Known Issues
5279 tracked MongoDB server issues from Jira with status, affected versions, and resolution details.
SERVER-135336ClosedDeclined
prototype independent mat frontier updates using existing log service apis
As part of the materialization lag improvement work, we need to decouple materialization frontier updates from Mongo-side ReadLog processing.
Currently, if ReadLog ingestion or processing becomes slow, PALI may continue observing a stale materialization frontier even though Log service has already progressed. The long-term solution is a dedicated log service frontier-only metadata stream, but that API is not currently available.
Implement a temp playground path using existing lo
SERVER-135319ClosedFixed
Add a dedicated LogFrontiers stream for standby materialization
The standby updated its materialized log position from ReadLog responses. When ReadLog processing stopped, the position stopped advancing, so checkpoints could not be installed and WT could not evict unmaterialized pages. This adds a second stream from the log service that carries the same positions independent of whether the ReadLog stream is stalled, to avoid these secondary effects.
Enabled by default under the {{logServerFrontierStreamEnabled}} server parameter, but it can be disabled for
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-137420ClosedDone
Annotate $group key fields that cannot be arrays
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-48526ClosedFixed
Lengthy oplog scans may cause difficulty resuming change streams in a sharded cluster
If a change stream is resumed in a sharded cluster, all shards that have no oplog entries for the relevant namespace (either because they never had any data for the collection, or because they have not taken any writes recently) will scan through their oplogs until reaching EOF before returning their first {{getMore}} batch to {{mongoS}}. If the oplogs are large and the requested resume point is early, then this may take so long that the cursors on shards which *do* have oplog entries time out,
Components
Aggregation FrameworkChange streams
Fixed Versions
9.0.0-rc0
SERVER-126330ClosedFixed
Rename "accurate" size and count in WiredTigerRecordStore
In SERVER-125800, we decided to use "accurate" as the qualifier to distinguish the new size/count stored in the {{WiredTigerRecordStore}} from the legacy, {{SizeInfo::numRecords}} and {{{}SizeInfo::dataSize{}}}. Accurate might be overselling our size/count, though, so we should choose a qualifier that better reflects the new size/count semantics.
As part of this change or in a follow-up, we should use the new size/count in {{WiredTigerRecordStore::numRecords()}} and {{{}WiredTigerRecordStore:
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-138339Needs Scheduling
Add BSONObj/BSONElement spatial safety fuzz testing
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-137718Open
max_time_ms_mongos_deadline.js counts config server connections as shard connection churn
h2. Problem
{{jstests/noPassthrough/network/max_time_ms_mongos_deadline.js}} asserts that mongos creates no new egress connections while a find hits MaxTimeMSExpired on the shard. Its helper {{getShardHostPoolStats()}} sums {{connPoolStats.hosts}} over every host, including the config server replica set members.
Shortly after startup, mongos issues config reads with nearest read preference. When such a read lands on a config node that has no pooled connection yet, mongos opens a new connec
SERVER-138299Needs Scheduling
Add bounds checking to BSONObj constructors
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-138292Needs Scheduling
Add type check to public BSONElement accessors
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-138257Needs Scheduling
Add bounds checks to BSONObj iteration
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-138359Needs Scheduling
Investigate libunwind in Fuzzer Test
Is libunwind causing fuzzers to take longer. Should libunwind be removed from the BSON Overread Fuzzer or Fuzzers in general.
SERVER-138307In Code Review
Add write path hooks to register update and delete query stats for bulkWrite
_Why:_ Registration must happen after parsing and before canonicalization, which is inside the shared update and delete write path.
_What:_ Add an opt-in hook in the shared update and delete write path that registers bulkWrite ops. Other callers are unaffected.
SERVER-136295ClosedFixed
Implement checkpoint adoption in continuous internode validation
The [technical design|https://docs.google.com/document/d/11UuywdF2CZctqzVqHdZ9dNcLDt1kaKf52ykgrqUBd5w/edit?tab=t.rasi4c5l19ke#heading=h.5llnuy105z27] mentions handling recovery by adopting the first checkpoint from the primary without validation after a standby restarts.
The {{SizeCountCheckpointValidator}} must handle this by initializing {{_baseline}} upon seeing the first checkpoint.
This behavior is also the reason we can see a checkpoint with no watermark:
* ts_0: watermark
* ts_1
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-138343Needs Scheduling
Complete TODO listed in SERVER-114203
There is a TODO in code owned by your team which references a resolved ticket.
The TODO appears in:
* src/mongo/s/commands/query_cmd/cluster_find_and_modify_cmd.cpp:949
You can also search for all references to the resolved ticket here:
https://github.com/10gen/mongo/search?q=SERVER-114203&type=Code
The next steps for this ticket are to either remove the outdated TODO or follow the steps in the TODO if it is correct. If the latter, please update the summary and description of this ticket to r
SERVER-114203ClosedDuplicate
FindAndModify should only implicitly create database if upsert is true
According to [this comment|https://github.com/10gen/mongo/blob/65d23b6e5e6f0141e9fb0f797d7fb2df12969d59/src/mongo/s/commands/query_cmd/cluster_find_and_modify_cmd.cpp#L1027-L1028], findAndModify should implicictly create the database only when upsert is true, but this is not the case right now.
The goal of this ticket is to avoid creating the database implicitly when upsert is not true.
SERVER-137314Open
Perf: Investigate caching size
Cache the size in the 3 bytes if it fits. Else recalculate. Requires an unlikely branch.
SERVER-137313Open
Perf: Investigate caching type
Investigate caching the type in the BSONElement.
SERVER-137309ClosedFix Not Needed
Validate size in Untrusted BSONElement Constructor
Validate the data type, fieldname, and value size fit in the buffer during untrusted construction.