Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | BUMP 2.2.0r2.2.0 | Eliot Horowitz | 2012-08-28 | 1 | -1/+1 |
| | |||||
* | post 2.2.0-rc2 | Eliot Horowitz | 2012-08-23 | 1 | -1/+1 |
| | |||||
* | fix compilation for test binary | Eric Milkie | 2012-08-22 | 1 | -1/+1 |
| | |||||
* | SERVER-6825 convert updates to upserts on oplog application | Eric Milkie | 2012-08-22 | 4 | -7/+9 |
| | | | | | | | | | | | | | | During initial sync, you must apply updates as updates and upserts as upserts, since it is possible that an in-place update will move a document backwards and cause the cloner to miss it. We need to detect this situation on the secondary. However, during normal replication, we need all updates to be upserts. Consider the situation where the oplog is replayed (possibly due to a server crash). If an update and subsequent delete are played and then replyed, the reply will hit an error on the update because the document will not exist. Converting the update to an upsert will allow oplog application to proceed; while it may result in an incomplete document, this document will be deleted soon afterward. | ||||
* | BUMP 2.2.0-rc2 | Eliot Horowitz | 2012-08-22 | 1 | -1/+1 |
| | |||||
* | SERVER-6757 Store holdCursor in a ClientCursor::Holder to prevent a double ↵ | Aaron | 2012-08-21 | 1 | -4/+3 |
| | | | | free on failed yield recovery. | ||||
* | SERVER-6785 fix leaking parallel cursor on unsharded reads | Eliot Horowitz | 2012-08-21 | 1 | -0/+9 |
| | |||||
* | SERVER-6757 - need to clear holdCursor when collection dropped during yield | Eliot Horowitz | 2012-08-21 | 1 | -0/+1 |
| | |||||
* | SERVER-6759 - work around for libm loading on some platforms | Eliot Horowitz | 2012-08-21 | 2 | -0/+25 |
| | |||||
* | SERVER-6778 use log() instead of cout for Windows stack trace | Tad Marshall | 2012-08-21 | 1 | -7/+11 |
| | |||||
* | SERVER-6809 SERVER-6811 fix chunk bounds for prefix shard keys | Kevin Matulef | 2012-08-21 | 4 | -27/+30 |
| | |||||
* | SERVER-6814 - fix memory leak in BtreeCursor::make | Eliot Horowitz | 2012-08-21 | 1 | -4/+4 |
| | |||||
* | SERVER-6766 Properly serialize DocumentSourceGroup for transmission to shards. | Aaron | 2012-08-20 | 4 | -22/+67 |
| | |||||
* | SERVER-6795 Correctly optimize reverse sense ExpressionCompare to ↵ | Aaron | 2012-08-20 | 2 | -4/+370 |
| | | | | ExpressionFieldRange. | ||||
* | SERVER-6779 Correctly serialize ExpressionCoerceToBool | Mathias Stearn | 2012-08-20 | 2 | -5/+55 |
| | |||||
* | SERVER-6793 - parallel writer batch lock needs to be involved in ↵ | Eliot Horowitz | 2012-08-20 | 4 | -6/+46 |
| | | | | yield/tempRelease | ||||
* | post rc1 | Eliot Horowitz | 2012-08-14 | 1 | -1/+1 |
| | |||||
* | BUMP 2.2.0-rc1r2.2.0-rc1 | Eliot Horowitz | 2012-08-13 | 1 | -1/+1 |
| | |||||
* | Revert "Speed-up dur::filename()" | Mathias Stearn | 2012-08-10 | 1 | -16/+5 |
| | | | | | | | This reverts commit 066d794339538fb020ec0fe9a27be981ea75b0a9. Conflicts: src/mongo/db/dur_recover.cpp | ||||
* | Fix dur::fileName() | Mathias Stearn | 2012-08-10 | 1 | -2/+7 |
| | |||||
* | SERVER-4683 Fix build break on older SCons, redux. | Andy Schwerin | 2012-08-10 | 1 | -2/+2 |
| | |||||
* | SERVER-4683 Fix build break on older SCons. | Andy Schwerin | 2012-08-10 | 1 | -2/+5 |
| | |||||
* | Speed-up dur::filename() | Mathias Stearn | 2012-08-10 | 1 | -5/+11 |
| | |||||
* | SERVER-4683 Support the imported tcmalloc, ystem tcmalloc or the default ↵ | Andy Schwerin | 2012-08-10 | 5 | -3/+70 |
| | | | | | | | | system allocator library. Pass --allocator=tcmalloc (default on Linux) or --allocator=system (default elsewhere) to control which allocator is used, and --use-system-tcmalloc to use the system- installed tcmalloc instead of the one in the mongo source tree if you use --allocator=tcmalloc. | ||||
* | SERVER-4683 Config files for tcmalloc on Linux built by 10gen. | Andy Schwerin | 2012-08-10 | 3 | -0/+403 |
| | |||||
* | SERVER-4683 Import gperftools source files. | Andy Schwerin | 2012-08-10 | 194 | -0/+56663 |
| | |||||
* | SERVER-6743 Do not add perftest.exe to the distribution archive or install ↵ | Andy Schwerin | 2012-08-09 | 1 | -1/+2 |
| | | | | it to /usr/bin. | ||||
* | Make C++ driver call GLE after storing gridFS chunks, before calling ↵ | Spencer T Brody | 2012-08-09 | 1 | -0/+6 |
| | | | | | | filemd5. SERVER-6742 This ensures that there are no pending writebacks from the chunks being inserted. | ||||
* | Make handling of stale configs in ParallelSortClusteredCursor safer | Spencer T Brody | 2012-08-09 | 2 | -9/+12 |
| | | | | by only checking the result object on commands. SERVER-6685 SERVER-6684 | ||||
* | Revert "SERVER-6076 implement loopCommands for benchRun" | Siddharth Singh | 2012-08-09 | 2 | -6/+2 |
| | | | | This reverts commit 02ec29038e81588c29cd463fa5e6610c23f20fae. | ||||
* | Revert "SERVER-6738 exclude bench_test4 from parallel suite" | Siddharth Singh | 2012-08-09 | 1 | -1/+0 |
| | | | | This reverts commit ea9d4e5c7451b25ac92bb00fa634f97116fdbe50. | ||||
* | SERVER-6738 exclude bench_test4 from parallel suite | Siddharth Singh | 2012-08-09 | 1 | -0/+1 |
| | |||||
* | SERVER-6572 - use pthread specifics on os x for performance | Eliot Horowitz | 2012-08-09 | 2 | -7/+45 |
| | |||||
* | SERVER-6572 fix Windows warning | Tad Marshall | 2012-08-08 | 1 | -1/+1 |
| | | | | | | | Explicitly make the target of the bit shift 'long long' to avoid MSVC warning: "src\mongo\db\record.cpp(242) : warning C4334: '<<' : result of 32-bit shift implicitly converted to 64 bits (was 64-bit shift intended?)". | ||||
* | SERVER-6572 - fix windows | Eliot Horowitz | 2012-08-08 | 1 | -2/+2 |
| | |||||
* | SERVER-6572 - can't use enum on windows | Eliot Horowitz | 2012-08-08 | 1 | -16/+13 |
| | |||||
* | SERVER-6572 - don't use a class so we can use simple thread locals | Eliot Horowitz | 2012-08-08 | 1 | -74/+77 |
| | |||||
* | SERVER-6076 implement loopCommands for benchRun | Siddharth Singh | 2012-08-08 | 2 | -2/+6 |
| | |||||
* | SERVER-6718 initialize member variable to intended default | Eric Milkie | 2012-08-08 | 1 | -2/+4 |
| | |||||
* | SERVER-6572 use a thread local cache for better perforamnce in record caching | Eliot Horowitz | 2012-08-08 | 1 | -5/+126 |
| | | | | needs work on os x | ||||
* | Revert "SERVER-6572 use a thread local cache for better perforamnce in ↵ | Andy Schwerin | 2012-08-08 | 1 | -76/+2 |
| | | | | | | record caching" This reverts commit f7fdee2e22b37c20a2999fe579111f0fb1f5eb91. | ||||
* | Check result object of command for stale config in ↵ | Spencer T Brody | 2012-08-08 | 1 | -0/+11 |
| | | | | ParallelSortClusteredCursor. SERVER-6684 SERVER-6685 | ||||
* | SERVER-6572 use a thread local cache for better perforamnce in record caching | Eliot Horowitz | 2012-08-08 | 1 | -2/+76 |
| | | | | needs work on osx | ||||
* | SERVER-6718 new parameter to control replication index prefetching behavior | Eric Milkie | 2012-08-08 | 8 | -9/+127 |
| | | | | | | | | | | | | | | | | New parameter replIndexPrefetch. Settable via the command line or use setParameter() command on a running server. Only valid on a server started with --replSet parameter. This allows a user to disable the prefetching of all index pages prior to the application of a replicated op on a secondary. For most use cases, prefetching all indexes for a given op will provide good performance. For use cases that do many in-place updates to documents in a collection with many indexes that are unaffected by such updates, using the '_id_only' setting may provide better performance than the 'all' setting. The '_id_only' setting only prefetches index pages for the _id index, which will always be used to find the document needed to be updated. The 'none' setting will be used for debugging only; there should be no production use cases for this setting. getParameter() supports getting the current value of this parameter. | ||||
* | Merge pull request #282 from tychoish/clone-collection-warning | Ian Whalen | 2012-08-08 | 1 | -2/+1 |
|\ | | | | | SERVER-6731 cloneCollection warning correction | ||||
| * | SERVER-6731 cloneCollection warning correction | Sam Kleinman | 2012-08-08 | 1 | -2/+1 |
| | | |||||
* | | Merge pull request #283 from samantharitter/SERVER--6708 | Ian Whalen | 2012-08-08 | 1 | -12/+6 |
|\ \ | | | | | | | SERVER-6708 sh._adminCommand is silent and sh helpers return result obj | ||||
| * | | SERVER--6708 sh._adminCommand is silent, and sh helpers return result object | Samantha Ritter | 2012-08-08 | 1 | -12/+6 |
| | | | |||||
* | | | Merge pull request #280 from samantharitter/SERVER-6474 | Ian Whalen | 2012-08-08 | 1 | -0/+8 |
|\ \ \ | |_|/ |/| | | SERVER-6474 give better error messages when syncFrom is misused | ||||
| * | | SERVER-6474 give more descriptive error messages when syncFrom is misused | Samantha Ritter | 2012-08-01 | 1 | -0/+8 |
| |/ |