| Commit message (Expand) | Author | Age | Files | Lines |
... | |
* | Two small changes in page-acquisition handling of forced-eviction: only | Keith Bostic | 2015-02-21 | 1 | -8/+18 |
* | Whitespace, use WT_PAGE_IS_INTERNAL instead of enumerating leaf page | Keith Bostic | 2015-02-18 | 1 | -9/+3 |
* | Combine the various checks for whether a page can be evicted into one place. | Michael Cahill | 2015-02-17 | 1 | -12/+6 |
* | Check if a page was recently split before doing forced eviction. We used to ... | Michael Cahill | 2015-02-11 | 1 | -6/+15 |
* | Once we decide to force-evict a page, do it directly rather than setting read... | Michael Cahill | 2015-02-02 | 1 | -2/+2 |
* | Use reads to measure cache pressure with shared caches. We previously tracke... | Michael Cahill | 2015-01-30 | 1 | -0/+1 |
* | If we fail to force evict a page, sleep to give another thread a chance. Oth... | Michael Cahill | 2015-01-14 | 1 | -1/+7 |
* | Don't attempt to force evict pages during checkpoints: that is guaranteed to ... | Michael Cahill | 2015-01-14 | 1 | -2/+5 |
* | Cache a maximum of 2MB per session for scratch buffers (configurable via an u... | Michael Cahill | 2015-01-05 | 1 | -1/+1 |
* | whitespace | Keith Bostic | 2015-01-04 | 1 | -1/+1 |
* | Copyright notices: add MongoDB, update to 2015. | Keith Bostic | 2015-01-04 | 1 | -0/+1 |
* | Start sleeping for 1ms, not 2ms | Michael Cahill | 2014-12-31 | 1 | -1/+1 |
* | Use WT_MIN not WT_MAX. | Susan LoVerso | 2014-12-30 | 1 | -1/+1 |
* | Yield 1000 times instead of 5 to reduce latency impact. | Michael Cahill | 2014-12-24 | 1 | -2/+2 |
* | busy-work: sort the stat options, make the stat options units match | Keith Bostic | 2014-12-19 | 1 | -9/+11 |
* | Add sleep backoff if waiting for page too long. SERVER-16269 | Susan LoVerso | 2014-12-19 | 1 | -6/+16 |
* | Add separate statistics for pages yielding because the WT_REF isn't | Keith Bostic | 2014-12-19 | 1 | -3/+11 |
* | Most WiredTiger calloc calls are to allocate a single structure, | Keith Bostic | 2014-12-15 | 1 | -2/+2 |
* | Move the READGEN_OLDEST logic into a function to make it less obscure. | Michael Cahill | 2014-12-05 | 1 | -2/+2 |
* | We don't yet need anything other than single-byte versions of the | Keith Bostic | 2014-10-04 | 1 | -1/+1 |
* | Add size specific atomics for 1, 4 & 8 byte atomic operations | Mark Benvenuto | 2014-10-03 | 1 | -2/+2 |
* | __value is a keyword in Visual Studio .NET. | Keith Bostic | 2014-10-01 | 1 | -1/+2 |
* | Split the WT_READ_NO_GEN flag to clarify: sometimes we mean "this read should... | Michael Cahill | 2014-08-15 | 1 | -1/+1 |
* | Add flags to page release, respect WT_READ_NO_GEN to avoid forced eviction. | Michael Cahill | 2014-08-13 | 1 | -1/+1 |
* | Update function naming to be consistent with our style. | Alex Gorrod | 2014-07-28 | 1 | -3/+3 |
* | __wt_eviction_force_check() is only called from one place, there's no | Keith Bostic | 2014-07-25 | 1 | -1/+39 |
* | Encode the location/length of simple values in the WT_ROW pointer, along | Keith Bostic | 2014-07-07 | 1 | -0/+8 |
* | List the cases in sorted order. | Keith Bostic | 2014-07-01 | 1 | -1/+1 |
* | Convert row-store fast on-page key processing from per-page to per key. | Keith Bostic | 2014-06-30 | 1 | -45/+17 |
* | Experimental change to avoid unpacking cells in order to access | Keith Bostic | 2014-05-30 | 1 | -10/+49 |
* | The disk header in a page is read-only: mark it const. | Michael Cahill | 2014-04-30 | 1 | -10/+10 |
* | Change the block manager to not touch the WT_ITEM.{mem,memsize} fields | Keith Bostic | 2014-04-28 | 1 | -2/+5 |
* | Fix a comment. | Keith Bostic | 2014-04-22 | 1 | -6/+1 |
* | Fix the problem where a truncated/deleted page might split (meaning | Keith Bostic | 2014-04-09 | 1 | -26/+28 |
* | Obscure the name of the internal page's index so code more naturally | Keith Bostic | 2014-03-31 | 1 | -12/+14 |
* | Bug when deepening the tree in a split, didn't connect the newly created | Keith Bostic | 2014-03-30 | 1 | -4/+4 |
* | Experimental change: move the WT_PAGE.{parent,ref_hint} fields into the | Keith Bostic | 2014-03-29 | 1 | -11/+17 |
* | A problem in the new-split branch is WT_REF structures remain with the | Keith Bostic | 2014-03-26 | 1 | -52/+51 |
* | Add a comment that we don't Huffman encode internal page keys. | Keith Bostic | 2014-03-20 | 1 | -0/+4 |
* | Generalize overflow keys to look like overflow values, that is, set the | Keith Bostic | 2014-03-18 | 1 | -4/+0 |
* | Another run at handling removed overflow objects. Michael suggested we could | Keith Bostic | 2014-03-18 | 1 | -36/+0 |
* | Try to fix the two problems with the original overflow work in the | Keith Bostic | 2014-03-17 | 1 | -2/+43 |
* | Remove __wt_eviction_force entirely, just set the page read-gen in | Keith Bostic | 2014-03-12 | 1 | -1/+1 |
* | We can't copy/move the WT_REF structures to the new child pages when | Keith Bostic | 2014-03-05 | 1 | -18/+22 |
* | Merge branch 'develop' into new-split | Keith Bostic | 2014-03-05 | 1 | -13/+20 |
|\ |
|
| * | [mq]: evict-walk-review | Michael Cahill | 2014-03-05 | 1 | -8/+9 |
| * | Remove the code associated with WT_REF_EVICT_WALK. | Michael Cahill | 2014-02-27 | 1 | -6/+7 |
| * | Switch to using hazard references for walks through the cache. | Michael Cahill | 2014-02-27 | 1 | -3/+8 |
* | | Merge branch 'develop' into new-split | Keith Bostic | 2014-02-26 | 1 | -9/+5 |
|\ \
| |/ |
|
| * | Allow some internal pages to be evicted without eviction getting aggressive. | Michael Cahill | 2014-02-25 | 1 | -9/+5 |