Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | SERVER-38311 Change out merging strategy | Charlie Swanson | 2019-02-13 | 3 | -3/+69 |
| | | | | | Allows an $out stage to run in parallel on all shards if the target collection is sharded and so is the input collection to the aggregate. | ||||
* | SERVER-39565 Add requires_document_locking tag. | Max Hirschhorn | 2019-02-13 | 1 | -0/+2 |
| | |||||
* | SERVER-39488 Look for storage stats in the find command's profiled entry. | nehakhatri5 | 2019-02-13 | 1 | -1/+2 |
| | | | | | | | The find command will always have the storage stats because the test has created a collection with btree spanning multiple pages. Scanning this collection after a server restart will trigger read from the disk and have the disk read stats. | ||||
* | SERVER-38621 Do not ignore regex options when specified first | Evan Nixon | 2019-02-12 | 1 | -0/+19 |
| | |||||
* | SERVER-39376 Canonicalize net.ssl.mode to net.tls.mode | Sara Golemon | 2019-02-12 | 1 | -0/+39 |
| | |||||
* | SERVER-38583 JS test fixes for write errors in transactions | A. Jesse Jiryu Davis | 2019-02-12 | 4 | -19/+41 |
| | |||||
* | SERVER-38583 Fix transaction insert writeError format | A. Jesse Jiryu Davis | 2019-02-12 | 5 | -10/+264 |
| | |||||
* | SERVER-38372 Fix rollback of index drop to update CollectionInfoCache. | David Storch | 2019-02-12 | 1 | -0/+35 |
| | |||||
* | SERVER-39528 inc timeouts for socket_disconnect_kills | Jason Carey | 2019-02-12 | 1 | -2/+2 |
| | | | | | The current timeouts are a bit too low for some variants. Let's increase them to fix some flakiness. | ||||
* | SERVER-38516 Tighten up tests that don't fail when not closing connections ↵ | Matthew Russotto | 2019-02-12 | 19 | -97/+35 |
| | | | | during stepdown bbut allow exceptions during stepdown | ||||
* | SERVER-38516 Remove “closeConnectionsOnStepdown” flag and fix tests ↵ | Matthew Russotto | 2019-02-12 | 37 | -307/+102 |
| | | | | broken by this | ||||
* | SERVER-39474 Fix incorrect use of 'skip' parameter to configureFailPoint in ↵ | Esha Maharishi | 2019-02-11 | 2 | -4/+5 |
| | | | | txn_two_phase_commit_failover.js | ||||
* | SERVER-39068 Replicate startIndexBuild and commitIndexBuild oplog entrires | Louis Williams | 2019-02-12 | 2 | -12/+99 |
| | |||||
* | SERVER-35551 Avoid restarting mongobridge processes. | Max Hirschhorn | 2019-02-11 | 3 | -10/+51 |
| | | | | | | | Changes ReplSetTest, RollbackTest, and RollbackTestDeluxe to avoid restarting the mongobridge process associated with the mongod process being restarted. This ensures that any partitioning which has been configured remains intact after the server is restarted. | ||||
* | SERVER-39333 Add assertions that operations work in coll_epoch_test2.js | jannaerin | 2019-02-11 | 1 | -20/+26 |
| | |||||
* | SERVER-39169 Add $_internalReadAtClusterTime option to find and dbHash. | Max Hirschhorn | 2019-02-11 | 4 | -86/+313 |
| | | | | | | | | | | | | The new $_internalReadAtClusterTime option replaces all usages of running the dbHash command inside of a multi-statement transaction. It can be used to read from a consistent snapshot in place of specifying an atClusterTime read concern. Unlike multi-statement transactions, the new $_internalReadAtClusterTime option doesn't cause locks to be left on the server after returning a network response. It instead restores the snapshot to read from as part of handling the request. | ||||
* | SERVER-39328 Ensure unacked write starts and finishes before stepping down ↵ | Matthew Russotto | 2019-02-11 | 1 | -7/+24 |
| | | | | in last_error_reported_after_stepdown | ||||
* | SERVER-36489 Start initial sync oplog fetching from the 'oldest active ↵ | Samy Lanka | 2019-02-11 | 4 | -2/+353 |
| | | | | | | | transaction timestamp' SERVER-36490 Initial sync should not actually prepare transactions on applying prepareTransaction oplog entries SERVER-36491 During initial sync, make commitTransaction oplog entries apply the transaction from the prepare oplog entry | ||||
* | SERVER-37283 Use stronger locks for system.views | Charlie Swanson | 2019-02-11 | 6 | -108/+129 |
| | | | | | | | | | | Readers of the view catalog depend on a MODE_IS DB lock preventing concurrent writes to the view catalog. This is true for regular view maintenance commands like collMod, create, and drop. However, on secondaries these commands are replicated as direct writes to system.views and do not hold as strong of a lock. Further, a user is permitted to write directly to system.views and so could hit a similar issue on the primary. | ||||
* | SERVER-39183 honor socket disconnect in $where | Jason Carey | 2019-02-10 | 1 | -0/+23 |
| | |||||
* | SERVER-39150 markKillOnClientDisconnect | Jason Carey | 2019-02-10 | 2 | -10/+220 |
| | |||||
* | SERVER-39149 Homogenize getMore behavior | Jason Carey | 2019-02-08 | 2 | -2/+4 |
| | | | | | | | | | | | | | Cursor's have subtly different semantics across mongos and mongod and between legacy getMore and command getMore. (as does the find command) Ensuring that all getMores can be paused by waitAfterPinningCursorBeforeGetMoreBatch, and that cursors are cleaned up if they are killed after verifying auth, makes testing a bit simpler and cursors less leaky And adding a check for waitInFindBeforeMakingBatch to runQuery similarly makes it easier to test | ||||
* | SERVER-39090 Remove sort7.js and fold coverage into sort_array.js | Charlie Swanson | 2019-02-08 | 2 | -28/+60 |
| | |||||
* | SERVER-39463 Add logging to plan_cache_shell_helpers.js | Ted Tuckman | 2019-02-08 | 1 | -36/+51 |
| | |||||
* | SERVER-37836 re-evaluate authorization for originating command in getMore | Isabella Siu | 2019-02-08 | 1 | -12/+121 |
| | |||||
* | SERVER-39274 Mongos shouldn't upconvert txn read concern level to snapshot ↵ | Jack Mulrow | 2019-02-08 | 4 | -15/+38 |
| | | | | if none is provided | ||||
* | Revert "SERVER-39149 Homogenize getMore behavior" | James Wahlin | 2019-02-08 | 2 | -4/+2 |
| | | | | This reverts commit 4d703e26c2801971d538f948a4dc3191994f0074. | ||||
* | SERVER-39269 flag gate sharded $lookup | Ian Boros | 2019-02-07 | 10 | -60/+214 |
| | |||||
* | SERVER-39149 Homogenize getMore behavior | Jason Carey | 2019-02-07 | 2 | -2/+4 |
| | | | | | | | | | | | | | Cursor's have subtly different semantics across mongos and mongod and between legacy getMore and command getMore. (as does the find command) Ensuring that all getMores can be paused by waitAfterPinningCursorBeforeGetMoreBatch, and that cursors are cleaned up if they are killed after verifying auth, makes testing a bit simpler and cursors less leaky And adding a check for waitInFindBeforeMakingBatch to runQuery similarly makes it easier to test | ||||
* | SERVER-39312 Rename targeted jstests for transaction two phase commit to ↵ | Esha Maharishi | 2019-02-07 | 6 | -0/+0 |
| | | | | txn_two_phase_commit_*.js | ||||
* | SERVER-39292 Do not evaluate failpoint twice per operation when multiple ↵ | jannaerin | 2019-02-07 | 1 | -0/+22 |
| | | | | options specified | ||||
* | SERVER-39215 Fix aggregation_currentop.js such that we compare ↵ | Suganthi Mani | 2019-02-07 | 1 | -3/+3 |
| | | | | transactionDocument.startWallClockTime is greater than epoch time (1970-01-01T00:00:00.000). | ||||
* | SERVER-39053 Transaction coordinator should retry sending decision until ↵ | Esha Maharishi | 2019-02-07 | 4 | -52/+89 |
| | | | | definitively hearing an ack | ||||
* | SERVER-39394 Traverse $lookup subpipeline for more involved collections | Charlie Swanson | 2019-02-06 | 1 | -0/+68 |
| | |||||
* | SERVER-39139 Disallow commitTransaction and abortTransaction commands on ↵ | Vesselina Ratcheva | 2019-02-06 | 5 | -176/+12 |
| | | | | secondaries | ||||
* | SERVER-35402 Add validators where applicable in query_knobs.cpp | Arun Banala | 2019-02-06 | 1 | -0/+171 |
| | |||||
* | SERVER-38986 Timestamp writes while draining | Louis Williams | 2019-02-06 | 1 | -4/+8 |
| | |||||
* | SERVER-38701 Prohibit two phase index build oplog entries through applyOps | Louis Williams | 2019-02-06 | 1 | -0/+45 |
| | |||||
* | SERVER-39145 Modify no_disconnect_on_stepdown.js to test that the write ↵ | Matthew Russotto | 2019-02-06 | 1 | -19/+29 |
| | | | | operations are killed during step down. | ||||
* | SERVER-38519 Test transactions with new stepdown sequence. | Suganthi Mani | 2019-02-06 | 2 | -36/+143 |
| | |||||
* | SERVER-38515 Test that initial sync continues through a step down. | Suganthi Mani | 2019-02-06 | 2 | -0/+178 |
| | |||||
* | SERVER-39375 fix trafficRecordingFileToBSONArr | Jason Carey | 2019-02-05 | 1 | -0/+4 |
| | | | | | | | | | The convertTrafficRecordingToBSON function in the shell calls trafficRecordingFileToBSONArr. When called on a non-existant file, this leaves inputFd <0, which causes crashes when ::close is invoked. The fix is to move the guard which cleans up the file until after the check that it is valid | ||||
* | SERVER-39365 shell_load_file.js check for windows filenames | Gabriel Russell | 2019-02-05 | 1 | -2/+3 |
| | |||||
* | SERVER-38435 Update assertion messages in txn broadcast writes fsm workloads | Jack Mulrow | 2019-02-05 | 2 | -11/+13 |
| | |||||
* | SERVER-38906 Test readConcern local txn isolation | A. Jesse Jiryu Davis | 2019-02-05 | 2 | -10/+79 |
| | |||||
* | SERVER-38161 Test that prepared transactios can span state transitions | Samy Lanka | 2019-02-04 | 2 | -7/+107 |
| | |||||
* | SERVER-37823 Server Side Traffic Capture | Jason Carey | 2019-02-04 | 11 | -5/+332 |
| | | | | | Adds support for special commands which dump wire protocol traffic to disk. | ||||
* | SERVER-39056 Further refine readWriteAnyDatabase | Shreyas Kalyan | 2019-02-04 | 3 | -78/+127 |
| | |||||
* | SERVER-39353 disable chaining in rollback_test.js | Pavi Vetriselvan | 2019-02-04 | 1 | -0/+1 |
| | |||||
* | SERVER-39272 reconnect tiebreaker node and restart repl after rollback completes | Pavi Vetriselvan | 2019-02-04 | 1 | -11/+5 |
| |