summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Cut WiredTiger release 2.0.12.0.1Alex Gorrod2013-12-127-15/+15
|
* Merge pull request #806 from wiredtiger/lsm-chunk-flags-atomicMichael Cahill2013-12-117-54/+63
|\ | | | | Switch to atomic flags for LSM chunks, to avoid racing when setting merge flags
| * Switch to atomic flags for LSM chunks, to avoid racing when setting merge flags.Michael Cahill2013-12-127-54/+63
|/
* spellingMichael Cahill2013-12-121-2/+2
|
* Add recurse to spelling list.Alex Gorrod2013-12-121-0/+1
|
* Merge pull request #803 from wiredtiger/lintAlex Gorrod2013-12-112-4/+9
|\ | | | | Coverity + lint
| * lint + CoverityKeith Bostic2013-12-112-4/+9
| |
* | Merge pull request #804 from wiredtiger/lsm-compact-fixAlex Gorrod2013-12-115-25/+60
|\ \ | |/ |/| Fix LSM compact so it doesn't try to compact the file inside an LSM tree
| * Fix LSM compact so it doesn't try to compact the file inside an LSM tree.Michael Cahill2013-12-125-25/+60
|/
* Cut WiredTiger release 2.0.02.0.0Alex Gorrod2013-12-116-16/+63
|
* Merge pull request #802 from wiredtiger/lsm-merge-cleanupAlex Gorrod2013-12-101-8/+11
|\ | | | | | | Drop partial results after an interrupted LSM merge.
| * Drop partial results after an interrupted LSM merge.Michael Cahill2013-12-111-8/+11
|/
* Merge pull request #792 from wiredtiger/compact-lsmMichael Cahill2013-12-1018-83/+249
|\ | | | | Allow LSM trees to be compacted.
| * Alphabetize LSM tree flags.Michael Cahill2013-12-111-5/+5
| | | | | | | | refs #792
| * LSM compaction review updates. Implement Sue's suggestions, be more ↵Michael Cahill2013-12-115-51/+47
| | | | | | | | | | | | | | aggressive about waking the merge threads, fix some merge pathology. --HG-- rename : bench/wtperf/runners/medium-lsm-sleep.wtperf => bench/wtperf/runners/medium-lsm-compact.wtperf
| * icount is uint32_t. #792Susan LoVerso2013-12-091-1/+1
| |
| * Show running total during populate. #792Susan LoVerso2013-12-091-2/+4
| |
| * Change wtperf to remove merge_sleep and add compact. #792Susan LoVerso2013-12-093-14/+17
| |
| * Nits.Alex Gorrod2013-12-062-2/+1
| |
| * Merge branch 'develop' into compact-lsmAlex Gorrod2013-12-0634-242/+871
| |\ | | | | | | | | | | | | Conflicts: src/lsm/lsm_merge.c
| * | Implement compact for LSM trees. Add a new timeout compact configuration.Alex Gorrod2013-12-0614-47/+211
| | |
* | | Merge pull request #797 from wiredtiger/overflow-keyMichael Cahill2013-12-106-23/+61
|\ \ \ | | | | | | | | Overflow key delete race.
| * | | Rename the overflow lock, now that it is used for keys and values.Michael Cahill2013-12-116-16/+16
| | | | | | | | | | | | | | | | refs #797
| * | | typoMichael Cahill2013-12-112-2/+2
| | | |
| * | | Merge pull request #799 from wiredtiger/overflow-instantiateKeith Bostic2013-12-101-9/+12
| |\ \ \ | | | | | | | | | | Remove call to instantiate a deleted overflow key during reconciliation.
| | * | | Turns out truncation does delete without instantiating keys -- putKeith Bostic2013-12-101-6/+12
| | | | | | | | | | | | | | | | | | | | the code back and update the comment for clarity.
| | * | | I don't believe it's possible for a key/value pair to be deleted withoutKeith Bostic2013-12-101-10/+7
| |/ / / | | | | | | | | | | | | | | | | instantiating the key, which makes this call to __wt_row_leaf_key_work unnecessary; assert the fact.
| * | | Don't let reconciliation delete overflow keys out from under readersKeith Bostic2013-12-102-3/+38
| | | | | | | | | | | | | | | | instantiating them. #789
* | | | Improve the install instructions from wtstats.py if python-nvd3 is not found.Michael Cahill2013-12-111-3/+3
| | | |
* | | | Revert 4df3f8d705c1176d970fe4ddd6ea6ae2d75c8ebf, pushed by mistake.Michael Cahill2013-12-111-10/+2
| | | |
* | | | imported patch schema-lock-timeout-abortMichael Cahill2013-12-111-2/+10
| | | |
* | | | In test/format, have bloom_oldest imply bloom=1 to avoid an error in ↵Michael Cahill2013-12-112-5/+7
|/ / / | | | | | | | | | WT_SESSION::create.
* | | Use a flag in LSM cursors to track whether the cursor has incremented the ↵Michael Cahill2013-12-102-13/+42
| | | | | | | | | | | | session count, clean it up in a few more places in case error paths leave it set.
* | | Avoid bouncing the session cursor count between zero and one when searching ↵Michael Cahill2013-12-106-48/+66
| | | | | | | | | | | | an LSM tree: increment the session's cursor count at the beginnging of an LSM search and decrement at the end. This also means we can get rid of the special purpose "cache busy" flag.
* | | Make LSM chunk size configurable in test/format.Michael Cahill2013-12-103-0/+7
| | |
* | | Fix a long-standing test/format bug that meant values repeated *way* more ↵Michael Cahill2013-12-101-1/+1
| | | | | | | | | | | | often than configured. This showed up because repeating overflow items with large pages result in pathological (quadratic) performance degradation.
* | | whitespaceKeith Bostic2013-12-091-2/+2
| | |
* | | Turn off archiving on primary connection. #710Susan LoVerso2013-12-093-5/+8
| | |
* | | Increase the cache size so eviction doesn't happen, and increase theKeith Bostic2013-12-091-3/+4
| | | | | | | | | | | | allowed pages for "success" so we don't randomly fail. Closes #787.
* | | Update test/format to include LSM configuration options.Alex Gorrod2013-12-092-6/+4
| | |
* | | Do a hot backup to test recovery, otherwise the copy can race with log ↵Michael Cahill2013-12-096-40/+30
| | | | | | | | | | | | archiving removing a log file. Switch a few tests to use the "standard" backup, use shutil.rmtree instead of our own version.
* | | We've already read the shared state, use the local copies.Michael Cahill2013-12-091-1/+1
| | |
* | | Use a field in the session to avoid locking during close rather than setting ↵Michael Cahill2013-12-093-6/+13
| | | | | | | | | | | | | | | | | | a flag -- the latter can race. refs #793
* | | Acquire the live system's lock around the save/restore of the first-firstKeith Bostic2013-12-071-2/+14
| | | | | | | | | | | | | | | flag -- we don't currently need it (because we only read the flag in one place), but it's safer going forward if that ever changes.
* | | Free up allocated memory -- it makes Coverity happier.Keith Bostic2013-12-071-21/+41
| | |
* | | Allow replay of threaded runs.Keith Bostic2013-12-076-53/+62
| | |
* | | Get rid of CONFIG.example, it's never correct, futz with the configKeith Bostic2013-12-073-99/+10
| | | | | | | | | | | | error output a little to make it more readable.
* | | Internal WT programs are using WT_UNUSED everywhere, test/bloom was theKeith Bostic2013-12-071-47/+11
| | | | | | | | | | | | | | | | | | last hold-out. While in the area, noticed test/bloom doesn't need signal or event handlers, fix function signatures, and there was an extraneous newline written by the usage function.
* | | comment spelling / wordingMichael Cahill2013-12-071-1/+1
| | |
* | | Replace an explicit (void)arg with WT_UNUSED.Keith Bostic2013-12-061-1/+1
| | |
View Lorry Controller Status