Known Issues
5279 tracked MongoDB server issues from Jira with status, affected versions, and resolution details.
SERVER-139044In Code Review
Create a second version of the materialized tracker
Code in the {{v2}} namespace will be the scratchpad for working on improvements to the materialized tracker. We start with an exact copy of the original tracker, including all of its unit tests.
We add a feature flag so that we are able to switch back to the old tracker if needed, but development will continue in the v2 tracker.
SERVER-131625ClosedFixed
Add Ubuntu 26.04 support for Server.
It looks like we want 26.04 to land in the 9.0 release after all, since the kernel has already been patched to work with mongod.
Fixed Versions
9.1.0-rc09.1.0-rc10249.0.4
SERVER-138742In Code Review
Run SBE blueprint dead slot elimination in the stage builder behind featureFlagSbeBlueprintRewrites
Run the dead slot elimination rule over the SBE blueprint in {{SlotBasedStageBuilder::build()}} when {{featureFlagSbeBlueprintRewrites}} is on.
* Debug builds tassert if the rewrite does not converge or leaves a stale slot summary; release builds log a warning on non-convergence
* Test that a sort's parallel arrays check survives the rewrite
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-139023In Code Review
Validate that any Fuzzer violations are invalid BSON.
Run Validate BSON on any values that cause an error to ensure that we don't have bugs in the product or fuzzing test code.
SERVER-128021ClosedFixed
Invariant defanging documentation changes
This ticket is to update all types of documentation API Guide,
Programming Guide, and Architecture Guide related to the project.
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-136754In Progress
Yielding support for SBE batched index scans and BatchedFetchStage
SERVER-138948Needs Scheduling
mongodb contains a vulnerability in the zlib code (CVE-2026-85091)
Our team came up with a backport to fix this vulnerability. See PR here:
https://github.com/mongodb/mongo/pull/1834
SERVER-137894ClosedFixed
Create ConsumesAllOrderInsensitive stage constraint
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-136509In Code Review
Remove / replace WiredTigerMetrics with the new WiredTiger otel stat generation
WiredTigerMetrics is hard coding otel metrics we want it gone
SERVER-138112ClosedFixed
Skip the second voteAbortIndexBuild after the index build attempt limit's abort is accepted
When the primary accepts the attempt limit abort from _signalPrimaryForAbortIfOverAttemptLimit, the build fails, and the normal failure path (_signalPrimaryForAbortAndWaitForExternalAbort) sends voteAbortIndexBuild again. It's harmless, since the primary has already aborted, but it's a wasted round trip and a confusing log line. The failure path should skip the vote if the abort was already accepted.
Components
Index Builds
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-138268Open
Fix missing backup mock setup in PALI chaos tests
PALI chaos tests mock backup service is not set up to answer requests about the tests log which causes S3 reads to fail after a PS crash.
Add the mock response already used by other S3 tests.
Follow up to BF-46485.
SERVER-138123ClosedFixed
Add the shared in-scope predicate for set-wide Replicated RecordIds
Add the single shared predicate for "this collection replicates rids under the set-wide feature": explicitly replicated AND non-clustered, excluding [implicitly replicated namespaces|https://github.com/10gen/mongo/blob/5fce3b99dea39b4f9453477819cedd935fba9a2c/src/mongo/db/namespace_string.cpp#L368]: {{config.transactions}}, {{config.image_collection}}, change stream pre-images).
h2. Details
- One function used by every consumer: op observer rid emission, applier rid expectations, the init
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-120246Needs Scheduling
Optimize calculation of cardinality, frequency and monotonicity metrics in analyzeShardKey command
SERVER-135859In Code Review
Test Kudzu workload for large concurrent writes during primary driven index build
SERVER-131487Backlog
Return array of responses in sharded createSearchIndexes to expose per-shard index IDs
Sharded createSearchIndexes does not return per-shard index IDs in its response. This prevents by-ID search index operations (update, drop, list) from being tested in e2e suites against a real mongot.
by-ID tests in jstests/with_mongot/search_mocked/text_search_index_commands.js remain mocked because index IDs are not reliably available after creating a search index on a sharded collection. See jstests/libs/query_integration_search/search.js ln#224 for the current helper limitation.
Came u
SERVER-114028Backlog
Generalize LiteParserRegistration to allow for all FeatureFlag types, not just IFR flags
See SERVER-137483 for supporting binary-compatible feature flags. The biggest obstacle to supporting the last class of feature flags, FCV-gated flags, is that the FCV may not be initialized yet at load time, and it can change at runtime. Thus, there must be a way to verify the FCV at fallback parser registration time and correctly check whether the flag was enabled.
SERVER-137703ClosedFixed
Fix race condition in ConnectionPool CRL-rejection resilience test
In this test, we assert that {{rejected > baseline}} in terms of rejected connections [here|https://github.com/10gen/mongo/blob/7f1db302f54d67ef62b2eee8c1bee4b25b897368/src/mongo/executor/egress_connection_pool_rate_limiter_resilience_integration_test.cpp#L450-L456] then disable the failpoint, then assert all responses failed. We should assert all responses failed before disabling the failpoint.
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-133791Open
Split the first/last send in log entry lifecycle
The current log lifecycle metrics recognise 4 stages:
Generate - Send - Commit - Materialize
but "send" is a bit subtle. Up to the first send is PALI-dominated queuing, but between the first send and the last send we are retrying on account of log service interruptions which isn't really pali controlled.
The way the current code works "re-send" has different effects for oplog vs phylog:
- for oplog, entirely new entries are generated for re-send, so even the current code does not capt
SERVER-135204Open
Prioritize oplog traffic over phylog writes in PALI
(Previous this was a ticket in SLS-11996, as this is a PALI request I am create a SERVER ticket here. See some initial discussion as comments on SLS-11996)
Coming as part of investigation done in [https://jira.mongodb.org/browse/SLS-11818]
Phylog/checkpoint write bursts can reach ~691 MiB/s and coincide with increased peer-message delivery latency and elevated p99 commit latency. Lower bursts around 430–462 MB/s do not show the same impact.
Investigate changes in PALI/log-service sched
SERVER-134926Open
Replace no_timestamp declarations on disaggregated writes with real commit timestamps
SERVER-134925 declared MongoDB's untimestamped writes to disaggregated (layered)
tables with {{allowOneUntimestampedWrite()}} so WT's commit-timestamp check
(WT-18585 / WT-18608) can be enabled. That suppresses the error, not the
problem: an untimestamped commit is never picked up by the ingest->stable drain,
so the content stays in the local ingest table and is lost when ingest is
discarded.
For each site marked {{TODO SERVER-134926}}, determine whether a real commit
timestamp can be supplied
Components
WiredTiger