When we generate write batches for time-series inserts in the bucket catalog, it is possible to create different batches which have the same bucketId but contain different sets of measurements. This breaks some assumptions the code makes later during the execution.
- is depended on by
-
SERVER-89825 Investigate why jstests/concurrency/fsm_workloads/timeseries_mirrored_writes.js fails in case of concurrent moveCollection
- Closed
-
SERVER-93149 Re-enable reshardingForTimeseriesFeatureFlagEnabled
- In Code Review
- is related to
-
SERVER-95234 Sharded concurrency suites should shard test-defined collections
- Closed