-
Type:
Task
-
Resolution: Fixed
-
Priority:
Major - P3
-
Affects Version/s: None
-
Component/s: None
-
None
-
Query Integration
-
Fully Compatible
-
None
-
None
-
None
-
None
-
None
-
None
-
None
Besides the "simple" write batches, UWE also contains different types of batches that are not handled:
- TwoPhaseWriteBatch
- InternalTransactionBatch (this seems to only be used for retryable time series updates and we are not supporting time series updates yet)
- MultiWriteBlockingMigrationsBatch
We should have tests for them, and understand what we can handle and what we cannot. We process these kinds of writes here: WriteBatchResponseProcessor::_onWriteBatchResponse(NoRetryWriteBatchResponse)
- depends on
-
SERVER-112050 Collect query stats for update in sharded
-
- Closed
-
- related to
-
SERVER-121266 Design behavior for retryable timeseries updates
-
- Backlog
-
-
SERVER-121325 Query stats double-counts update when _id but no shard key present
-
- Backlog
-
-
SERVER-121267 Design behavior for would-change-owning-shard updates
-
- Needs Scheduling
-
-
SERVER-121005 Make sure metrics are correct for different types of writes
-
- Needs Scheduling
-