This pull request introduces a new caching mechanism for client options in the Alternator and transport layers, refactors how client metadata is stored and accessed, and extends the `system.clients` virtual table to surface richer client information. The changes improve efficiency by deduplicating commonly used strings (like driver names/versions and client options), and ensure that client data is handled in a way that's safe for cross-shard access. Additionally, the test suite and virtual table schema are updated to reflect the new client options data.
**Caching and client metadata refactoring:**
* The largest and most repeatable items in the connection state before this PR were a `driver_name` and a `driver_version` which were stored as an `sstring` object which means that the corresponding memory consumption was 16 bytes per each such value at least (the smallest size of the `seastar`'s `sstring` object) **per-connection**. In reality the driver name is usually longer than 15 characters, e.g. "ScyllaDB Python Driver" is 23 characters and this is not the longest driver name there is. In such cases the actual memory usage of a corresponding `sstring` object jumps to 8 + 4 + 1 + (string length, 23 in our example) + 1.
So, for "ScyllaDB Python Driver" it would be 37 bytes (in reality it would be a bit more due to natural alignment of other allocations since the `contents` size is not well aligned (13 bytes), but let's ignore this for now).
* These bytes add up quickly as there are more connections and, sometimes we are talking about millions of connections per-shard.
* Using a smart pointer (`lw_shared_ptr`) referencing a corresponding cached value will effectively reduce the per-connection memory usage to be 8 bytes (a size of a pointer on 64-bit CPU platform) for each such value. While storing a corresponding `sstring` value only once.
* This will would reduce the "variable" (per-connection) memory usage by **at least 50%**. And in case of "ScyllaDB Python Driver" driver version - by 78%!
* And all this for a price of a single `loading_shared_values` object **per-shard** (implements a hash table) and a minor overhead for each value **stored** in it.
* Introduced a new cache type (`client_options_cache_type`) for deduplicating and sharing client option strings, and refactored `client_data`, `client_state`, and related classes to use `foreign_ptr<std::unique_ptr<client_data>>` and cached entry types for fields like driver name, driver version, and client options. (`client_data.hh`, `service/client_state.hh`, `alternator/server.hh`, `alternator/controller.hh`, `transport/controller.hh`, `transport/protocol_server.hh`) [[1]](diffhunk://#diff-664a3b19e905481bdf8eb3843fc4d34691067bb97ab11cfd6e652e74aac51d9fR33-R36) [[2]](diffhunk://#diff-664a3b19e905481bdf8eb3843fc4d34691067bb97ab11cfd6e652e74aac51d9fL40-R56) [[3]](diffhunk://#diff-daadce1a2de3667511e59558f3a8f077b5ee30a14bcc6a99d588db90d0fcd2bdL105-R107) [[4]](diffhunk://#diff-daadce1a2de3667511e59558f3a8f077b5ee30a14bcc6a99d588db90d0fcd2bdL154-R182) [[5]](diffhunk://#diff-5fce246edf5abffb2351bd02e2eb1e9850880f7a00607ccaa90c3eee7ef57c6bL91-R92) [[6]](diffhunk://#diff-5fce246edf5abffb2351bd02e2eb1e9850880f7a00607ccaa90c3eee7ef57c6bL110-R111) [[7]](diffhunk://#diff-31730ba8e7374f784a88dc27c1512291cf73b7f24e08768f7466a3c8cfcc7a1aL96-R96) [[8]](diffhunk://#diff-19a97c0247cc08155ee49b277e43859ca32d6ef8cbff0ed7368ec5fa19e0a11eL172-R172) [[9]](diffhunk://#diff-eea7e2db5d799a25e717a72ac8ce5842bd4adb72b694d38d8f47166d9cd926faL356-R356) [[10]](diffhunk://#diff-d0b4ec3a144bbc5dc993866cf0b940850a457ff6156064f7e2b4b10ad0a95fefL80-R80) [[11]](diffhunk://#diff-4293b94c444d9bd5ecd17ce7eda8c00685d35ecf6e07f844efc91a91bbe85be1L46-R48)
* Updated the methods for setting and getting driver name, driver version, and client options in `client_state` to be asynchronous and use the new cache. (`service/client_state.hh`, `service/client_state.cc`) [[1]](diffhunk://#diff-daadce1a2de3667511e59558f3a8f077b5ee30a14bcc6a99d588db90d0fcd2bdL154-R182) [[2]](diffhunk://#diff-99634aae22e2573f38b4e2f050ed2ac4f8173ff27f0ae8b3609d1f0cc1aeb775R347-R362)
**Virtual table and API enhancements:**
* Extended the `system.clients` virtual table schema and implementation to include a new `client_options` column (a map of option key/value pairs), and updated the table population logic to use the new cached types and foreign pointers. (`db/virtual_tables.cc`) [[1]](diffhunk://#diff-05f7bff3edb39fb8759c90b445e860189f2f30e04717ed58bae42716082af3d1R752) [[2]](diffhunk://#diff-05f7bff3edb39fb8759c90b445e860189f2f30e04717ed58bae42716082af3d1L769-R770) [[3]](diffhunk://#diff-05f7bff3edb39fb8759c90b445e860189f2f30e04717ed58bae42716082af3d1L809-R816) [[4]](diffhunk://#diff-05f7bff3edb39fb8759c90b445e860189f2f30e04717ed58bae42716082af3d1L828-R879)
**API and interface changes:**
* Changed the signatures of `get_client_data` methods throughout the codebase to return vectors of `foreign_ptr<std::unique_ptr<client_data>>` instead of plain `client_data` objects, to ensure safe cross-shard access. (`alternator/controller.hh`, `alternator/controller.cc`, `alternator/server.hh`, `alternator/server.cc`, `transport/controller.hh`, `transport/protocol_server.hh`) [[1]](diffhunk://#diff-31730ba8e7374f784a88dc27c1512291cf73b7f24e08768f7466a3c8cfcc7a1aL96-R96) [[2]](diffhunk://#diff-19a97c0247cc08155ee49b277e43859ca32d6ef8cbff0ed7368ec5fa19e0a11eL172-R172) [[3]](diffhunk://#diff-5fce246edf5abffb2351bd02e2eb1e9850880f7a00607ccaa90c3eee7ef57c6bL110-R111) [[4]](diffhunk://#diff-a7e2cda866c03a75afcf3b087de1c1dcd2e7aa996214db67f9a11ed6451e596dL988-R995) [[5]](diffhunk://#diff-eea7e2db5d799a25e717a72ac8ce5842bd4adb72b694d38d8f47166d9cd926faL356-R356) [[6]](diffhunk://#diff-d0b4ec3a144bbc5dc993866cf0b940850a457ff6156064f7e2b4b10ad0a95fefL80-R80) [[7]](diffhunk://#diff-4293b94c444d9bd5ecd17ce7eda8c00685d35ecf6e07f844efc91a91bbe85be1L46-R48)
**Testing and validation:**
* Updated the Python test for the `system.clients` table to verify the new `client_options` column and its contents, ensuring that driver name and version are present in the options map. (`test/cqlpy/test_virtual_tables.py`) [[1]](diffhunk://#diff-6dd8bd4a6a82cd642252a29dc70726f89a46ceefb991c3e63fc67e283f323f03R79) [[2]](diffhunk://#diff-6dd8bd4a6a82cd642252a29dc70726f89a46ceefb991c3e63fc67e283f323f03R88-R90)
Closesscylladb/scylladb#25746
* github.com:scylladb/scylladb:
transport/server: declare a new "CLIENT_OPTIONS" option as supported
service/client_state and alternator/server: use cached values for driver_name and driver_version fields
system.clients: add a client_options column
controller: update get_client_data to use foreign_ptr for client_data
Optimize memory usage changing types of driver_name and driver_version be
a reference to a cached value instead of an sstring.
These fields very often have the same value among different connections hence
it makes sense to cache these values and use references to them instead of duplicating
such strings in each connection state.
Signed-off-by: Vlad Zolotarov <vladz@scylladb.com>
This new column is going to contain all OPTIONS sent in the
STARTUP frame of the corresponding CQL session.
The new column has a `frozen<map<text, text>>` type, and
we are also optimizing the amount of required memory for storing
corresponding keys and values by caching them on each shard level.
Signed-off-by: Vlad Zolotarov <vladz@scylladb.com>
Make the removenode operation go through the `left_token_ring` state, similar to decommission. This ensures that when removenode completes, all nodes in the cluster are aware of the topology change through a global token metadata barrier.
Previously, removenode would skip the `left_token_ring` state and go directly from `write_both_read_new` to `left` state. This meant that when the operation completed, some nodes might not yet know about the topology change, potentially causing issues with subsequent data plane requests.
Key changes:
- Both decommission and removenode now transition to `left_token_ring` state in the `write_both_read_new` handler
- In `left_token_ring` state, only decommissioning nodes receive the shutdown RPC (removed nodes are already dead)
- Updated documentation to reflect that both operations use this state
This change improves consistency guarantees for removenode operations by ensuring cluster-wide awareness before completion.
The change is protected by "REMOVENODE_WITH_LEFT_TOKEN_RING" feature flag to also support mixed clusters during e.g. upgrade.
Fixes: scylladb/scylladb#25530
No backport: This fixes and issue found in tests. It can theoretically happen in production too, but wasn't reported in any customer issue, so a backport is not needed.
Closesscylladb/scylladb#26931
* https://github.com/scylladb/scylladb:
topology: make removenode use left_token_ring state for global barrier
topology: allow removing nodes not having tokens
features: add feature flag for removenode via left token ring
When calling a migration notification from the context of a notification
callback, this could lead to a deadlock with unregistering a listener:
A: the parent notification is called. it calls thread_for_each, where it
acquires a read lock on the vector of listeners, and calls the
callback function for each listener while holding the lock.
B: a listener is unregistered. it calls `remove` and tries to acquire a
write lock on the vector of listeners. it waits because the lock is
held.
A: the callback function calls another notification and calls
thread_for_each which tries to acquire the read lock again. but it
waits since there is a waiter.
Currently we have such concrete scenario when creating a table, where
the callback of `before_create_column_family` in the tablet allocator
calls `before_allocate_tablet_map`, and this could deadlock with node
shutdown where we unregister listeners.
Fix this by not acquiring the read lock again in the nested
notification. There is no need because the read lock is already held by
the parent notification while the child notification is running. We add
a function `thread_for_each_nested` that is similar to `thread_for_each`
except it assumes the read lock is already held and doesn't acquire it,
and it should be used for nested notifications instead of
`thread_for_each`.
Fixesscylladb/scylladb#27364Closesscylladb/scylladb#27637
Make the removenode operation go through the `left_token_ring` state,
similar to decommission. This ensures that when removenode completes,
all nodes in the cluster are aware of the topology change through a
global token metadata barrier.
Previously, removenode would skip the `left_token_ring` state and go
directly from `write_both_read_new` to `left` state. This meant that
when the operation completed, some nodes might not yet know about the
topology change, potentially causing issues with subsequent data plane
requests.
Key changes:
- Both decommission and removenode now transition to `left_token_ring`
state in the `write_both_read_new` handler
- In `left_token_ring` state, only decommissioning nodes receive the
shutdown RPC (removed nodes may already be dead)
- Updated documentation to reflect that both operations use this state
This change improves consistency guarantees for removenode operations
by ensuring cluster-wide awareness before completion.
Fixes: scylladb/scylladb#25530
If a keyspace has a numeric replication factor in a DC and rf < #racks,
then the replicas of tablets in this keyspace can be distributed among
all racks in the DC (different for each tablet). With rack list, we need all
tablet replicas to be placed on the same racks. Hence, the conversion
requires tablet co-location.
After this series, the conversion can be done using ALTER KEYSPACE
statement. The statement that does this conversion in any DC is not
allowed to change a rf in any DC. So, if we have dc1 and dc2 with 3 racks
each and a keyspace ks then with a single ALTER KEYSPACE we can do:
- {dc1 : 2} -> {dc1 : [r1, r2]};
- {dc1 : 2, dc2: 2} -> {dc1 : [r1, r2], dc2: [r2,r3]};
- {dc1 : 2, dc2: 2} -> {dc1 : [r1, r2], dc2: 2}
- {dc1 : 2} -> {dc1 : 2, dc2 : [r1]}
But we cannot do:
- {dc1 : 2} -> {dc1 : [r1, r2, r3]};
- {dc1 : 1, dc2 : [r1, r2] → dc1: [r1], dc2: [r1].
In order to do the co-locations rf change request is paused. Tablet
load balancer examines the paused rf change requests and schedules
necessary tablet migrations. During the process of co-location, no other
cross-rack migration is allowed.
Load balancer checks whether any paused rf change request is
ready to be resumed. If so, it puts the request back to global topology
request queue.
While an rf change request for a keyspace is running, any other rf change
of this keyspace will fail.
Fixes: #26398.
New feature, no backport
Closesscylladb/scylladb#27279
* github.com:scylladb/scylladb:
test: add est_rack_list_conversion_with_two_replicas_in_rack
test: test creating tablet_rack_list_colocation_plan
test: add test_numeric_rf_to_rack_list_conversion test
tasks: service: add global_topology_request_virtual_task
cql3: statements: allow altering from numeric rf to rack list
service: topology_coordinator: pause keyspace_rf_change request
service: implement make_rack_list_colocation_plan
service: add tablet_rack_list_colocation_plan
cql3: reject concurrent alter of the same keyspace
test: check paused rf change requests persistence
db: service: add paused_rf_change_requests to system.topology
service: pass topology and system_keyspace to load_balancer ctor
service: tablet_allocator: extract load updates
service: tablet_allocator: extract ensure_node
tasks, system_keyspace: Introduce get_topology_request_entry_opt()
node_ops: Drop get_pending_ids()
node_ops: Drop redundant get_status_helper()
Previously, the scheduling_group column was updated during the switch_tenant function, which meant the update occurred only after the tenant change operation completed—updating rows one by one. With this change, the scheduling_group column is now updated before the switch_tenant logic runs, ensuring that the table reflects the correct scheduling groups for all rows as early as possible.
fixes: #26060fixes: #27295
backport: not required
this is a minor bug fix. Internal logic worked but the user couldnt see the change in the table if they would read the system.clients table
Closesscylladb/scylladb#26404
* github.com:scylladb/scylladb:
test: cqlpy: Remove test_switch_tenants and add test in cluster testing. The test needs to run twice, in two separate Scylla runs, using two different modes: gossip and raft. The cluster framework supports this setup, while cqlpy only runs against Scylla instances in raft mode. Therefore, the test was moved from cqlpy to the cluster-based framework. This commit both adds the test in cluster/ and removes the old version in cqlpy/.
server: Refactor update_control_connection_scheduling_group functionality This refactoring moves the logic that retrieves the scheduling group for driver_service_level_name out of switch_tenant. This change is possible because the scheduling group for the driver is retrieved from a map (LOOKUP). The lookup function is fully synchronized, non-coroutine, and returns immediately. For that reason, it’s better to perform this lookup outside of the switch_tenant function.
server: Refactor scheduling group update functionality. This change generalizes the scheduling-group update functionality and removes some copy-paste code, improving overall readability and maintainability. To achieve this, capturing lambdas were introduced. As a result, self-deducing this was added to those lambdas to avoid coroutine-related issues (“coroutine fiasco”).
server: Fix switch_tenant problem, When running on a V2 server, service-level data comes from service level cache. Because of this, we can use synchronized function to get the schedualing group. Since we are transitioning to a Raft-based architecture where all servers will be V2, we can safely implement this fix specifically for that case. This change adds get_cached_user_scheduling_group functionality and moves its usage out of switch_tenant function in update_scheduling_group_v2 usage.
server: Add update_service_level_scheduling_group_v1 functions to create placehholder for functionality that will introduce v2 implementation. The new functionality will allow usage of service level cache
Add a service::topo::global_topology_request_virtual_task, which
covers the replication factor changes.
Currently, the global_topology_request_virtual_task can be aborted
only if it is paused.
The progress of the rf change isn't counted.
To do the conversion from numeric rf to rack list, we need to co-locate
tablets of the keyspace, so that all of them have replicas on the same racks.
Pause the keyspace_rf_change global topology request, so that the co-location
could be done before the ALTER KEYSPACE changes are applied.
The pause is needed if in any dc rf changes from numeric to rack list
and the co-location is necessary. In this case we don't finish the request.
Instead, we add the request to the paused request vector. No migrations are
started.
The make_rack_list_colocation_plan consists of two phases.
In the first phase (realized with find_required_rack_list_colocations),
we find the pairs of (replica to be co-located, destination dc and rack).
We skip the pairs related to the tablets that are in transition or for
which the load balancer migration is already planned. We group the pairs
by destination dc and rack. Thanks to that in the second phase we can
calculate the least loaded nodes and shards only once for each rack.
In the second phase, we calculate the load of the nodes in a cluster
based on current transition and previously scheduled migrations.
We utilize the map created in the first phase and choose the least
loaded targets in each rack. We skip the tablets for which the
co-location was already scheduled.
find_required_rack_list_colocations isn't a method of load_balancer,
because in the following changes it is going to be reused by topology
coordinator to determine whether the rf change should be paused.
Add tablet_rack_list_colocation_plan. Keep it in migration_plan.
The plan includes a request that is ready to resume. There can be
more than one such request at the time, but we consider them one
by one for clarity of code. Rack list co-locations will be kept together
with normal load balancer migrations.
Consider normal load balancer migrations before rack list co-locations.
During rack list co-location, allow load balancer migrations to happen
only within a single rack. Do not create the merge co-location plan if
there is ongoing rack list co-location (if there are any rf changes paused).
Generate rf change resume based on the plan. Add _request_to_resume back
to global requests queue.
make_rack_list_colocation_plan will be implemented in the following change.
Reject ALTER KEYSPACE request if there is unfinished (queued, pending,
or paused) alter request of the same keyspace.
This is required as in the following changes, global request queue
will contain rf change requests meant to be resumed.
In the following changes, we allow to alter from numeric rf to rack list.
Before the alter, two tablets of the same keyspace can have replicas
on different racks. To switch to rack list, we need to co-locate
the replicas. It will be achieved by pausing the keyspace_rf_change
and scheduling migrations.
We need to persist the ids of requests that are paused. A new column -
paused_rf_change_requests is added to system.topology table.
In this commit no data is kept in the new column.
Pass a pointer to service::topology and db::system_keyspace to load
balancer. It will be used in the following patches to create
rack_list_colocation plan.
`system.client_routes` is a system table that sets the target address and ports for each `host_id`, for one or more connection (e.g., Private Link) represented by `connection_id`. Cloud will write the table via REST, and drivers will read it via CQL to override values obtained from `system.local` and `system.peers`.
This patch series contains:
- Introduction of `CLIENT_ROUTES` feature flag.
- Implementation of raft-based `system.client_routes` table
- Implementation of `v2/client-routes` POST/DELETE/GET endpoints
- Implementation of new `CLIENT_ROUTES_CHANGE` event that is sent to drivers when `system.client_routes` is changed
- New tests that verifies the aforementioned features
Ref: scylladb/scylla-enterprise#5699
For now, no automatic backport. However, the changes are planned to be release on `2025.4` either as a backport or a private build.
Closesscylladb/scylladb#27323
* https://github.com/scylladb/scylladb:
docs: describe CLIENT_ROUTES_CHANGE extension
test: add test for CLIENT_ROUTES event
service: transport: add CLIENT_ROUTES_CHANGE event
test: add cluster tests for client routes
test: add API tests for client_routes endpoints
test: add `timeout` parameter to `delete` in RESTClient
test: allow json_body in send
api: implement client_routes endpoints
api: add client_routes.json
service: main: add client_routes_service
db: add system.client_routes table
gms: add CLIENT_ROUTES feature
Introduce the CLIENT_ROUTES_CHANGE event to let drivers refresh
connections when `system.client_routes` is modified. Some deployments
(e.g., Private Link) require specific address/port mappings that can
change without topology changes and drivers need to adapt promptly
to avoid connectivity issues.
This new EVENT type carries a change indicator plus the affected
`connection_ids` and `host_ids`. The only change value is
`UPDATE_NODES`, meaning one or more client routes were inserted,
updated, or deleted.
Drivers subscribe using the existing events mechanism, so no additional
`cql_protocol_extension` key is required.
Ref: scylladb/scylla-enterprise#5699
Fix multiple cases where the captured `std::exception_ptr` has been re-thrown via simple `throw eptr;`, which results in losing the original exception type and details.
Resolved at various places found by clang-tidy:
1. db::schema_applier
When applying schema changes, the previous implementation attempted to handle exceptions by catching and rethrowing them, but did so incorrectly: using `throw ex` with a `std::exception_ptr` loses the original exception type and details.
However, in this case, explicit exception handling is unnecessary. The only reason for catching was to ensure `ap.destroy()` is called before propagating the exception. This can be more cleanly and safely achieved using Seastar's `.finally()` continuation, which guarantees cleanup regardless of success or failure.
2. directories
The `std::exception_ptr()` has been captured for logging and then again re-thrown incorrectly via `throw ex;`. We could use `std::rethrow_exception()` here instead, but it seems to be simpler to just use regular `throw;` to rethrow the original exception, and only use the `std::current_exception()` for logging (which is a pattern used in other places as well).
3. storage_service
Here the exception has been re-thrown incorrectly in a coroutine. There it is best to use the `co_await coroutine::return_exception_ptr` to propagate exception more efficiently in a coroutine-friendly manner.
Fixes: SCYLLADB-94
Refs: scylladb/scylladb#27501
No backport: This fixes an error logging issue, that isn't a production problem by itself (only found in test), therefore not backporting to older branches.
Closesscylladb/scylladb#27613
* https://github.com/scylladb/scylladb:
db: schema_applier: improve exception-safe cleanup
directories: fix exception rethrowing
storage_service: use coroutine-friendly exception propagation in join_node_response_handler
Introduce `system.client_routes`, a system table that sets the target
address and ports for each `host_id`, for one or more connections
(e.g., Private Link) represented by `connection_id`. Cloud will write
the table via REST, and drivers will read it via CQL to override
values obtained from `system.local` and `system.peers`.
The table is Raft-managed to provide consistent replication across
nodes.
Schema overview: each row is identified by `(connection_id, host_id)`
and describes where clients should connect: `address` and one or more of
`port`, `tls_port`, `alternator_port`, `alternator_https_port`.
`host_id` is a UUID (just as in ScyllaDB) but `connection_id` can be
any string to accept formats of all cloud providers. `address` is
also a regular string because it can represent either an IP address or
a domain. Ports are optional in the sense that at least one of
the four must be provided.
Ref: scylladb/scylla-enterprise#5699
The batchlog table contains an entry for each logged batch that is processed by the local node as coordinator. These entries are typically very short lived, they are inserted when the batch is processed and deleted immediately after the batch is successfully applied.
When a table has `tombstone_gc = {'mode': 'repair'}` enabled, every repair has to flush all hints and batchlogs, so that we can be certain that there is no live data in any of these, older than the last repair. Since batches can contain member queries from any number of tables, the whole batchlog has to be flushed, even if repair-mode tombstone-gc is enabled for a single table.
Flushing the batchlog table happens by doing a batchlog replay. This involves reading the entire content of this table, and attempting to replay+delete any live entries (that are old enough to be replayed). Under normal operating circumstances, 99%+ of the content of the batchlog table is partition tombstones. Because of this, scanning the content of this table has to process thousands to millions of tombstones. This was observed to require up to 20 minutes to finish, causing repairs to slow down to a crawl, as the batchlog-flush has to be repeated at the end of the repair of each token-range.
When trying to address this problem, the first idea was that we should expedite the garbage-collection of these accumulated tombstones. This experiment failed, see https://github.com/scylladb/scylladb/pull/23752. The commitlog proved to be an impossible to bypass barrier, preventing quick garbage-collection of tombstones. So long as a single commit-log segment is alive, holding content from the batchlog table, all tombstones written after are blocked from GC.
The second approach, represented by this PR, is to not rely in tombstone GC to reduce the tombstone amount. Instead restructure the table such that a single higher-order tombstone can be used to shadow and allow for the eviction of the myriads of individual batchlog entry tombstones. This is realized by reorganizing the batchlog table such that individual batches are rows, not partitions.
This new schema is introduced by the new `system.batchlog_v2` table, introduced by this PR:
CREATE TABLE system.batchlog_v2 (
version int,
stage int,
shard int,
written_at timestamp,
id uuid,
data blob,
PRIMARY KEY ((version, stage, shard), written_at, id));
The new schema organization has the following goals:
1) Make post-replay batchlog cleanup possible with a simple range-tombstone. This allows dropping the individual dead batchlog entries, as they are shadowed by a higher level tombstone. This enables dropping tombstones without tombstone GC.
2) To make the above possible, introduce the stage key component: batchlog entries that fail the first replay attempt, are moved to the failed_replay stage, so the initial stage can be cleaned up safely.
3) Spread out the data among Scylla shards, via the batchlog shard column.
4) Make batchlog entries ordered by the batchlog create time (id). This allows for selecting batchlogs to replay, without post-filtering of batchlogs that are too young to be replayed.
Fixes: https://github.com/scylladb/scylladb/issues/23358
This is an improvement, normally not a backport-candidate. We might override this and backport to allow wider use of `tombstone_gc: {'mode': 'repair'}`.
Closesscylladb/scylladb#26671
* github.com:scylladb/scylladb:
db/config: change batchlog_replay_cleanup_after_replays default to 1
test/boost/batchlog_manager_test: add test for batchlog cleanup
replica/mutation_dump: always set position weight for clustering positions
service/storage_proxy: s/batch_replay_throw/storage_proxy_fail_replay_batch/
test/lib: introduce error_injection.hh
utils/error_injection: add debug log to disable() and disable_all()
test/lib/cql_test_env: forward config to batchlog
test/lib/cql_test_env: add batch type to execute_batch()
test/lib/cql_assertions: add with_size(predicate) overload
test/lib/cql_assertions: add source location to fail messages
test/lib/cql_assertions: columns_assertions: add assert_for_columns_of_each_row()
test/lib/cql_assertions: rows_assertions::assert_for_columns_of_row(): add index bound check
test/lib/cql_assertions: columns_assertions: add T* with_typed_column() overload
db/batchlog_manager: config: s/write_timeout/reply_timeot/
db,service: switch to system.batchlog_v2
db/system_keyspace: introduce system.batchlog_v2
service,db: extract generation of batchlog delete mutation
service,db: extract get_batchlog_mutation_for() from storage-proxy
db/batchlog_manager: only consider propagation delay with tombstone-gc=repair
db/batchlog_manager: don't drop entire batch if one mutations' table was dropped
data_dictionary: table: add get_truncation_time()
db/batchlog_manager: batch(): replace map_reduce() with simple loop
db/batchlog_manager: finish coroutinizing replay_all_failed_batches
db/batchlog_manager: improve replayAllFailedBatches logs
Since we are transitioning to a Raft-based architecture where all servers will be V2, we can safely implement this fix specifically for that case.
This change adds get_cached_user_scheduling_group functionality and moves its usage out of switch_tenant function in update_scheduling_group_v2 usage.
Improve exception handling in join_node_response_handler by using
`co_await coroutine::return_exception_ptr` to propagate exceptions.
This replaces the incorrect direct throw of `std::exception_ptr` and
ensures proper coroutine-friendly exception propagation.
This reverts commit faad0167d7. It causes
a regression in
test_two_tablets_concurrent_repair_and_migration_repair_writer_level
in debug mode (with ~5%-10% probability).
Fixes#27510.
Closesscylladb/scylladb#27560
This patch adds tablet repair progress report support so that the user
could use the /task_manager/task_status API to query the progress.
In order to support this, a new system table is introduced to record the
user request related info, i.e, start of the request and end of the
request.
The progress is accurate when tablet split or merge happens in the
middle of the request, since the tokens of the tablet are recorded when
the request is started and when repair of each tablet is finished. The
original tablet repair is considered as finished when the finished
ranges cover the original tablet token ranges.
After this patch, the /task_manager/task_status API will report correct
progress_total and progress_completed.
Fixes#22564Fixes#26896Closesscylladb/scylladb#26924
In the next commit we want to add an optimization that relies on
precise control over the lifetime of cas_request. In particular, we
want the implementation of this interface in Alternator to operate on
raw references that are guaranteed to remain valid only until the
cas() future is resolved. We already depend on the same lifetime
assumptions in cas_request when used by modification_statement.
However, these assumptions are not clearly expressed in the current
interface: cas_request is taken by shared_ptr, and nothing prevents
cas() from storing that pointer inside paxos_response_handler, which
may outlive the cas() future.
This commit fixes that by taking cas_request by raw reference. This
makes it explicit that cas() does not assume ownership of the object.
Callers must ensure that the referenced object remains valid until
the returned future is resolved.
We saw that in large clusters direct failure detector may cause large task queues to be accumulated. The series address this issue and also moves the code into the correct scheduling group.
Fixes https://github.com/scylladb/scylladb/issues/27142
Backport to all version where 60f1053087 was backported to since it should improve performance in large clusters.
Closesscylladb/scylladb#27387
* github.com:scylladb/scylladb:
direct_failure_detector: run direct failure detector in the gossiper scheduling group
raft: drop invoke_on from the pinger verb handler
direct_failure_detector: pass timeout to direct_fd_ping verb
The PRUNE MATERALIZED VIEW statement is performed as follows:
1. Perform a range scan of the view table from the view replicas based
on the ranges specified in the statement.
2. While reading the paged scan above, for each view row perform a read
from all base replicas at the corresponding primary key. If a discrepancy
is detected, delete the row in the view table.
When reading multiple rows, this is very slow because for each view row
we need to performe a single row query on multiple replicas.
In this patch we add an option to speed this up by performing many of the
single base row reads concurrently, at the concurrency specified in the
USING CONCURRENCY clause.
Aside from the unit test, I checked manually on a 3-node cluster with 10M rows, using vnodes. There were actually no ghost rows in the test, but we still had to iterate over all view rows and read the corresponding base rows. And actual ghost rows, if there are any, should be a tiny fraction of all rows. I compared concurrencies 1,2,10,100 and the results were:
* Pruning with concurrency 1 took total 1416 seconds
* Pruning with concurrency 2 took total 731 seconds
* Pruning with concurrency 10 took total 234 seconds
* Pruning with concurrency 100 took total 171 seconds
So after a concurrency of 10 or so we're hitting diminishing returns (at least in this setup). At that point we may be no longer bottlenecked by the reads, but by CPU on the shard that's handling the PRUNE
Fixes https://github.com/scylladb/scylladb/issues/27070Closesscylladb/scylladb#27097
* github.com:scylladb/scylladb:
mv: allow setting concurrency in PRUNE MATERIALIZED VIEW
cql: add CONCURRENCY to the USING clause
When direct failure detector was introduces the idea was that it will
run on the same connection raft group0 verbs are running, but in
60f1053087 raft verbs were moved to run on the gossiper connection
while DIRECT_FD_PING was left where it was. This patch move it to
gossiper connection as well and fix the pinger code to run in gossiper
scheduling group.
Currently raft direct pinger verb jumps to shard 0 to check if group0 is
alive before replying. The verb runs relatively often, so it is not very
efficient. The patch distributes group0 liveness information (as it
changes) to all shard instead, so that the handler itself does not need
to jump to shard 0.
Consider the following scenario:
1. A table has RF=3 and writes use CL=QUORUM
2. One node is down
3. There is a pending tablet migration from the unavailable node
that is reverted
During the revert, there can be a time window where the pending replica
being cleaned up still accepts writes. This leads to write failures,
as only two nodes (out of four) are able to acknowledge writes.
This patch fixes the issue by adding a barrier to the cleanup_target
tablet transition state, ensuring that the coordinator switches back to
the previous replica set before cleanup is triggered.
Fixes https://github.com/scylladb/scylladb/issues/26512
It's a pre existing issue. Backport is required to all recent 2025.x versions.
Closesscylladb/scylladb#27413
* github.com:scylladb/scylladb:
topology_coordinator: Fix the indentation for the cleanup_target case
topology_coordinator: Add barrier to cleanup_target
test_node_failure_during_tablet_migration: Increase RF from 2 to 3
Consider the following scenario:
1. A table has RF=3 and writes use CL=QUORUM
2. One node is down
3. There is a pending tablet migration from the unavailable node
that is reverted
During the revert, there can be a time window where the pending replica
being cleaned up still accepts writes. This leads to write failures,
as only two nodes (out of four) are able to acknowledge writes.
This patch fixes the issue by adding a barrier to the cleanup_target
tablet transition state, ensuring that the coordinator switches back to
the previous replica set before cleanup is triggered.
Fixes https://github.com/scylladb/scylladb/issues/26512
Currently direct_fd_ping runs without timeout, but the verb is not
waited forever, the wait is canceled after a timeout, this timeout
simply is not passed to the rpc. It may create a situation where the
rpc callback can runs on a destination but it is no longer waited on.
Change the code to pass timeout to rpc as well and return earlier from
the rpc handler if the timeout is reached by the time the callback is
called. This is backwards compatible since timeout is passed as
optional.
Now that batchlog cleanup is cheap, on account of memtable flush on the
system.batchlog table garbage-collecting tombstones (previous patch), we
can afford to do cleanup on each replay, keeping the memtable size small
and more importantly -- the amount of tombstones in the memtable small.
Rename to make it more explicit where the error injection happens.
Also change how the error is injected, use the lambda overload instead
of is_enabled(), the former leaves better trace in logs, which helps
when debugging tests.
New batchlogs are written to the batchlog_v2 table and replay also uses
the v2 table.
The content of system.batchlog is attempted to be migrated to
system.batchlog_v2 after each start of the batchlog_manager service.
The migration is retried on each replay if it fails. This is reduntant
but simple.
Batchlog cleanup now doesn't involve flushing memtables, the only
remaining user of replica/database.hh is gone, so the include is
dropped.
Don't build batchlog delete mutations in storage-proxy code. Move this
code into db/batchlog_manager.cc, exposed via db/batchlog.hh.
This serves multiple goals:
1) Concentrates low-level batchlog related logic in
db/batchlog_manager.cc
2) Reduce current and future code duplication.
3) Make future changes to this logic easier.
Don't build batchlog mutations in storage-proxy code. Move this code
into db/batchlog_manager.cc, exposed via db/batchlog.hh.
This serves multiple goals:
1) Concentrates low-level batchlog related logic in
db/batchlog_manager.cc
2) Reduce current and future code duplication.
2) Make future changes to this logic easier.
Consider this:
1) n1 is the topology coordinator
2) n1 schedules and executes a tablet repair with session id s1 for a
tablet on n3 an n4.
3) n3 and n4 take and store the in _rs._repair_compaction_locks[s1]
4) n1 steps down before it executes
locator::tablet_transition_stage::end_repair
5) n2 becomes the new topology coordinator
6) n2 runs locator::tablet_transition_stage::repair again
7) n3 and n4 try to take the lock again and hangs since the lock is
already taken.
To avoid the deadlock, we can throw in step 7 so that n2 will
proceed to end_repair stage and release the lock. After that, the
scheduler could schedule the tablet repair request again.
Fixes#26346Closesscylladb/scylladb#27163
In several exception handlers, only raft::request_aborted was being
caught and rethrown, while seastar::abort_requested_exception was
falling through to the generic catch(...) block. This caused the
exception to be incorrectly treated as a failure that triggers
rollback, instead of being recognized as an abort signal.
For example, during tablet draining, the error log showed:
"tablets draining failed with seastar::abort_requested_exception
(abort requested). Aborting the topology operation"
This change adds seastar::abort_requested_exception handling
alongside raft::request_aborted in all places where it was missing.
When rethrown, these exceptions propagate up to the main run() loop
where handle_topology_coordinator_error() recognizes them as normal
abort signals and allows the coordinator to exit gracefully without
triggering unnecessary rollback operations.
Fixes: scylladb/scylladb#27255
No backport: The problem was only seen in tests and not reported in
customer tickets, so it's enough to fix it in the main branch.
Closesscylladb/scylladb#27314
