|
| 1 | +.. _4.2.19-changelog: |
| 2 | + |
| 3 | +4.2.19 Changelog |
| 4 | +---------------- |
| 5 | + |
| 6 | +Sharding |
| 7 | +~~~~~~~~ |
| 8 | + |
| 9 | +- :issue:`SERVER-26755` Collection chunk loading is aborted entirely if fetching all the chunks takes more than 30 seconds |
| 10 | +- :issue:`SERVER-45149` replSetStepDown command in txn_two_phase_commit_failover.js should not timeout |
| 11 | +- :issue:`SERVER-53335` Queries, updates, and deletes with non-"simple" collations may miss documents when using hashed sharding |
| 12 | +- :issue:`SERVER-59929` unexpected slower update/insert operation bease of splitchunk and moveChunk |
| 13 | +- :issue:`SERVER-60682` TransactionCoordinator may block acquiring WiredTiger write ticket to persist its decision, prolonging transactions being in the prepared state |
| 14 | +- :issue:`SERVER-61816` cancel_coordinate_txn_commit_with_tickets_exhausted.js can hang forever due to race condition between transaction reaper and transaction coordinator |
| 15 | +- :issue:`SERVER-61924` retryable_mongos_write_errors.js fails in 4.2 multiversion suites due to missing failpoint command in 4.0 |
| 16 | +- :issue:`SERVER-62065` Upgrade path from 3.6 to 4.0 can leave chunk entries without history on the shards |
| 17 | +- :issue:`SERVER-62739` cancel_coordinate_txn_commit_with_tickets_exhausted.js test should not be run with the ephemeral storage engine |
| 18 | +- :issue:`SERVER-62906` Add a check in the createCollection/shardCollection path verifying the collection name length |
| 19 | + |
| 20 | +Query |
| 21 | +~~~~~ |
| 22 | + |
| 23 | +- :issue:`SERVER-23664` $mod match expression should raise an error when the remainder is not a number |
| 24 | +- :issue:`SERVER-57588` Inconsistent query results when an array position is indexed whose value is an array |
| 25 | +- :issue:`SERVER-59754` Incorrect logging of queryHash/planCacheKey for operations that share the same $lookup shape |
| 26 | +- :issue:`SERVER-62147` Exhaust query using the OP_QUERY protocol is broken when more than one getMore batch is required |
| 27 | + |
| 28 | +Aggregation |
| 29 | +~~~~~~~~~~~ |
| 30 | + |
| 31 | +:issue:`SERVER-44484` Changestream with updateLookup uasserts on updates from before collection was sharded |
| 32 | + |
| 33 | +Storage |
| 34 | +~~~~~~~ |
| 35 | + |
| 36 | +:issue:`SERVER-55483` Add a new startup parameter that skips verifying the table log settings |
| 37 | + |
| 38 | +Internals |
| 39 | +~~~~~~~~~ |
| 40 | + |
| 41 | +- :issue:`SERVER-51087` Add testing utility for creating an empty sharded collection with the specified bounds |
| 42 | +- :issue:`SERVER-53726` Platform Support: Remove RHEL6 zSeries |
| 43 | +- :issue:`SERVER-57312` Pin transitive Python dependencies and use the pinned file for installation in Evergreen |
| 44 | +- :issue:`SERVER-57676` Wait for a checkpoint before releasing the "hangOplogCapMaintainerThread" in oplog_rollover.js |
| 45 | +- :issue:`SERVER-60392` Fix timing in priority_takeover_two_nodes_equal_priority test. |
| 46 | +- :issue:`SERVER-60685` TransactionCoordinator may interrupt locally executing update with non-Interruption error category, leading to server crash |
| 47 | +- :issue:`SERVER-61662` SCons configure checks should always run with verbosity |
| 48 | +- :issue:`SERVER-61743` The -fno-builtin-memcmp flag shouldn't be applied except on x86_64 platforms |
| 49 | +- :issue:`SERVER-62005` [4.2] Add requires_document_locking to unique_index_insert_during_collection_scan.js |
| 50 | +- :issue:`SERVER-62164` Remove several obsoleted build variants across all stable branches |
| 51 | +- :issue:`SERVER-62320` (v4.2) Fix use after std::move in find command |
| 52 | +- :issue:`SERVER-62668` Synchronize access to ImpersonatedUserMetadata in OperationContext. |
| 53 | +- :issue:`SERVER-62825` Robustify causally_consistent_index_builds.js |
| 54 | +- :issue:`SERVER-63197` Pin microbenchmarks genny version |
| 55 | +- :issue:`SERVER-63203` Chunk splitter never splits if more than 8192 split points are found |
| 56 | +- :issue:`WT-7993` If gathering handles and not in aggressive eviction mode, have eviction sleep to let checkpoint grab a contentious spinlock. |
0 commit comments