principal",
principalType: "Federated",
want: "aws:FederatedProvider",
},
{
name: "AWS principal",
principalType: "AWS",
want: "aws:PrincipalArn",
},
{
name: "Service principal",
principalType: "Service",
want: "aws:PrincipalServiceName",
},
{
name: "Custom principal type",
principalType: "CustomType",
want: "aws:PrincipalCustomType",
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
result := getPrincipalContextKey(tt.principalType)
assert.Equal(t, tt.want, result, "Context key mapping failed for: %s", tt.name)
})
}
}
-426
View File
@@ -1,426 +0,0 @@
package policy
import (
"context"
"testing"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
// TestPolicyEngineInitialization tests policy engine initialization
func TestPolicyEngineInitialization(t *testing.T) {
tests := []struct {
name string
config *PolicyEngineConfig
wantErr bool
}{
{
name: "valid config",
config: &PolicyEngineConfig{
DefaultEffect: "Deny",
StoreType: "memory",
},
wantErr: false,
},
{
name: "invalid default effect",
config: &PolicyEngineConfig{
DefaultEffect: "Invalid",
StoreType: "memory",
},
wantErr: true,
},
{
name: "nil config",
config: nil,
wantErr: true,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
engine := NewPolicyEngine()
err := engine.Initialize(tt.config)
if tt.wantErr {
assert.Error(t, err)
} else {
assert.NoError(t, err)
assert.True(t, engine.IsInitialized())
}
})
}
}
// TestPolicyDocumentValidation tests policy document structure validation
func TestPolicyDocumentValidation(t *testing.T) {
tests := []struct {
name string
policy *PolicyDocument
wantErr bool
errorMsg string
}{
{
name: "valid policy document",
policy: &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "AllowS3Read",
Effect: "Allow",
Action: []string{"s3:GetObject", "s3:ListBucket"},
Resource: []string{"arn:aws:s3:::mybucket/*"},
},
},
},
wantErr: false,
},
{
name: "missing version",
policy: &PolicyDocument{
Statement: []Statement{
{
Effect: "Allow",
Action: []string{"s3:GetObject"},
Resource: []string{"arn:aws:s3:::mybucket/*"},
},
},
},
wantErr: true,
errorMsg: "version is required",
},
{
name: "empty statements",
policy: &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{},
},
wantErr: true,
errorMsg: "at least one statement is required",
},
{
name: "invalid effect",
policy: &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Effect: "Maybe",
Action: []string{"s3:GetObject"},
Resource: []string{"arn:aws:s3:::mybucket/*"},
},
},
},
wantErr: true,
errorMsg: "invalid effect",
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
err := ValidatePolicyDocument(tt.policy)
if tt.wantErr {
assert.Error(t, err)
if tt.errorMsg != "" {
assert.Contains(t, err.Error(), tt.errorMsg)
}
} else {
assert.NoError(t, err)
}
})
}
}
// TestPolicyEvaluation tests policy evaluation logic
func TestPolicyEvaluation(t *testing.T) {
engine := setupTestPolicyEngine(t)
// Add test policies
readPolicy := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "AllowS3Read",
Effect: "Allow",
Action: []string{"s3:GetObject", "s3:ListBucket"},
Resource: []string{
"arn:aws:s3:::public-bucket/*", // For object operations
"arn:aws:s3:::public-bucket", // For bucket operations
},
},
},
}
err := engine.AddPolicy("", "read-policy", readPolicy)
require.NoError(t, err)
denyPolicy := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "DenyS3Delete",
Effect: "Deny",
Action: []string{"s3:DeleteObject"},
Resource: []string{"arn:aws:s3:::*"},
},
},
}
err = engine.AddPolicy("", "deny-policy", denyPolicy)
require.NoError(t, err)
tests := []struct {
name string
context *EvaluationContext
policies []string
want Effect
}{
{
name: "allow read access",
context: &EvaluationContext{
Principal: "user:alice",
Action: "s3:GetObject",
Resource: "arn:aws:s3:::public-bucket/file.txt",
RequestContext: map[string]interface{}{
"aws:SourceIp": "192.168.1.100",
},
},
policies: []string{"read-policy"},
want: EffectAllow,
},
{
name: "deny delete access (explicit deny)",
context: &EvaluationContext{
Principal: "user:alice",
Action: "s3:DeleteObject",
Resource: "arn:aws:s3:::public-bucket/file.txt",
},
policies: []string{"read-policy", "deny-policy"},
want: EffectDeny,
},
{
name: "deny by default (no matching policy)",
context: &EvaluationContext{
Principal: "user:alice",
Action: "s3:PutObject",
Resource: "arn:aws:s3:::public-bucket/file.txt",
},
policies: []string{"read-policy"},
want: EffectDeny,
},
{
name: "allow with wildcard action",
context: &EvaluationContext{
Principal: "user:admin",
Action: "s3:ListBucket",
Resource: "arn:aws:s3:::public-bucket",
},
policies: []string{"read-policy"},
want: EffectAllow,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
result, err := engine.Evaluate(context.Background(), "", tt.context, tt.policies)
assert.NoError(t, err)
assert.Equal(t, tt.want, result.Effect)
// Verify evaluation details
assert.NotNil(t, result.EvaluationDetails)
assert.Equal(t, tt.context.Action, result.EvaluationDetails.Action)
assert.Equal(t, tt.context.Resource, result.EvaluationDetails.Resource)
})
}
}
// TestConditionEvaluation tests policy conditions
func TestConditionEvaluation(t *testing.T) {
engine := setupTestPolicyEngine(t)
// Policy with IP address condition
conditionalPolicy := &PolicyDocument{
Version: "2012-10-17",
Statement: []Statement{
{
Sid: "AllowFromOfficeIP",
Effect: "Allow",
Action: []string{"s3:*"},
Resource: []string{"arn:aws:s3:::*"},
Condition: map[string]map[string]interface{}{
"IpAddress": {
"aws:SourceIp": []string{"192.168.1.0/24", "10.0.0.0/8"},
},
},
},
},
}
err := engine.AddPolicy("", "ip-conditional", conditionalPolicy)
require.NoError(t, err)
tests := []struct {
name string
context *EvaluationContext
want Effect
}{
{
name: "allow from office IP",
context: &EvaluationContext{
Principal: "user:alice",
Action: "s3:GetObject",
Resource: "arn:aws:s3:::mybucket/file.txt",
RequestContext: map[string]interface{}{
"aws:SourceIp": "192.168.1.100",
},
},
want: EffectAllow,
},
{
name: "deny from external IP",
context: &EvaluationContext{
Principal: "user:alice",
Action: "s3:GetObject",
Resource: "arn:aws:s3:::mybucket/file.txt",
RequestContext: map[string]interface{}{
"aws:SourceIp": "8.8.8.8",
},
},
want: EffectDeny,
},
{
name: "allow from internal IP",
context: &EvaluationContext{
Principal: "user:alice",
Action: "s3:PutObject",
Resource: "arn:aws:s3:::mybucket/newfile.txt",
RequestContext: map[string]interface{}{
"aws:SourceIp": "10.1.2.3",
},
},
want: EffectAllow,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
result, err := engine.Evaluate(context.Background(), "", tt.context, []string{"ip-conditional"})
assert.NoError(t, err)
assert.Equal(t, tt.want, result.Effect)
})
}
}
// TestResourceMatching tests resource ARN matching
func TestResourceMatching(t *testing.T) {
tests := []struct {
name string
policyResource string
requestResource string
want bool
}{
{
name: "exact match",
policyResource: "arn:aws:s3:::mybucket/file.txt",
requestResource: "arn:aws:s3:::mybucket/file.txt",
want: true,
},
{
name: "wildcard match",
policyResource: "arn:aws:s3:::mybucket/*",
requestResource: "arn:aws:s3:::mybucket/folder/file.txt",
want: true,
},
{
name: "bucket wildcard",
policyResource: "arn:aws:s3:::*",
requestResource: "arn:aws:s3:::anybucket/file.txt",
want: true,
},
{
name: "no match different bucket",
policyResource: "arn:aws:s3:::mybucket/*",
requestResource: "arn:aws:s3:::otherbucket/file.txt",
want: false,
},
{
name: "prefix match",
policyResource: "arn:aws:s3:::mybucket/documents/*",
requestResource: "arn:aws:s3:::mybucket/documents/secret.txt",
want: true,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
result := matchResource(tt.policyResource, tt.requestResource)
assert.Equal(t, tt.want, result)
})
}
}
// TestActionMatching tests action pattern matching
func TestActionMatching(t *testing.T) {
tests := []struct {
name string
policyAction string
requestAction string
want bool
}{
{
name: "exact match",
policyAction: "s3:GetObject",
requestAction: "s3:GetObject",
want: true,
},
{
name: "wildcard service",
policyAction: "s3:*",
requestAction: "s3:PutObject",
want: true,
},
{
name: "wildcard all",
policyAction: "*",
requestAction: "filer:CreateEntry",
want: true,
},
{
name: "prefix match",
policyAction: "s3:Get*",
requestAction: "s3:GetObject",
want: true,
},
{
name: "no match different service",
policyAction: "s3:GetObject",
requestAction: "filer:GetEntry",
want: false,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
result := matchAction(tt.policyAction, tt.requestAction)
assert.Equal(t, tt.want, result)
})
}
}
// Helper function to set up test policy engine
func setupTestPolicyEngine(t *testing.T) *PolicyEngine {
engine := NewPolicyEngine()
config := &PolicyEngineConfig{
DefaultEffect: "Deny",
StoreType: "memory",
}
err := engine.Initialize(config)
require.NoError(t, err)
return engine
}
-246
View File
@@ -1,246 +0,0 @@
package providers
import (
"context"
"testing"
"time"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
// TestIdentityProviderInterface tests the core identity provider interface
func TestIdentityProviderInterface(t *testing.T) {
tests := []struct {
name string
provider IdentityProvider
wantErr bool
}{
// We'll add test cases as we implement providers
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
// Test provider name
name := tt.provider.Name()
assert.NotEmpty(t, name, "Provider name should not be empty")
// Test initialization
err := tt.provider.Initialize(nil)
if tt.wantErr {
assert.Error(t, err)
return
}
require.NoError(t, err)
// Test authentication with invalid token
ctx := context.Background()
_, err = tt.provider.Authenticate(ctx, "invalid-token")
assert.Error(t, err, "Should fail with invalid token")
})
}
}
// TestExternalIdentityValidation tests external identity structure validation
func TestExternalIdentityValidation(t *testing.T) {
tests := []struct {
name string
identity *ExternalIdentity
wantErr bool
}{
{
name: "valid identity",
identity: &ExternalIdentity{
UserID: "user123",
Email: "user@example.com",
DisplayName: "Test User",
Groups: []string{"group1", "group2"},
Attributes: map[string]string{"dept": "engineering"},
Provider: "test-provider",
},
wantErr: false,
},
{
name: "missing user id",
identity: &ExternalIdentity{
Email: "user@example.com",
Provider: "test-provider",
},
wantErr: true,
},
{
name: "missing provider",
identity: &ExternalIdentity{
UserID: "user123",
Email: "user@example.com",
},
wantErr: true,
},
{
name: "invalid email",
identity: &ExternalIdentity{
UserID: "user123",
Email: "invalid-email",
Provider: "test-provider",
},
wantErr: true,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
err := tt.identity.Validate()
if tt.wantErr {
assert.Error(t, err)
} else {
assert.NoError(t, err)
}
})
}
}
// TestTokenClaimsValidation tests token claims structure
func TestTokenClaimsValidation(t *testing.T) {
tests := []struct {
name string
claims *TokenClaims
valid bool
}{
{
name: "valid claims",
claims: &TokenClaims{
Subject: "user123",
Issuer: "https://provider.example.com",
Audience: "seaweedfs",
ExpiresAt: time.Now().Add(time.Hour),
IssuedAt: time.Now().Add(-time.Minute),
Claims: map[string]interface{}{"email": "user@example.com"},
},
valid: true,
},
{
name: "expired token",
claims: &TokenClaims{
Subject: "user123",
Issuer: "https://provider.example.com",
Audience: "seaweedfs",
ExpiresAt: time.Now().Add(-time.Hour), // Expired
IssuedAt: time.Now().Add(-time.Hour * 2),
Claims: map[string]interface{}{"email": "user@example.com"},
},
valid: false,
},
{
name: "future issued token",
claims: &TokenClaims{
Subject: "user123",
Issuer: "https://provider.example.com",
Audience: "seaweedfs",
ExpiresAt: time.Now().Add(time.Hour),
IssuedAt: time.Now().Add(time.Hour), // Future
Claims: map[string]interface{}{"email": "user@example.com"},
},
valid: false,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
valid := tt.claims.IsValid()
assert.Equal(t, tt.valid, valid)
})
}
}
// TestProviderRegistry tests provider registration and discovery
func TestProviderRegistry(t *testing.T) {
// Clear registry for test
registry := NewProviderRegistry()
t.Run("register provider", func(t *testing.T) {
mockProvider := &MockProvider{name: "test-provider"}
err := registry.RegisterProvider(mockProvider)
assert.NoError(t, err)
// Test duplicate registration
err = registry.RegisterProvider(mockProvider)
assert.Error(t, err, "Should not allow duplicate registration")
})
t.Run("get provider", func(t *testing.T) {
provider, exists := registry.GetProvider("test-provider")
assert.True(t, exists)
assert.Equal(t, "test-provider", provider.Name())
// Test non-existent provider
_, exists = registry.GetProvider("non-existent")
assert.False(t, exists)
})
t.Run("list providers", func(t *testing.T) {
providers := registry.ListProviders()
assert.Len(t, providers, 1)
assert.Equal(t, "test-provider", providers[0])
})
}
// MockProvider for testing
type MockProvider struct {
name string
initialized bool
shouldError bool
}
func (m *MockProvider) Name() string {
return m.name
}
func (m *MockProvider) Initialize(config interface{}) error {
if m.shouldError {
return assert.AnError
}
m.initialized = true
return nil
}
func (m *MockProvider) Authenticate(ctx context.Context, token string) (*ExternalIdentity, error) {
if !m.initialized {
return nil, assert.AnError
}
if token == "invalid-token" {
return nil, assert.AnError
}
return &ExternalIdentity{
UserID: "test-user",
Email: "test@example.com",
DisplayName: "Test User",
Provider: m.name,
}, nil
}
func (m *MockProvider) GetUserInfo(ctx context.Context, userID string) (*ExternalIdentity, error) {
if !m.initialized || userID == "" {
return nil, assert.AnError
}
return &ExternalIdentity{
UserID: userID,
Email: userID + "@example.com",
DisplayName: "User " + userID,
Provider: m.name,
}, nil
}
func (m *MockProvider) ValidateToken(ctx context.Context, token string) (*TokenClaims, error) {
if !m.initialized || token == "invalid-token" {
return nil, assert.AnError
}
return &TokenClaims{
Subject: "test-user",
Issuer: "test-issuer",
Audience: "seaweedfs",
ExpiresAt: time.Now().Add(time.Hour),
IssuedAt: time.Now(),
Claims: map[string]interface{}{"email": "test@example.com"},
}, nil
}
-109
View File
@@ -1,109 +0,0 @@
package providers
import (
"fmt"
"sync"
)
// ProviderRegistry manages registered identity providers
type ProviderRegistry struct {
mu sync.RWMutex
providers map[string]IdentityProvider
}
// NewProviderRegistry creates a new provider registry
func NewProviderRegistry() *ProviderRegistry {
return &ProviderRegistry{
providers: make(map[string]IdentityProvider),
}
}
// RegisterProvider registers a new identity provider
func (r *ProviderRegistry) RegisterProvider(provider IdentityProvider) error {
if provider == nil {
return fmt.Errorf("provider cannot be nil")
}
name := provider.Name()
if name == "" {
return fmt.Errorf("provider name cannot be empty")
}
r.mu.Lock()
defer r.mu.Unlock()
if _, exists := r.providers[name]; exists {
return fmt.Errorf("provider %s is already registered", name)
}
r.providers[name] = provider
return nil
}
// GetProvider retrieves a provider by name
func (r *ProviderRegistry) GetProvider(name string) (IdentityProvider, bool) {
r.mu.RLock()
defer r.mu.RUnlock()
provider, exists := r.providers[name]
return provider, exists
}
// ListProviders returns all registered provider names
func (r *ProviderRegistry) ListProviders() []string {
r.mu.RLock()
defer r.mu.RUnlock()
var names []string
for name := range r.providers {
names = append(names, name)
}
return names
}
// UnregisterProvider removes a provider from the registry
func (r *ProviderRegistry) UnregisterProvider(name string) error {
r.mu.Lock()
defer r.mu.Unlock()
if _, exists := r.providers[name]; !exists {
return fmt.Errorf("provider %s is not registered", name)
}
delete(r.providers, name)
return nil
}
// Clear removes all providers from the registry
func (r *ProviderRegistry) Clear() {
r.mu.Lock()
defer r.mu.Unlock()
r.providers = make(map[string]IdentityProvider)
}
// GetProviderCount returns the number of registered providers
func (r *ProviderRegistry) GetProviderCount() int {
r.mu.RLock()
defer r.mu.RUnlock()
return len(r.providers)
}
// Default global registry
var defaultRegistry = NewProviderRegistry()
// RegisterProvider registers a provider in the default registry
func RegisterProvider(provider IdentityProvider) error {
return defaultRegistry.RegisterProvider(provider)
}
// GetProvider retrieves a provider from the default registry
func GetProvider(name string) (IdentityProvider, bool) {
return defaultRegistry.GetProvider(name)
}
// ListProviders returns all provider names from the default registry
func ListProviders() []string {
return defaultRegistry.ListProviders()
}
+1
View File
@@ -124,6 +124,7 @@ const (
ActionAssumeRole = "sts:AssumeRole"
ActionAssumeRoleWithWebIdentity = "sts:AssumeRoleWithWebIdentity"
ActionAssumeRoleWithCredentials = "sts:AssumeRoleWithCredentials"
ActionGetFederationToken = "sts:GetFederationToken"
ActionValidateSession = "sts:ValidateSession"
)
-503
View File
@@ -1,503 +0,0 @@
package sts
import (
"context"
"testing"
"time"
"github.com/golang-jwt/jwt/v5"
"github.com/seaweedfs/seaweedfs/weed/iam/oidc"
"github.com/seaweedfs/seaweedfs/weed/iam/providers"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
// Test-only constants for mock providers
const (
ProviderTypeMock = "mock"
)
// createMockOIDCProvider creates a mock OIDC provider for testing
// This is only available in test builds
func createMockOIDCProvider(name string, config map[string]interface{}) (providers.IdentityProvider, error) {
// Convert config to OIDC format
factory := NewProviderFactory()
oidcConfig, err := factory.convertToOIDCConfig(config)
if err != nil {
return nil, err
}
// Set default values for mock provider if not provided
if oidcConfig.Issuer == "" {
oidcConfig.Issuer = "http://localhost:9999"
}
provider := oidc.NewMockOIDCProvider(name)
if err := provider.Initialize(oidcConfig); err != nil {
return nil, err
}
// Set up default test data for the mock provider
provider.SetupDefaultTestData()
return provider, nil
}
// createMockJWT creates a test JWT token with the specified issuer for mock provider testing
func createMockJWT(t *testing.T, issuer, subject string) string {
token := jwt.NewWithClaims(jwt.SigningMethodHS256, jwt.MapClaims{
"iss": issuer,
"sub": subject,
"aud": "test-client",
"exp": time.Now().Add(time.Hour).Unix(),
"iat": time.Now().Unix(),
})
tokenString, err := token.SignedString([]byte("test-signing-key"))
require.NoError(t, err)
return tokenString
}
// TestCrossInstanceTokenUsage verifies that tokens generated by one STS instance
// can be used and validated by other STS instances in a distributed environment
func TestCrossInstanceTokenUsage(t *testing.T) {
ctx := context.Background()
// Dummy filer address for testing
// Common configuration that would be shared across all instances in production
sharedConfig := &STSConfig{
TokenDuration: FlexibleDuration{time.Hour},
MaxSessionLength: FlexibleDuration{12 * time.Hour},
Issuer: "distributed-sts-cluster", // SAME across all instances
SigningKey: []byte(TestSigningKey32Chars), // SAME across all instances
Providers: []*ProviderConfig{
{
Name: "company-oidc",
Type: ProviderTypeOIDC,
Enabled: true,
Config: map[string]interface{}{
ConfigFieldIssuer: "https://sso.company.com/realms/production",
ConfigFieldClientID: "seaweedfs-cluster",
ConfigFieldJWKSUri: "https://sso.company.com/realms/production/protocol/openid-connect/certs",
},
},
},
}
// Create multiple STS instances simulating different S3 gateway instances
instanceA := NewSTSService() // e.g., s3-gateway-1
instanceB := NewSTSService() // e.g., s3-gateway-2
instanceC := NewSTSService() // e.g., s3-gateway-3
// Initialize all instances with IDENTICAL configuration
err := instanceA.Initialize(sharedConfig)
require.NoError(t, err, "Instance A should initialize")
err = instanceB.Initialize(sharedConfig)
require.NoError(t, err, "Instance B should initialize")
err = instanceC.Initialize(sharedConfig)
require.NoError(t, err, "Instance C should initialize")
// Set up mock trust policy validator for all instances (required for STS testing)
mockValidator := &MockTrustPolicyValidator{}
instanceA.SetTrustPolicyValidator(mockValidator)
instanceB.SetTrustPolicyValidator(mockValidator)
instanceC.SetTrustPolicyValidator(mockValidator)
// Manually register mock provider for testing (not available in production)
mockProviderConfig := map[string]interface{}{
ConfigFieldIssuer: "http://test-mock:9999",
ConfigFieldClientID: TestClientID,
}
mockProviderA, err := createMockOIDCProvider("test-mock", mockProviderConfig)
require.NoError(t, err)
mockProviderB, err := createMockOIDCProvider("test-mock", mockProviderConfig)
require.NoError(t, err)
mockProviderC, err := createMockOIDCProvider("test-mock", mockProviderConfig)
require.NoError(t, err)
instanceA.RegisterProvider(mockProviderA)
instanceB.RegisterProvider(mockProviderB)
instanceC.RegisterProvider(mockProviderC)
// Test 1: Token generated on Instance A can be validated on Instance B & C
t.Run("cross_instance_token_validation", func(t *testing.T) {
// Generate session token on Instance A
sessionId := TestSessionID
expiresAt := time.Now().Add(time.Hour)
tokenFromA, err := instanceA.GetTokenGenerator().GenerateSessionToken(sessionId, expiresAt)
require.NoError(t, err, "Instance A should generate token")
// Validate token on Instance B
claimsFromB, err := instanceB.GetTokenGenerator().ValidateSessionToken(tokenFromA)
require.NoError(t, err, "Instance B should validate token from Instance A")
assert.Equal(t, sessionId, claimsFromB.SessionId, "Session ID should match")
// Validate same token on Instance C
claimsFromC, err := instanceC.GetTokenGenerator().ValidateSessionToken(tokenFromA)
require.NoError(t, err, "Instance C should validate token from Instance A")
assert.Equal(t, sessionId, claimsFromC.SessionId, "Session ID should match")
// All instances should extract identical claims
assert.Equal(t, claimsFromB.SessionId, claimsFromC.SessionId)
assert.Equal(t, claimsFromB.ExpiresAt.Unix(), claimsFromC.ExpiresAt.Unix())
assert.Equal(t, claimsFromB.IssuedAt.Unix(), claimsFromC.IssuedAt.Unix())
})
// Test 2: Complete assume role flow across instances
t.Run("cross_instance_assume_role_flow", func(t *testing.T) {
// Step 1: User authenticates and assumes role on Instance A
// Create a valid JWT token for the mock provider
mockToken := createMockJWT(t, "http://test-mock:9999", "test-user")
assumeRequest := &AssumeRoleWithWebIdentityRequest{
RoleArn: "arn:aws:iam::role/CrossInstanceTestRole",
WebIdentityToken: mockToken, // JWT token for mock provider
RoleSessionName: "cross-instance-test-session",
DurationSeconds: int64ToPtr(3600),
}
// Instance A processes assume role request
responseFromA, err := instanceA.AssumeRoleWithWebIdentity(ctx, assumeRequest)
require.NoError(t, err, "Instance A should process assume role")
sessionToken := responseFromA.Credentials.SessionToken
accessKeyId := responseFromA.Credentials.AccessKeyId
secretAccessKey := responseFromA.Credentials.SecretAccessKey
// Verify response structure
assert.NotEmpty(t, sessionToken, "Should have session token")
assert.NotEmpty(t, accessKeyId, "Should have access key ID")
assert.NotEmpty(t, secretAccessKey, "Should have secret access key")
assert.NotNil(t, responseFromA.AssumedRoleUser, "Should have assumed role user")
// Step 2: Use session token on Instance B (different instance)
sessionInfoFromB, err := instanceB.ValidateSessionToken(ctx, sessionToken)
require.NoError(t, err, "Instance B should validate session token from Instance A")
assert.Equal(t, assumeRequest.RoleSessionName, sessionInfoFromB.SessionName)
assert.Equal(t, assumeRequest.RoleArn, sessionInfoFromB.RoleArn)
// Step 3: Use same session token on Instance C (yet another instance)
sessionInfoFromC, err := instanceC.ValidateSessionToken(ctx, sessionToken)
require.NoError(t, err, "Instance C should validate session token from Instance A")
// All instances should return identical session information
assert.Equal(t, sessionInfoFromB.SessionId, sessionInfoFromC.SessionId)
assert.Equal(t, sessionInfoFromB.SessionName, sessionInfoFromC.SessionName)
assert.Equal(t, sessionInfoFromB.RoleArn, sessionInfoFromC.RoleArn)
assert.Equal(t, sessionInfoFromB.Subject, sessionInfoFromC.Subject)
assert.Equal(t, sessionInfoFromB.Provider, sessionInfoFromC.Provider)
})
// Test 3: Session revocation across instances
t.Run("cross_instance_session_revocation", func(t *testing.T) {
// Create session on Instance A
mockToken := createMockJWT(t, "http://test-mock:9999", "test-user")
assumeRequest := &AssumeRoleWithWebIdentityRequest{
RoleArn: "arn:aws:iam::role/RevocationTestRole",
WebIdentityToken: mockToken,
RoleSessionName: "revocation-test-session",
}
response, err := instanceA.AssumeRoleWithWebIdentity(ctx, assumeRequest)
require.NoError(t, err)
sessionToken := response.Credentials.SessionToken
// Verify token works on Instance B
_, err = instanceB.ValidateSessionToken(ctx, sessionToken)
require.NoError(t, err, "Token should be valid on Instance B initially")
// Validate session on Instance C to verify cross-instance token compatibility
_, err = instanceC.ValidateSessionToken(ctx, sessionToken)
require.NoError(t, err, "Instance C should be able to validate session token")
// In a stateless JWT system, tokens remain valid on all instances since they're self-contained
// No revocation is possible without breaking the stateless architecture
_, err = instanceA.ValidateSessionToken(ctx, sessionToken)
assert.NoError(t, err, "Token should still be valid on Instance A (stateless system)")
// Verify token is still valid on Instance B
_, err = instanceB.ValidateSessionToken(ctx, sessionToken)
assert.NoError(t, err, "Token should still be valid on Instance B (stateless system)")
})
// Test 4: Provider consistency across instances
t.Run("provider_consistency_affects_token_generation", func(t *testing.T) {
// All instances should have same providers and be able to process same OIDC tokens
providerNamesA := instanceA.getProviderNames()
providerNamesB := instanceB.getProviderNames()
providerNamesC := instanceC.getProviderNames()
assert.ElementsMatch(t, providerNamesA, providerNamesB, "Instance A and B should have same providers")
assert.ElementsMatch(t, providerNamesB, providerNamesC, "Instance B and C should have same providers")
// All instances should be able to process same web identity token
testToken := createMockJWT(t, "http://test-mock:9999", "test-user")
// Try to assume role with same token on different instances
assumeRequest := &AssumeRoleWithWebIdentityRequest{
RoleArn: "arn:aws:iam::role/ProviderTestRole",
WebIdentityToken: testToken,
RoleSessionName: "provider-consistency-test",
}
// Should work on any instance
responseA, errA := instanceA.AssumeRoleWithWebIdentity(ctx, assumeRequest)
responseB, errB := instanceB.AssumeRoleWithWebIdentity(ctx, assumeRequest)
responseC, errC := instanceC.AssumeRoleWithWebIdentity(ctx, assumeRequest)
require.NoError(t, errA, "Instance A should process OIDC token")
require.NoError(t, errB, "Instance B should process OIDC token")
require.NoError(t, errC, "Instance C should process OIDC token")
// All should return valid responses (sessions will have different IDs but same structure)
assert.NotEmpty(t, responseA.Credentials.SessionToken)
assert.NotEmpty(t, responseB.Credentials.SessionToken)
assert.NotEmpty(t, responseC.Credentials.SessionToken)
})
}
// TestSTSDistributedConfigurationRequirements tests the configuration requirements
// for cross-instance token compatibility
func TestSTSDistributedConfigurationRequirements(t *testing.T) {
_ = "localhost:8888" // Dummy filer address for testing (not used in these tests)
t.Run("same_signing_key_required", func(t *testing.T) {
// Instance A with signing key 1
configA := &STSConfig{
TokenDuration: FlexibleDuration{time.Hour},
MaxSessionLength: FlexibleDuration{12 * time.Hour},
Issuer: "test-sts",
SigningKey: []byte("signing-key-1-32-characters-long"),
}
// Instance B with different signing key
configB := &STSConfig{
TokenDuration: FlexibleDuration{time.Hour},
MaxSessionLength: FlexibleDuration{12 * time.Hour},
Issuer: "test-sts",
SigningKey: []byte("signing-key-2-32-characters-long"), // DIFFERENT!
}
instanceA := NewSTSService()
instanceB := NewSTSService()
err := instanceA.Initialize(configA)
require.NoError(t, err)
err = instanceB.Initialize(configB)
require.NoError(t, err)
// Generate token on Instance A
sessionId := "test-session"
expiresAt := time.Now().Add(time.Hour)
tokenFromA, err := instanceA.GetTokenGenerator().GenerateSessionToken(sessionId, expiresAt)
require.NoError(t, err)
// Instance A should validate its own token
_, err = instanceA.GetTokenGenerator().ValidateSessionToken(tokenFromA)
assert.NoError(t, err, "Instance A should validate own token")
// Instance B should REJECT token due to different signing key
_, err = instanceB.GetTokenGenerator().ValidateSessionToken(tokenFromA)
assert.Error(t, err, "Instance B should reject token with different signing key")
assert.Contains(t, err.Error(), "invalid token", "Should be signature validation error")
})
t.Run("same_issuer_required", func(t *testing.T) {
sharedSigningKey := []byte("shared-signing-key-32-characters-lo")
// Instance A with issuer 1
configA := &STSConfig{
TokenDuration: FlexibleDuration{time.Hour},
MaxSessionLength: FlexibleDuration{12 * time.Hour},
Issuer: "sts-cluster-1",
SigningKey: sharedSigningKey,
}
// Instance B with different issuer
configB := &STSConfig{
TokenDuration: FlexibleDuration{time.Hour},
MaxSessionLength: FlexibleDuration{12 * time.Hour},
Issuer: "sts-cluster-2", // DIFFERENT!
SigningKey: sharedSigningKey,
}
instanceA := NewSTSService()
instanceB := NewSTSService()
err := instanceA.Initialize(configA)
require.NoError(t, err)
err = instanceB.Initialize(configB)
require.NoError(t, err)
// Generate token on Instance A
sessionId := "test-session"
expiresAt := time.Now().Add(time.Hour)
tokenFromA, err := instanceA.GetTokenGenerator().GenerateSessionToken(sessionId, expiresAt)
require.NoError(t, err)
// Instance B should REJECT token due to different issuer
_, err = instanceB.GetTokenGenerator().ValidateSessionToken(tokenFromA)
assert.Error(t, err, "Instance B should reject token with different issuer")
assert.Contains(t, err.Error(), "invalid issuer", "Should be issuer validation error")
})
t.Run("identical_configuration_required", func(t *testing.T) {
// Identical configuration
identicalConfig := &STSConfig{
TokenDuration: FlexibleDuration{time.Hour},
MaxSessionLength: FlexibleDuration{12 * time.Hour},
Issuer: "production-sts-cluster",
SigningKey: []byte("production-signing-key-32-chars-l"),
}
// Create multiple instances with identical config
instances := make([]*STSService, 5)
for i := 0; i < 5; i++ {
instances[i] = NewSTSService()
err := instances[i].Initialize(identicalConfig)
require.NoError(t, err, "Instance %d should initialize", i)
}
// Generate token on Instance 0
sessionId := "multi-instance-test"
expiresAt := time.Now().Add(time.Hour)
token, err := instances[0].GetTokenGenerator().GenerateSessionToken(sessionId, expiresAt)
require.NoError(t, err)
// All other instances should validate the token
for i := 1; i < 5; i++ {
claims, err := instances[i].GetTokenGenerator().ValidateSessionToken(token)
require.NoError(t, err, "Instance %d should validate token", i)
assert.Equal(t, sessionId, claims.SessionId, "Instance %d should extract correct session ID", i)
}
})
}
// TestSTSRealWorldDistributedScenarios tests realistic distributed deployment scenarios
func TestSTSRealWorldDistributedScenarios(t *testing.T) {
ctx := context.Background()
t.Run("load_balanced_s3_gateway_scenario", func(t *testing.T) {
// Simulate real production scenario:
// 1. User authenticates with OIDC provider
// 2. User calls AssumeRoleWithWebIdentity on S3 Gateway 1
// 3. User makes S3 requests that hit S3 Gateway 2 & 3 via load balancer
// 4. All instances should handle the session token correctly
productionConfig := &STSConfig{
TokenDuration: FlexibleDuration{2 * time.Hour},
MaxSessionLength: FlexibleDuration{24 * time.Hour},
Issuer: "seaweedfs-production-sts",
SigningKey: []byte("prod-signing-key-32-characters-lon"),
Providers: []*ProviderConfig{
{
Name: "corporate-oidc",
Type: "oidc",
Enabled: true,
Config: map[string]interface{}{
"issuer": "https://sso.company.com/realms/production",
"clientId": "seaweedfs-prod-cluster",
"clientSecret": "supersecret-prod-key",
"scopes": []string{"openid", "profile", "email", "groups"},
},
},
},
}
// Create 3 S3 Gateway instances behind load balancer
gateway1 := NewSTSService()
gateway2 := NewSTSService()
gateway3 := NewSTSService()
err := gateway1.Initialize(productionConfig)
require.NoError(t, err)
err = gateway2.Initialize(productionConfig)
require.NoError(t, err)
err = gateway3.Initialize(productionConfig)
require.NoError(t, err)
// Set up mock trust policy validator for all gateway instances
mockValidator := &MockTrustPolicyValidator{}
gateway1.SetTrustPolicyValidator(mockValidator)
gateway2.SetTrustPolicyValidator(mockValidator)
gateway3.SetTrustPolicyValidator(mockValidator)
// Manually register mock provider for testing (not available in production)
mockProviderConfig := map[string]interface{}{
ConfigFieldIssuer: "http://test-mock:9999",
ConfigFieldClientID: "test-client-id",
}
mockProvider1, err := createMockOIDCProvider("test-mock", mockProviderConfig)
require.NoError(t, err)
mockProvider2, err := createMockOIDCProvider("test-mock", mockProviderConfig)
require.NoError(t, err)
mockProvider3, err := createMockOIDCProvider("test-mock", mockProviderConfig)
require.NoError(t, err)
gateway1.RegisterProvider(mockProvider1)
gateway2.RegisterProvider(mockProvider2)
gateway3.RegisterProvider(mockProvider3)
// Step 1: User authenticates and hits Gateway 1 for AssumeRole
mockToken := createMockJWT(t, "http://test-mock:9999", "production-user")
assumeRequest := &AssumeRoleWithWebIdentityRequest{
RoleArn: "arn:aws:iam::role/ProductionS3User",
WebIdentityToken: mockToken, // JWT token from mock provider
RoleSessionName: "user-production-session",
DurationSeconds: int64ToPtr(7200), // 2 hours
}
stsResponse, err := gateway1.AssumeRoleWithWebIdentity(ctx, assumeRequest)
require.NoError(t, err, "Gateway 1 should handle AssumeRole")
sessionToken := stsResponse.Credentials.SessionToken
accessKey := stsResponse.Credentials.AccessKeyId
secretKey := stsResponse.Credentials.SecretAccessKey
// Step 2: User makes S3 requests that hit different gateways via load balancer
// Simulate S3 request validation on Gateway 2
sessionInfo2, err := gateway2.ValidateSessionToken(ctx, sessionToken)
require.NoError(t, err, "Gateway 2 should validate session from Gateway 1")
assert.Equal(t, "user-production-session", sessionInfo2.SessionName)
assert.Equal(t, "arn:aws:iam::role/ProductionS3User", sessionInfo2.RoleArn)
// Simulate S3 request validation on Gateway 3
sessionInfo3, err := gateway3.ValidateSessionToken(ctx, sessionToken)
require.NoError(t, err, "Gateway 3 should validate session from Gateway 1")
assert.Equal(t, sessionInfo2.SessionId, sessionInfo3.SessionId, "Should be same session")
// Step 3: Verify credentials are consistent
assert.Equal(t, accessKey, stsResponse.Credentials.AccessKeyId, "Access key should be consistent")
assert.Equal(t, secretKey, stsResponse.Credentials.SecretAccessKey, "Secret key should be consistent")
// Step 4: Session expiration should be honored across all instances
assert.True(t, sessionInfo2.ExpiresAt.After(time.Now()), "Session should not be expired")
assert.True(t, sessionInfo3.ExpiresAt.After(time.Now()), "Session should not be expired")
// Step 5: Token should be identical when parsed
claims2, err := gateway2.GetTokenGenerator().ValidateSessionToken(sessionToken)
require.NoError(t, err)
claims3, err := gateway3.GetTokenGenerator().ValidateSessionToken(sessionToken)
require.NoError(t, err)
assert.Equal(t, claims2.SessionId, claims3.SessionId, "Session IDs should match")
assert.Equal(t, claims2.ExpiresAt.Unix(), claims3.ExpiresAt.Unix(), "Expiration should match")
})
}
// Helper function to convert int64 to pointer
func int64ToPtr(i int64) *int64 {
return &i
}

Some files were not shown because too many files have changed in this diff Show More