Eliot Horowitz
|
e521301d18
|
make forcing a split work better SERVER-2419 180?
|
2011-03-06 20:18:32 -05:00 |
|
Eliot Horowitz
|
13608bb283
|
null is ok in shard key, blank is not SERVER-2648
|
2011-03-01 13:21:26 -05:00 |
|
Eliot Horowitz
|
ed6273ba49
|
make sure not to use memory in a cursor when yielding SERVER-2447
|
2011-03-01 10:51:48 -05:00 |
|
Eliot Horowitz
|
54175da517
|
splitVector perf
|
2011-02-01 18:03:00 -05:00 |
|
Eliot Horowitz
|
ea4585664e
|
add an assert in splitVector
|
2011-01-29 02:43:01 -05:00 |
|
Eliot Horowitz
|
1b68b2f53b
|
fix some split logging
|
2011-01-29 01:25:24 -05:00 |
|
Alberto Lerner
|
95883498e5
|
SERVER-2060 allow an index to be validated for sharding
|
2011-01-10 09:12:34 -08:00 |
|
Eliot Horowitz
|
b828d21630
|
ran astyle SERVER-2304
|
2011-01-04 00:40:41 -05:00 |
|
Alberto Lerner
|
0472e3396a
|
splitChunk entry message to like like moveChunk's
|
2010-12-29 11:32:18 -05:00 |
|
Alberto Lerner
|
acb40fb684
|
SERVER-2306 split request now carries expected shard information
|
2010-12-29 10:53:02 -05:00 |
|
Alberto Lerner
|
5e5ed8bb03
|
message warns explicitly that operation aborted
|
2010-12-28 14:54:29 -05:00 |
|
Alberto Lerner
|
83b6d04f0d
|
put likely cause on message
|
2010-12-28 12:53:14 -05:00 |
|
Alberto Lerner
|
84a30c24e9
|
log if moveChunk/splitChunk did not pass admission checks
|
2010-12-27 14:30:54 -05:00 |
|
Alberto Lerner
|
1fa6403ef6
|
make message consistent
|
2010-12-27 08:54:36 -05:00 |
|
Alberto Lerner
|
b9c944466f
|
hook new versioning machinery and allow mongod to maintain per-chunk state
|
2010-12-04 14:49:09 -05:00 |
|
Alberto Lerner
|
c8fd4df699
|
disambiguate
|
2010-12-02 13:09:27 -05:00 |
|
Alberto Lerner
|
b08dcf6e6d
|
fix potential seg fault
|
2010-11-23 09:25:24 -05:00 |
|
Alberto Lerner
|
ab7fa6504c
|
log command start
|
2010-11-22 08:28:38 -05:00 |
|
Alberto Lerner
|
f9d7cdab63
|
ChunkMatcher now has per-chunk boundaries
|
2010-11-18 17:59:20 -05:00 |
|
Alberto Lerner
|
910044f28a
|
splits are now done by the mongod side
|
2010-10-27 17:13:47 -04:00 |
|
Alberto Lerner
|
ba7cfbbd8c
|
fix memory ownership
|
2010-10-26 18:24:55 -04:00 |
|
Alberto Lerner
|
7e074d00e4
|
split chunk maybe the first time a sharded command hit mongod
|
2010-10-25 15:38:22 -04:00 |
|
Alberto Lerner
|
3a8d363fd4
|
tighter new splitChunk (misses factoring out common code to migrate path)
|
2010-10-22 19:15:48 -04:00 |
|
Alberto Lerner
|
4f1a41872f
|
add logging to splitChunk command in mongod
|
2010-10-22 16:44:37 -04:00 |
|
Alberto Lerner
|
9266002d96
|
splitVector can now be 'force'-d to split a small chunk
|
2010-10-21 14:11:48 -04:00 |
|
Alberto Lerner
|
1e450f7b00
|
do not pick a split key if it is not different from the chunks' min key (on behalf of Matt Taylor)
|
2010-10-19 13:12:28 -04:00 |
|
Alberto Lerner
|
41f97ca850
|
Initial mongod-side split logic. Not yet hooked up.
|
2010-10-18 15:44:56 -04:00 |
|
Alberto Lerner
|
2ad85ece41
|
Prepare initial move of split logic to the mongod side
|
2010-10-14 20:35:24 -04:00 |
|
dwight
|
c1b68a13de
|
towards durability
|
2010-09-27 12:35:22 -04:00 |
|
Alberto Lerner
|
f5d055953c
|
SERVER-1807 better logging and tighter test
|
2010-09-17 17:02:28 -04:00 |
|
Alberto Lerner
|
d55c4af161
|
SERVER-1807 quicker auto-split if in a jumbo chunk case
|
2010-09-16 18:21:59 -04:00 |
|
Eliot Horowitz
|
85cdf5bfed
|
yield on medianKey
|
2010-09-15 22:04:39 -04:00 |
|
Eliot Horowitz
|
e70006635a
|
fix potential inf. loop in splitVector
|
2010-09-13 12:51:34 -04:00 |
|
Eliot Horowitz
|
5a4bfedb88
|
yield in splitVector so even if it takes a while won't block SERVER-1779
|
2010-09-13 12:50:33 -04:00 |
|
Eliot Horowitz
|
2418da510e
|
use splitVector instead of datasize and mediankey. 1 command vs. 2 and should be faster and more extensible.
still some code cleanup to do probably
SERVER-1550 SERVER-1779
|
2010-09-12 18:23:27 -04:00 |
|
Eliot Horowitz
|
e172d7d74f
|
tiny bit of cleaning for splitVector
|
2010-09-12 17:51:39 -04:00 |
|
Eliot Horowitz
|
c6bae28ff6
|
cleaning split
|
2010-08-24 01:58:21 -04:00 |
|
Eliot Horowitz
|
06917cf3bc
|
safety in split
|
2010-08-24 01:53:54 -04:00 |
|
Alberto Lerner
|
9e951e98b4
|
SERVER-1545 splitVector now takes ranges.
|
2010-08-23 22:53:14 -04:00 |
|
Alberto Lerner
|
2af7fa477c
|
SERVER-1545 Leave new chunks half-full instead of at 90%.
|
2010-08-23 20:00:14 -04:00 |
|
Alberto Lerner
|
4bcf64d3d1
|
SERVER-1545 Add a fast path to splitVector command.
|
2010-08-23 19:42:14 -04:00 |
|
Eliot Horowitz
|
0de85ad5c2
|
safety in median cmd
|
2010-07-21 00:04:07 -04:00 |
|
Alberto Lerner
|
d0d79d0fbf
|
Fix warnings
|
2010-07-19 09:40:15 -04:00 |
|
Aaron
|
c2f1a2318d
|
merge
|
2010-07-16 05:10:42 -07:00 |
|
Aaron
|
5371a5e448
|
SERVER-802 range / in skipping
|
2010-07-16 01:49:15 -07:00 |
|
Mathias Stearn
|
cfcc881427
|
compound shard keys are almost done SERVER-951
|
2010-07-14 20:08:14 -04:00 |
|
Aaron
|
4798e64b1c
|
SERVER-371 checkpoint
|
2010-07-07 12:26:20 -07:00 |
|
Alberto Lerner
|
480e89e582
|
Fix timer for 'dataSize', 'median', and 'splitVector' commands
|
2010-07-03 14:20:18 -04:00 |
|
Alberto Lerner
|
8672d8855e
|
SERVER-1185 maxChunkSize unit is MBs
|
2010-07-03 12:46:02 -04:00 |
|
Alberto Lerner
|
3d27b6bd3f
|
SERVER-1185 Initial tests for splitVector
|
2010-07-02 16:37:47 -04:00 |
|