Known Issues
5279 tracked MongoDB server issues from Jira with status, affected versions, and resolution details.
SERVER-139187In Progress
Fix ECs measurement from raw acquisitions to a better measurement of yield count
SERVER-138398In Code Review
In-process-derived stack traces are mis-symbolized due to BOLT rewriting
Stack traces all in crashes, asserts, and the heap profiler are mis-symbolized due to BOLT. Frames will resolve to unrelated data symbols and wrong function names. The bug is that both abseil and libunwind assume a load relocation for program counters that is now incorrect due to BOLT adding a second {{LOAD}} segment. Both abseil and libunwind assume that the first load segment is the only load segment, and look up symbol names relative to that segment.
The impact is that emitted stack traces
Fixed Versions
9.0 Required
SERVER-133311ClosedDuplicate
Should a replicated ident drop be fatal when a reader still holds the ident?
SERVER-132623 fixed one holder {{AutoGetOplogFastPath}} was retaining a whole {{CollectionCatalog}} snapshot and pinning every ident in it. However, Vishnu brought up the point about handling the replicated ident drop when a user read that started just before the drop (PIT lock-free reads).
SERVER-138781Open
Run sys-perf for the incremental sorter spilling stack before merge
The stack makes container sorters spill incrementally unconditionally, so there is nothing to enable by default and no kill switch: the fallback is a revert.
The index-build write path is hot, so before any merge the stack is measured with a sys-perf run comparing a base build against a patched build (no off/on arms): flow-control engagement and standby lag should improve as in SERVER-132056's measurements, sorter memory should stay within the budget, and the foreground p50 change should be re-
SERVER-139235In Progress
Add incremental_sorter_spilling.js end-to-end jstest
The container spiller always spills incrementally: SERVER-138779's two-buffer machinery is always on, decided in the SERVER-132056 review, so there is no parameter to add and no wiring in {{makeSpiller}}.
This adds {{incremental_sorter_spilling.js}} to the {{no_passthrough_primary_driven_index_builds}} matrix suite. The collection is sized so the scan spills many times against the minimum sorter memory budget, and the insertion batch size is small enough that each spill is emitted in many batch
GODRIVER-4172Backlog
Maintain the error chain when returning a mongo.WriteException
h3. *Context*
The [{{processWriteError}}|https://github.com/mongodb/mongo-go-driver/blob/3f2f028ee04bcb9e287ab4e23181873c540ec736/mongo/errors.go#L907] func converts all errors chains that contain type {{driver.WriteCommandError}} into a {{mongo.WriteException}}, but discards the error chain. As a result, the additional context from all wrapped and joined errors is completely missing from the returned error. That means that joined retry errors (added with GODRIVER-3600) are never returned from
Components
Error Handling
SERVER-139299In Code Review
Remove BSONElementValue from src/mongo/db/exec/sbe/values/bsoncolumn_materializer.h
Since SP team wants to delete BSONElementValue another stepping stone is removing it from src/mongo/db/exec/sbe/values/bsoncolumn_materializer.h
SERVER-134745ClosedFixed
Sharded local change-stream updateLookup fails getMore with NamespaceNotFound when the database was dropped, instead of a null fullDocument
h3. Summary
On a sharded cluster, a cluster-wide change stream with {{fullDocument: "updateLookup"}} aborts a {{getMore}} with {{NamespaceNotFound}} (26) when an {{update}}/{{replace}} event's database has been dropped before the stream reads that event:
The expected result is a null {{fullDocument}}. This is explicitly the intended behavior per SERVER-129059's own acceptance criteria, which state: "keep {{NamespaceNotFound}} and {{CollectionUUIDMismatch}} mapping to a null fullDocument".
T
Affected Versions
9.0.2
Fixed Versions
9.1.0-rc09.1.0-rc10249.0.4
SERVER-138883In Code Review
Add facility for labelling a chunk as "irregularly sized"
Create a new command to take proper chunkOp lock and mark chunk as "irregularly sized". To keep in simple in the first iteration, let's clear this flag after a merge/split operation. This is to try to avoid unnecessarily "spreading" the "irregularly sized" label to the entire chunk range.
SERVER-136803ClosedDone
Chain validation across ORP and LS entries
Each applied LSN must contiguously chain from the previous one regardless of which source served it. Right now the chain gap check (12908000) only ever sees entries from the ReadLog stream
Make sure entries applied from the ORP read stream pass the same contiguity check, and that the first LS entry after a cutback is checked against the last ORP applied LSN (and the same when we go back to ORP). Any gap here should crash like currently. The ORP start LSN is exclusive so we shouldn't see dupli
SERVER-130498ClosedFixed
Strictly validate the names of the FLE2 ESC and ECOC state collections
h1. Security Context
See SECBUG-1031 for problem description, security impact, preconditions, reproduction steps, and severity rationale.
h1. Root Cause
Authorization is checked only on the user-facing EDC namespace; the internal state collection names supplied by the client in encryptionInformation are never validated against the EDC collection or authorized separately.
+*Gap 1 — isFLE2StateCollection validates format only, not derivation*+
src/mongo/crypto/encryption_fields_valida
Components
Queryable Encryption
Fixed Versions
9.1.0-rc09.1.0-rc10258.3.169.0.78.0.377.0.48
SERVER-139270In Code Review
Add per-batch internode validation to POA's advancer
Add per-worker validator buffers and watermark/checkpoint bookkeeping directly to {{InflightBatch}}, alongside its existing multikey metadata. Give each work item a pointer to its worker's buffer for that batch, pass it into the apply function, and scope delta recording to the item so persistent workers cannot mix results from different batches. Keep those pointers valid as the batch moves through the tracker and until the advancer consumes its buffers. Evaluate validation enablement once per ba
NODE-7850ClosedDone
Run unit tests against the bundled barrel in CI
h3. Use Case
As a... Node Driver developer
I want... bundled unit tests to run during CI
So that... bundled unit tests failures block a release
h3. User Experience
* bundled unit tests start to run as part of CI
* bundled unit tests currently do not part run as part of any of our CI runs
h3. Dependencies
* none
h3. Risks/Unknowns
* none
* I tested this with an Evergreen patch and all tests still pass: https://spruce.corp.mongodb.com/version/6aaad3ed424df3000722dc1b/tasks?so
Components
CITesting
Fixed Versions
7.9.0
SERVER-139282ClosedFixed
Mark QSN move constructors as noexcept
This was flagged on the SERVER-136115 pr but was missed.
https://github.com/10gen/mongo/blob/master/docs/exception_architecture.md#using-noexcept
Fixed Versions
9.1.0-rc09.1.0-rc1026
SERVER-100669Needs Scheduling
Remove version checks in transactions_wait_for_write_concern.js and speculative_read_transaction.js when v9.0 becomes last-lts
SERVER-125009Needs Scheduling
mongorestore from config can run into error due to validation
to get around this and make restore complete, I had to pass --nsExclude "config.settings"
SERVER-113271Needs Scheduling
Complete TODO listed in SERVER-89498
There is a TODO in the codebase referencing a resolved ticket which is assigned to you.
Please follow this link to see the lines of code referencing this resolved ticket:
https://github.com/10gen/mongo/search?q=SERVER-89498&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 represent the work you're actually doing.
SERVER-107144Needs Scheduling
Complete TODO listed in SERVER-67835
There is a TODO in the codebase referencing a resolved ticket which is assigned to you.
Please follow this link to see the lines of code referencing this resolved ticket:
https://github.com/10gen/mongo/search?q=SERVER-67835&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 represent the work you're actually doing.
SERVER-132881Needs Scheduling
Create regression test for BF-44603
BF-44603 was fixed, but we should have a regression test on master to ensure the issue can't appear again.
Specifically create test to ensure the following scenario does not happen:
* The resharding DDL coordinator fails over.
* It re-sends `_configsvrReshardCollection` to start or join an existing resharding coordinator on the config server.
* A concurrent FCV transition is happening. So the command early returns with an "fcv transition in progress" error, preventing it from joining an
SERVER-138508Backlog
Change error code for multiple reshardings to reflect maximum reshardings reached
Right now, the number of concurrent resharding operations allowed is capped (SERVER-136621), and when the max is reached, we return the error code {{ConflictingOperationInProgress}} which isn't reflective of the actual error (it is a relic of invoking a second resharding operation at all while one is ongoing). We should change it to something like {{ReshardingConcurrencyLimitReached}} or {{MaximumReshardingOperationsExceeded}} to be more reflective.
There is also a test plan for the multiple