summaryrefslogtreecommitdiff
AgeCommit message (Expand)Author
2014-07-29Cut WiredTiger release 2.3.02.3.0Alex Gorrod
2014-07-29Sanity check the LevelDB configuration at build time, not runtime.Michael Cahill
2014-07-29Make the LevelDB API more resilient to invalid configurations. Also clean up...Michael Cahill
2014-07-29Add upgrading section to docs for 2.3.0 release.Alex Gorrod
2014-07-29Whitespace.Alex Gorrod
2014-07-29Merge pull request #1137 from wiredtiger/changelog-2.3.0Alex Gorrod
2014-07-29Add a page to the documentation describing how to use the LevelDB API.Michael Cahill
2014-07-29Minor changelog edits.Michael Cahill
2014-07-29Add details for query performance improvements to changelog.Alex Gorrod
2014-07-29Merge pull request #1123 from wiredtiger/force-log-ckptMichael Cahill
2014-07-28Fix type warning.Susan LoVerso
2014-07-28Force a full database checkpoint at the end of recovery. #1115Susan LoVerso
2014-07-28Add draft changelog notes for upcoming 2.3.0 release.Alex Gorrod
2014-07-28Update function naming to be consistent with our style.Alex Gorrod
2014-07-25Swap out WT_ERR for WT_RET, there's no cleanup needed.Keith Bostic
2014-07-25Remove unused temporary buffer.Keith Bostic
2014-07-25I don't expect to ever drill a path so we don't need the WT_BTREE.ckpt field,Keith Bostic
2014-07-25There's no need to test if the handle is non-NULL before callingKeith Bostic
2014-07-25Fix comment alignment.Keith Bostic
2014-07-25__wt_eviction_force_check() is only called from one place, there's noKeith Bostic
2014-07-25Combine the not-modified and not-checkpoint code in one place rather thanKeith Bostic
2014-07-25Touch elements in order (no real change).Keith Bostic
2014-07-24Merge pull request #1126 from wiredtiger/bulk-checkpointMichael Cahill
2014-07-23Switch from using the checkpoint lock to serialize bulk-load close andKeith Bostic
2014-07-23Revert "Switch from using the checkpoint lock to serialize bulk-load close and"Keith Bostic
2014-07-23Merge branch 'develop' into bulk-checkpointKeith Bostic
2014-07-23Don't initialize name_alloc twice.Keith Bostic
2014-07-22Switch from using the checkpoint lock to serialize bulk-load close andKeith Bostic
2014-07-22Add a compact to the evict-lsm job to reduce variability during the query phase.Michael Cahill
2014-07-20Remove some initialization code that's no longer needed.Keith Bostic
2014-07-20Don't release the page if called from the split code, it's referencedKeith Bostic
2014-07-20Tiny variable name change for consistency.Keith Bostic
2014-07-20Remove another few instructions from the read-only row-search code,Keith Bostic
2014-07-20Do tree-append tracking in the cursor instead of in the btree handle.Keith Bostic
2014-07-18Fix a search past the end of a skiplist: we need to set `cbt->ins` to the las...Michael Cahill
2014-07-18Merge pull request #1121 from wiredtiger/row-search-slimMichael Cahill
2014-07-18Merge pull request #1122 from wiredtiger/row-search-slim-fast-append-firstMichael Cahill
2014-07-17Merge branch 'row-search-slim' into row-search-slim-fast-append-firstKeith Bostic
2014-07-17Clean up __cursor_pos_clear: only clear the fields we need to clear andKeith Bostic
2014-07-17updateKeith Bostic
2014-07-17Rename __cursor_search_clear __cursor_pos_clear; it's the code thatKeith Bostic
2014-07-17Force a checkpoint at the end of recovery so that the checkpoint LSNSusan LoVerso
2014-07-17Move connection config out to a macro.Susan LoVerso
2014-07-17Do the fast-append test before we search the leaf page.Keith Bostic
2014-07-17Remove a comment that no longer applies (and even lint realizes it now).Keith Bostic
2014-07-17Use "ins" instead of "ret_ins", it doesn't add any additional information.Keith Bostic
2014-07-17Add a comment explaining the WT_BTREE.appending behavior.Keith Bostic
2014-07-17Inline the fast-path insert-list append code and only call it when doingKeith Bostic
2014-07-17Don't call the insert list search routine unless there's an insert listKeith Bostic
2014-07-17Some minor code-shuffling in the leaf page row-search code toKeith Bostic