Known Issues
5279 tracked MongoDB server issues from Jira with status, affected versions, and resolution details.
PYTHON-6249ClosedDone
Enable ruff D rules
h3. *Context*
{color:#505f79}Describe the background behind the problem.{color}
Ruff 0.16.0 added the {{D}} category of rules derived from the pydocstyle linter.
h3. *Definition of done*
{color:#505f79}What must be done to consider the task complete?{color}
Remove the suppression of {{D}} rules or add an explicit comment explaining why keeping it suppressed is correct.
h3. *Pitfalls*
{color:#505f79}What should the implementer watch out for? What are the risks?{color}
PYTHON-6246ClosedDone
Enable ruff FA rules
h3. *Context*
{color:#505f79}Describe the background behind the problem.{color}
Ruff 0.16.0 added the {{FA}} category of rules derived from the flake8-future-annotations linter.
h3. *Definition of done*
{color:#505f79}What must be done to consider the task complete?{color}
Remove the suppression of {{FA}} rules or add an explicit comment explaining why keeping it suppressed is correct.
h3. *Pitfalls*
{color:#505f79}What should the implementer watch out for? What are the risks?{color}
PYTHON-6243ClosedDone
Enable ruff TC rules
h3. *Context*
{color:#505f79}Describe the background behind the problem.{color}
Ruff 0.16.0 added the {{TC}} category of rules derived from the flake8-type-checking linter.
h3. *Definition of done*
{color:#505f79}What must be done to consider the task complete?{color}
Remove the suppression of {{TC}} rules or add an explicit comment explaining why keeping it suppressed is correct.
h3. *Pitfalls*
{color:#505f79}What should the implementer watch out for? What are the risks?{color}
PYTHON-6239ClosedDone
Enable ruff FURB rules
h3. *Context*
{color:#505f79}Describe the background behind the problem.{color}
Ruff 0.16.0 added the {{FURB}} category of rules derived from the refurb linter.
h3. *Definition of done*
{color:#505f79}What must be done to consider the task complete?{color}
Remove the suppression of {{FURB}} rules or add an explicit comment explaining why keeping it suppressed is correct.
h3. *Pitfalls*
{color:#505f79}What should the implementer watch out for? What are the risks?{color}
PYTHON-6236ClosedDone
Enable ruff rule B905 for clearer zip() calls
h3. *Context*
{color:#505f79}Describe the background behind the problem.{color}
Ruff's B905 rule enforces {{strict=True/False}} be present on all {{zip()}} calls. This is a clarity improvement.
h3. *Definition of done*
{color:#505f79}What must be done to consider the task complete?{color}
Remove the suppression of B905.
h3. *Pitfalls*
{color:#505f79}What should the implementer watch out for? What are the risks?{color}
PYTHON-6242ClosedDone
Enable ruff PYI rules
h3. *Context*
{color:#505f79}Describe the background behind the problem.{color}
Ruff 0.16.0 added the {{PYI}} category of rules derived from the flake8-pyi linter.
h3. *Definition of done*
{color:#505f79}What must be done to consider the task complete?{color}
Remove the suppression of {{PYI}} rules or add an explicit comment explaining why keeping it suppressed is correct.
h3. *Pitfalls*
{color:#505f79}What should the implementer watch out for? What are the risks?{color}
SERVER-139217Investigating
Add backoff and retry limits for ORP requests
Handle ORP errors differently (overload, network failure, terminal errors). Retry recoverable errors with exponentially increasing backoff + a randomised jitter so request don't storm and all retry together.
We should limit retries over time (including overload retries) and wait when that allowance is exhausted. This recovery policy should apply to both the GetIndexedFrontier requests and the oplog stream requests
Make waits interruptible and we should test that repeated failures are eve
SERVER-139172ClosedWon't Do
Prettier doesn't reflow comments over 100 characters
Would be nice if existing formatting tools auto-formatted long comments in jstests like they do in C++.
SERVER-136810Open
Distinguish different ORP fallback failures
We need to be able to tell apart from the log alone the different reasons we gave up. A few cases I can think of that we'd need separate IDs for are ORP unreachable after bounded retries, requested range doesn't exist in ORP, no progress across retries, the loop threshold hit and we have a newer checkpoint available (so we want to crash reload).
Each time we encounter these failures we should log the last applied LSN, the ORP frontier we saw, first_untruncated_lsn_in_domain from the last LS r
SERVER-122866In Code Review
Query stats collectionType type is incorrect for view over timeseries
Query stats reports "timeseries" as the collectionType for a view over a timeseries collection, when it should be "view".
NODE-7626ClosedFixed
Deprecate ObjectId.cacheHexString
Deprecate the static {{ObjectId.cacheHexString}} opt-in flag, with removal planned for the next major version.
Rationale:
* {{ObjectId.cacheHexString}} is opt-in and {{undefined}} by default. It cannot simply be enabled by default: caching per-instance hex state breaks deep clone (when stored as a {{#private}} field, reverted in NODE-7271) or deep equality (when stored as a visible property). Default-on was rejected in NODE-6480 (Won't Do).
* The packed-integer ObjectId representation (NODE
Components
BSON
Fixed Versions
bson-7.3.5
NODE-7743ClosedFixed
Deprecate SelectServerOptions
Deprecate the {{SelectServerOptions}} interface, with internalization planned for the next major version.
Rationale:
* {{SelectServerOptions}} is marked {{@public}} in src/sdam/topology.ts but is not used by any public API, so it is public by accident rather than by design.
* Making it internal is tracked in NODE-7342, whose stated dependency is that the type must be deprecated first.
* Deprecating it in a 7.x minor gives consumers a release of warning before the type stops being exported.
Sco
Components
TypeScript
Fixed Versions
7.9.0
SERVER-139036ClosedFixed
Add deterministic lints for backports_required_for_multiversion_tests.yml and resmoke suite yamls to check_feature_flag_tags
h2. Summary
Add static, deterministic checks to the existing {{check_feature_flag_tags}} Evergreen task (SERVER-132548) that validate {{etc/backports_required_for_multiversion_tests.yml}} and the resmoke suite yamls. These checks are false-positive-free and run blocking pre-merge.
h2. Background
~21% of Repl's recent test-only hot BFs failed on issues catchable via static analysis of jstests and yaml files. This covers the deterministic half (BF-44237, BF-44795).
h2. Checks
* Veri
Components
Testing Infrastructure
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-139043ClosedFixed
Make SBE BlockHashAgg stage remap input slots
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-138905ClosedFixed
Remove unnecessary uses of SBE ArraySet
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-138888ClosedFixed
Add helper to read getField() results through MakeObj in query code
Fixed Versions
9.1.0-rc09.1.0-rc1025
SERVER-139216Backlog
Add testing for transactions with pipelined oplog application
* Unprepared mixed-CRUD transactions contribute exactly once across workers and collections.
* Large transactions spanning batches/watermarks contribute only at commit.
* Retryable writes and findAndModify/image handling do not duplicate hash contributions
* Command-bearing transactions (e.g. create, drop) retain validation accounting on the serialized path.
GODRIVER-4130Investigating
Introduce Custom Callback logic in Configurable DNS
h2. Context
Consolidates the work from epic GODRIVER-3953 (DRIVERS-3329), this ticket (DRIVERS-3632), and GODRIVER-4173 (DRIVERS-3664) into one implementation.
h2. Definition of Done
*Spec sync*
* DRIVERS-3329: [https://github.com/mongodb/specifications/commit/6f64d0ee3ae49edbdb30eb995f3e29549e8cfa6a]
and [https://github.com/mongodb/specifications/commit/131ab56f64916380f959bb2423a4842bce519e79]
(the latter also defines how to use and parse the public suffix list)
* DRIVERS-3632: [h
Components
Connection String & URI OptionsSRV-based Cluster Discovery
SERVER-124438Open
[Join Optimization] Consider the size of the OS filesystem cache for IO estimation
Currently the join cost model's usage of Mackert Lohman formula for estimation of IOs only considers the size of the WT page cache. In reality, MongoDB also caches pages in the OS page cache as well, making the effective size of the buffer cache larger, though potentially storing duplicates.
We could improve our estimate of INLJ/Index scans for collections which result sets which don't fit in memory by considering the effect of the addition cache.
SERVER-139211ClosedDuplicate
Join cost model picks COLLSCAN + hash join over index nested loop join for in-memory data once cardinality estimates are accurate
h2. Summary
With more accurate cardinality estimates, the join optimizer switches some TPC\-H plans from an index nested loop join \(INLJ\) to a COLLSCAN \+ hash join \(HJ\), and the new plans are 2.8–7.3x slower when the data is in memory, the regime the join plan stability tests and most working sets run in. When the cache is cold they are no faster \(0.98–1.04x\). The join cost model appears to misprice INLJ probes against COLLSCAN \+ HJ in the CPU\-bound \(cached\) regime.
The previous plan