summaryrefslogtreecommitdiff
Commit message (Expand)AuthorAgeFilesLines
* Cut WiredTiger release 2.2.12.2.1Alex Gorrod2014-06-253-5/+5
* Merge pull request #1083 from wiredtiger/221-release-changelogMichael Cahill2014-06-251-0/+44
|\
| * Fix variable naming: it's called "enabled", not "sync_on_commit".Michael Cahill2014-06-251-1/+1
| * minor reformat for markdown (in the documentation)Michael Cahill2014-06-251-22/+22
| * Add draft ChangeLog for 2.2.1 release.Alex Gorrod2014-06-241-0/+44
* | Slightly shuffle some code in __wt_cache_full_check() to minimize theKeith Bostic2014-06-241-20/+34
* | Update Java example to catch WiredTiger exceptions.Alex Gorrod2014-06-231-6/+21
* | Merge pull request #1076 from wiredtiger/txn-syncMichael Cahill2014-06-2014-215/+291
|\ \
| * | Clarify that the transaction sync setting overrides the global setting. Add ...Michael Cahill2014-06-205-11/+28
| * | Minor cleanup. Add enabled/none test case back in. #1074Susan LoVerso2014-06-195-4/+17
| * | Split the global transaction_sync configuration into two parts: a sync method...Michael Cahill2014-06-1911-81/+90
| * | Fix a formatting problem with configuration defaults: sub-configurations of t...Michael Cahill2014-06-192-170/+168
| * | Add per-txn sync configuration support. #1074Susan LoVerso2014-06-189-26/+65
* | | whitespaceKeith Bostic2014-06-191-7/+7
* | | Merge pull request #1077 from wiredtiger/lsm_lock_flagAlex Gorrod2014-06-192-10/+10
|\ \ \
| * | | Revert name change of flag to WT_SESSION_NO_CACHE_CHECKAlex Gorrod2014-06-194-6/+6
| * | | Replace WT_SESSION_NO_CACHE_CHECK with a more general-purpose flag,Keith Bostic2014-06-185-14/+14
|/ / /
* | | Merge pull request #1075 from wiredtiger/debug-lsm-dirty-after-checkpointAlex Gorrod2014-06-199-183/+260
|\ \ \ | |/ / |/| |
| * | Fix problems with reopening an LSM tree: don't set the primary chunk until we...Michael Cahill2014-06-181-17/+20
| * | Wait for ordinary LSM merges to quiesce before starting compaction, to avoid ...Michael Cahill2014-06-183-34/+40
| * | When flushing LSM chunks, only check switch_txn for in-memory chunks.Michael Cahill2014-06-181-6/+6
| * | Cleanup: undo debugging changes to get this branch ready for review.Michael Cahill2014-06-1811-94/+18
| * | Merge branch 'develop' into debug-lsm-dirty-after-checkpointMichael Cahill2014-06-183-60/+107
| |\ \ | |/ / |/| |
* | | The split transaction ID may not be globally visible by the time weKeith Bostic2014-06-171-1/+1
* | | Add sentence about close event handler alternative.Susan LoVerso2014-06-171-0/+3
* | | Add sentence about waiting for worker threads.Susan LoVerso2014-06-171-1/+3
* | | Have wtperf script run 3 times and check for stability #1061Susan LoVerso2014-06-171-58/+100
| * | Instead of having update operations set the highest update transaction in an ...Michael Cahill2014-06-189-94/+85
| * | Wait until after the LSM tree state is updated (including getting an exclusiv...Michael Cahill2014-06-171-21/+28
| * | Merge branch 'develop' into debug-lsm-dirty-after-checkpointMichael Cahill2014-06-179-39/+107
| |\ \ | |/ / |/| |
* | | Add support for random isolation configurations, where each new sessionKeith Bostic2014-06-164-29/+78
* | | Add a new field to the WT_CURSOR_BTREE structure so we can pass back aKeith Bostic2014-06-164-9/+28
* | | Remove tab characters.Keith Bostic2014-06-161-1/+1
| * | Check for overflow before attempting an LSM put. Also relax the assertions t...Michael Cahill2014-06-173-95/+138
| * | Fix the LSM readonly flag, only assert that files older than the current one ...Michael Cahill2014-06-173-4/+7
| * | Only set the read-only flag if a checkpoint completed successfullyMichael Cahill2014-06-161-1/+1
| * | Merge branch 'develop' into debug-lsm-dirty-after-checkpointMichael Cahill2014-06-1668-534/+1044
| |\ \ | |/ / |/| |
* | | Merge pull request #1071 from wiredtiger/invisible-updatesMichael Cahill2014-06-167-133/+472
|\ \ \
| * | | Add a comment.Keith Bostic2014-06-151-1/+5
| * | | Flip the sense of the __cursor_invalid function to __cursor_validKeith Bostic2014-06-151-34/+31
| * | | Add a slightly different test, where there are visible records in theKeith Bostic2014-06-151-6/+71
| * | | Fixes for WT_INSERT objects with invisible WT_UPDATE objects. ForKeith Bostic2014-06-143-37/+183
| * | | update a comment.Keith Bostic2014-06-141-2/+2
| * | | Change WT_CURSOR.next-random to handle insert lists where the updatesKeith Bostic2014-06-144-41/+114
| * | | Confirm WT_CURSOR.next_random works for both in-memory and on-diskKeith Bostic2014-06-141-2/+21
| * | | Use self.reopen_conn() instead of rolling our own.Keith Bostic2014-06-141-3/+2
| * | | When returning from WT_CURSOR.update, we previously called __wt_kv_returnKeith Bostic2014-06-141-3/+10
| * | | Once we find an update object we can use, remember it and pass it intoKeith Bostic2014-06-134-59/+88
* | | | Merge pull request #1063 from wiredtiger/lsm-compact-improvementsAlex Gorrod2014-06-165-28/+65
|\ \ \ \ | |/ / / |/| | |
| * | | Set split_pct by default in wtperf: we use a 4KB allocation size and a 4KB le...Michael Cahill2014-06-131-1/+2
View Lorry Controller Status