Known Issues
5279 tracked MongoDB server issues from Jira with status, affected versions, and resolution details.
SERVER-138894ClosedWon't Fix
Stop decompressing columns if prior column evaluation eliminates all documents
SERVER-117906Backlog
$idLookup could be responsible for apply limit to DistributedPlanLogic
When a stage desugars with an $idLookup, it can choose to pass a limit value to be applied to the document stream. That limit only applies per-node though, so a sharded $idLookup could return N * limit documents overall, where N is the number of data-bearing nodes involved in the query.
Today, $vectorSearch and $search (which desugar with an $idLookup node) compensate by being responsible themselves for adding a $limit to the DistributedPlanLogic merging pipeline. We could consider having $id
SERVER-110278Backlog
Modify Split Pipeline logic to allow pushdown of sharding stages to shards pipeline after a split point
SERVER-110277 Introduced some initial changes to the split pipeline logic which provide a special case for splitting when we have a source stage followed by $_internalSearchIdLookup.
With the above change, we should now be splitting a pipeline at the $_internalSearchIdLookup stage (running on shards).
We’ll also need to modify existing methods in SplitPipeline, which allow for pushing down stages which can run on shards after the split point has been determined (i.e $match, $project, etc.
SERVER-134721Open
Grafana dashboard updates for query availability
Including write conflict storm mitigations, maxEstimatedScan, query memory load shedding, idle cursor memory consumption, etc.
Add this under the [mongotune dashboards|https://grafana.helix.corp.mongodb.com/dashboards/f/feyy5iih33ncwa/mongotune] group.
Some non-obvious things to cover:
* For QMLS, add a dashboard panel that flags clusters where QMLS is disabled but the low mark isn't -1.
SERVER-130333ClosedFixed
Probabilistic query memory load shedding
See the [README|https://github.com/mongodb/mongo/blob/84f627f05fd7803e586b169331c17deade82c550/src/mongo/db/memory_tracking/README.query_memory_load_shedding.md]
Fixed Versions
9.1.0-rc09.0.0-rc2
SERVER-132056ClosedDone
Investigate incremental spilling in primary-driven index builds
During index builds, the sorter can accumulate significant amount of index keys in memory (default 200MB) before spilling, resulting in a workload where we have large bursts of writes that can be fairly spaced out. This can interact poorly with flow control, which can kick in due to standby replication lag.
We should investigate if we can smooth out this workload with some form of incremental spilling, with the goal of getting better performance by avoiding flow control kicking in due to sta
SERVER-138165ClosedFixed
Add hash coverage to existing validation checks and metric tracking
After SERVER-137708 and SERVER-136298 land, extend the hash validation tests in size_count_checkpoint_validator_test.cpp and checkpoint_validation_steady_state.js to assert watermarksRecorded, watermarksEvicted and checkpointsSkipped, as SERVER-137708 does for the size and count tests.
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-83091ClosedFixed
$or query can trigger an infinite loop during plan enumeration
Certain $or queries can trigger an infinite loop in the {{PlanEnumerator}}. While the plan space can be quite large for indexed $or queries, we have logic which is supposed to bail out of plan enumeration once a certain number of alternatives have been generated. This logic is not kicking in correctly.
Note that this only affects the so-called "lockstep $or" enumeration order which was first implemented in SERVER-50291. Lockstep $or enumeration is intended to make it more likely that the syst
Affected Versions
7.0.27.0.37.2.0-rc1
Fixed Versions
7.3.0-rc07.2.0-rc27.0.56.0.135.0.24
SERVER-138857ClosedFixed
Split db_integration_test into two
{{db_integration_test}} runs two completely unrelated tests: one tests checks on a BSON nesting depth limit and the other checks how {{$currentOp}} handles exhaust cursors. In preparation for some infinite burndown-related test task splitting, this change breaks this integration test binary into two cohesive ones.
Fixed Versions
9.1.0-rc09.1.0-rc1025
CSHARP-6271In Code Review
Mandate raising writeConcernError (change SHOULD to MUST)
This ticket was split from {{DRIVERS-3684}}, please see that ticket for a detailed description.
Components
CRUD & Command-Level Operations
SERVER-138787ClosedFixed
Emit endOfTransaction for prepared replica-set and single-shard transaction
Change streams synthesized an {{endOfTransaction}} event only for unprepared multi-document transactions, whose commit is the final {{applyOps}} entry.
A prepared replica-set/single-shard transaction commits through a standalone {{commitTransaction}} oplog entry instead, so it currently produces no event.
This change teaches the replica-set/single-shard unwind stage that commit shape and records on the commit entry that the write path for end-of-transaction was enabled at transaction commit time
Components
Change streams
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-138104In Progress
Ignore oplog entries carrying a document hash with no size delta
h2. Background
Size metadata ({{{}m{}}}) on CRUD oplog entries can include {{sz}} (replicated fast count size delta), {{h}} (continuous internode validation per-document hash), or both. The primary can emit an entry with {{h}} and no {{{}sz{}}}: {{setSizeMetadata}} in {{op_observer_impl.cpp}} attaches {{m}} whenever either field is set. However, we haven't clearly defined whether that case is valid. Each consumer handles it on its own, and the code and comments disagree with each other.
h2.
SERVER-136702Open
Make a test command for a collection being eligible collection hash having an absent hash
GODRIVER-4253Ready for Work
Move CSOT unified tests into the Go driver repository
h3. *Context*
{color:#505f79}Since the spec is being deprecated and removed, we should move the unified tests into our codebase in order to avoid losing coverage. This will also allow us to decouple these tests more easily from the unified runner by converting them to bespoke Go driver prose tests as needed.{color}
h3. *Definition of done*
* {color:#505f79}Copy all current CSOT unified spec tests from their respective spec test submodule directories into a directory inside the go driver re
Components
CSOTTesting
SERVER-133623Open
Investigate impact of ingress rate limiter on large FLE2 writes in a sharded cluster