summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Merge pull request #1510 from wiredtiger/none-empty-defaultsmongodb-2.8-rc32.8-rc3Alex Gorrod2014-12-234-26/+27
|\ | | | | Use "none" for configurations in documentation but store empty defaults....
| * Use "none" for configurations in documentation but store empty defaults. ↵Michael Cahill2014-12-224-26/+27
| | | | | | | | | | | | This is both more efficient than storing "none" and backwards compatible. refs #1417
* | Merge pull request #1511 from wiredtiger/pkg-configAlex Gorrod2014-12-231-2/+2
|\ \ | | | | | | Remove version numbering from the pkg-config file.
| * | Remove version numbering from the pkg-config file. We don't createDon Anderson2014-12-191-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | include files that are named by version. Programs linked using -lwiredtiger will follow the symlink to wiredtiger-a.b.c.so, so their referred library name is forever stamped as wiredtiger-a.b.c.so, which won't conflict even when we ship wiredtiger-a.b.d.so . Refs #1458.
* | | Don't clear the WT_CONFIG_ITEM before calling __wt_config_getones_none(),Keith Bostic2014-12-221-7/+4
| | | | | | | | | | | | it's not needed, whitespace.
* | | Prettiness police, NULL vs. nul, KNF, no functional change.Keith Bostic2014-12-223-13/+12
| | |
* | | If __wt_getenv fails to allocate memory, we need to fail gracefully.Keith Bostic2014-12-221-1/+3
| | |
* | | Fix bugs reported by Coverity:Alex Gorrod2014-12-221-3/+2
| |/ |/| | | | | | | | | | | 1260223 1260224 One potential memory leak, and one redundant NULL check.
* | Merge pull request #1476 from wiredtiger/format-recoverMichael Cahill2014-12-225-80/+159
|\ \ | | | | | | use format as a simple recovery tester
| * | Add configuration options for log archive and pre-allocation, currentlyKeith Bostic2014-12-154-78/+113
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | at 50% each. Turn on logging all the time in the recovery test script. Explode the configuration snprintf() for wiredtiger_open. The test to check for configuration buffer overrun in WT_SESSION.create was wrong, fix it.
| * | Clean up run reporting.Keith Bostic2014-12-151-3/+7
| | |
| * | Add a simple test of recovery -- on timed runs, the monitor threadKeith Bostic2014-12-154-3/+43
| | | | | | | | | | | | | | | optionally indirects through a NULL and core dumps, then we verify the object.
* | | Merge pull request #1500 from wiredtiger/evict-trickleMichael Cahill2014-12-223-6/+32
|\ \ \ | | | | | | | | Use the eviction server to write pages with READGEN_OLDEST set.
| * | | Rename flags to call this phase "eviction of pages that would block ↵Michael Cahill2014-12-223-12/+12
| | | | | | | | | | | | | | | | application threads".
| * | | Eviction should do update-restore if upper layers are trying to force out a ↵Michael Cahill2014-12-221-4/+5
| | | | | | | | | | | | | | | | page, regardless of its size. Also, only look at ref->page after checking for exclusive access. It is possible (but very unlikely) that a child page pointer could be replaced in the window where we are checking hazard pointers.
| * | | Only trigger the eviction server to look for pages that would force eviction ↵Michael Cahill2014-12-226-11/+14
| | | | | | | | | | | | | | | | if regular eviction either fails or does an in-memory split.
| * | | Merge branch 'develop' into evict-trickleAlex Gorrod2014-12-2213-34/+111
| |\ \ \
| * | | | Check that handles are not being walked by eviction before discarding.Michael Cahill2014-12-192-2/+3
| | | | | | | | | | | | | | | | | | | | refs #1497
| * | | | Use the eviction server to write pages with READGEN_OLDEST set.Alex Gorrod2014-12-197-7/+29
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Even before the eviction trigger has been reached. This should mean that we clear those pages out of cache earlier, and hopefully will save application threads from doing the evictions (at least sometimes).
* | | | | Use the original page's first_dirty_txn when restoring updates to match what ↵Michael Cahill2014-12-221-6/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | we do for in-memory splits. refs #1475
* | | | | Merge pull request #1508 from wiredtiger/avoid-checkpoint-ebusyAlex Gorrod2014-12-221-0/+9
|\ \ \ \ \ | | | | | | | | | | | | Avoid EBUSY returns to verify and salvage caused by checkpoints
| * | | | | Add comments before getting the checkpoint lock.Michael Cahill2014-12-221-0/+3
| | | | | |
| * | | | | Revert some debugging changes.Michael Cahill2014-12-221-7/+3
| | | | | |
| * | | | | Avoid EBUSY returns to verify and salvage caused by checkpoints. The "fix" ↵Michael Cahill2014-12-222-3/+13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | involves blocking checkpoints while salvage or verify are in progress. refs #1404, SERVER-16457
* | | | | | Merge pull request #1507 from wiredtiger/extractor-none-configAlex Gorrod2014-12-221-26/+25
|\ \ \ \ \ \ | |/ / / / / |/| | | | | Have extractor configuration checking match collator.
| * | | | | Have extractor configuration checking match collator.Alex Gorrod2014-12-221-26/+25
|/ / / / / | | | | | | | | | | | | | | | Clean up - no intended functionality change.
* | | | | Make the cache bytes-written and bytes-read match, both shouldKeith Bostic2014-12-211-2/+2
| | | | | | | | | | | | | | | | | | | | ignore compression. Reference #1505.
* | | | | Merge pull request #1506 from wiredtiger/extractor-terminateMichael Cahill2014-12-222-3/+10
|\ \ \ \ \ | |_|/ / / |/| | | | Fix a bug where a custom extractor terminate was being called twice.
| * | | | Fix a bug where a custom extractor terminate was being called twice.Alex Gorrod2014-12-222-3/+10
|/ / / / | | | | | | | | | | | | | | | | | | | | | | | | Resolves issue #1503. Clarify the custom extractor and collator terminate documentation while I'm here.
* | | | Merge pull request #1504 from markbenvenuto/wt_getenvAlex Gorrod2014-12-228-13/+87
|\ \ \ \ | | | | | | | | | | Add __wt_getenv to workaround caching in MSVC CRT's getenv implementation
| * | | | Add __wt_getenv to workaround caching in MSVC CRT's getenv implementation.Mark Benvenuto2014-12-198-13/+87
| | |_|/ | |/| |
* | | | Merge pull request #1502 from markbenvenuto/win_test_schema03Alex Gorrod2014-12-221-3/+13
|\ \ \ \ | |/ / / |/| | | Skip test_schema03.py on Window since rlimits are unix specific
| * | | Skip test_schema03.py on Window since rlimits are unix specificMark Benvenuto2014-12-191-3/+13
|/ / /
* | | Remove the SET/CLEAR_BTREE_IN_SESSION macros: now we have WT_WITH_DHANDLE, ↵Michael Cahill2014-12-194-18/+7
| | | | | | | | | | | | they don't make much sense.
* | | Insulate against language-specific sort order in s_win.Michael Cahill2014-12-191-0/+4
| | |
* | | Check that handles are not being walked by eviction before discarding.Michael Cahill2014-12-192-2/+3
|/ / | | | | | | refs #1497
* | Merge pull request #1499 from wiredtiger/sweep-no-write-leavesAlex Gorrod2014-12-191-30/+20
|\ \ | | | | | | Don't try to write leaves from the sweep server
| * | Don't try to write leaves from the sweep server. Previously, this was done ↵Michael Cahill2014-12-191-30/+20
|/ / | | | | | | | | | | | | | | without locking the handle, and so could race with LSM discarding a handle. We know the handle has been idle, so there is a good chance that a checkpoint has run since the last update and the write leaves was wasted effort. If not, this change will keep the handle locked for longer preventing new opens, but it has been idle for a while anyway. refs #1495, #1497 (maybe)
* | Merge pull request #1496 from wiredtiger/server-16546Alex Gorrod2014-12-199-40/+77
|\ \ | | | | | | Don't try to set and clear session->split_gen in WT_INTL_FOREACH
| * | Set split-gen when deepening the tree, add paranoia that the eviction server ↵Michael Cahill2014-12-183-2/+7
| | | | | | | | | | | | doesn't keep split-gen pinned.
| * | Don't try to set and clear session->split_gen in WT_INTL_FOREACH_BEGIN: ↵Michael Cahill2014-12-187-38/+70
| | | | | | | | | | | | | | | | | | there are too many of those loops to ensure that none of them skip clearing it. Instead, make sure all calls are wrapped in WT_WITH_APAGE_INDEX. refs SERVER-16546
* | | Merge pull request #1494 from markbenvenuto/win_test_priv01_skipMichael Cahill2014-12-191-0/+5
|\ \ \ | | | | | | | | Skip test_priv01.py on Windows since it is Unix specific
| * | | add call to superMark Benvenuto2014-12-171-0/+1
| | | |
| * | | Skip test_priv01.py on Windows since it is Unix specificMark Benvenuto2014-12-171-0/+4
| | | |
* | | | Merge pull request #1498 from markbenvenuto/test_fix_salvageAlex Gorrod2014-12-191-1/+2
|\ \ \ \ | |_|/ / |/| | | fix test_salvage.py on Windows
| * | | fix test_salvage on WindowsMark Benvenuto2014-12-181-1/+2
|/ / /
* | | Add a Windows-specific file, os_win/os_snprintf.c.Keith Bostic2014-12-172-1/+3
| | |
* | | Clarify that setting statistics_log.wait to non-zero value starts theKeith Bostic2014-12-172-6/+8
| | | | | | | | | | | | statistics logging.
* | | Stop maintaining a separate Windows version of the file list, generate itKeith Bostic2014-12-174-19/+61
| | | | | | | | | | | | from the POSIX list (it shouldn't change that often).
* | | Merge pull request #1493 from markbenvenuto/snprintf_fixAlex Gorrod2014-12-185-8/+53
|\ \ \ | | | | | | | | Windows - Implement a custom version of snprintf
View Lorry Controller Status