Commit Graph

74 Commits

Author SHA1 Message Date
Eliot Horowitz
357f336ff9 if a chunk is too big to migrate, mark it as jumbo and move on SERVER-2571 2011-09-21 21:14:55 -04:00
Eliot Horowitz
4f0d2349c7 fix error message on Balancer init 2011-08-20 01:08:24 -04:00
Mathias Stearn
0e20be567e use LOG(x) everywhere in mongos 2011-08-08 19:52:56 -04:00
Jens Rantil
2c8088f180 Fixing a bunch of spelling mistakes, and a few grammatical errors. No functionality has been changed. 2011-05-26 11:59:13 +02:00
Eliot Horowitz
1183afb5c6 don't sleep with lock held SERVER-3020 2011-04-29 10:54:12 -04:00
Eliot Horowitz
2b76720b72 ping config server even if not doing balancing round SERVER-3015 2011-04-28 16:37:58 -04:00
Eliot Horowitz
6db7015120 try to fix double free SERVER-2704 2011-03-17 14:50:38 -04:00
Eliot Horowitz
39942c950a use nextSafe 180 2011-03-16 11:14:28 -04:00
greg
737a20b8fa Fixes for distributed lock handling time skew.
In general, the current distributed lock is susceptible to time skew in config servers and mongos clients.  This update
uses an algorithm that checks for time skew, as well as providing some more robust tests to make sure a distributed
lock cannot be forced prematurely.  Some documentation on the website, currently hidden from non-10gen, gives more
info.
2011-03-01 14:15:26 -05:00
Eliot Horowitz
0f533dd181 don't take balancer lock if balancing is disable SERVER-2632 2011-02-27 13:35:32 -05:00
Eliot Horowitz
a4153eb6ed If migrate fails due to oversized chunk, try to split that chunk SERVER-2519 2011-02-16 15:51:25 -05:00
Eliot Horowitz
b828d21630 ran astyle SERVER-2304 2011-01-04 00:40:41 -05:00
Alberto Lerner
7512156d5a SERVER-1358 add a size beyond which a chunk migrate request should not go through 2010-12-17 00:59:55 -05:00
dwight
a774b9c5d3 unref local var 2010-12-13 20:07:40 -05:00
Alberto Lerner
2224288933 SERVER-2213 balancer tolerates servers being down at mongos startup 2010-12-13 16:11:16 -05:00
Alberto Lerner
d598524214 don't split after a balance move fails (a lock taken would cause it) and defer splitting jumbos to SERVER-1358 2010-10-27 16:42:02 -04:00
Alberto Lerner
86b6240226 better method name 2010-10-25 12:28:13 -04:00
Alberto Lerner
0b621a2079 all split requests now go through new logic 2010-10-22 09:38:36 -04:00
Alberto Lerner
b6fd57882e SERVER-1979 check max key before requesting a move 2010-10-20 16:19:56 -04:00
Alberto Lerner
0f3b16781c SERVER-1713 balancer doesn't move to a server with writebacks pending 2010-09-22 11:58:55 -04:00
Alberto Lerner
68df06feab SERVER-1713 if a shard has writebacks pending don't use it as a receiver 2010-09-21 10:43:28 -04:00
Alberto Lerner
a3986b4040 SERVER-1713 move testing code to proper place and adhere to dbtest coding practices 2010-09-21 09:53:34 -04:00
Eliot Horowitz
1c778d93ed abort chunk move if too big SERVER-1806 2010-09-16 16:57:08 -04:00
Alberto Lerner
98dcdbe9db better message 2010-09-12 16:59:57 -04:00
Alberto Lerner
b322f76107 Say hello even if in log level 0. 2010-08-10 17:55:21 -07:00
Alberto Lerner
bceacf2544 SERVER-1292 A bit more prep, separate grid.{h,cpp} from config 2010-07-28 14:24:55 -04:00
Alberto Lerner
a976d6ed88 SERVER-1280 Bring back the stop balancer marker 2010-07-21 19:56:18 -04:00
Eliot Horowitz
1d7c6c8b3a sleep on balancer error 2010-07-19 10:57:04 -04:00
Alberto Lerner
daf8c0fbd0 Revert "SERVER-1270 Prep for hooking StringData"
This reverts commit b7c9a9fc57.
2010-07-17 16:07:38 -04:00
Alberto Lerner
b7c9a9fc57 SERVER-1270 Prep for hooking StringData 2010-07-17 14:33:17 -04:00
Alberto Lerner
38eb41bf4a Balancer refreshes shard state at every round 2010-07-16 15:06:20 -04:00
Eliot Horowitz
e953d1482e fix balance sleep schedule 2010-07-13 22:45:22 -04:00
Alberto Lerner
13708352c2 More logging messages 2010-07-13 20:22:25 -04:00
Eliot Horowitz
130f70d3d4 use distlock for balancer SERVER-1354 2010-07-13 14:50:53 -04:00
Eliot Horowitz
453398b8b2 distlock pings, get rid of ourHostname 2010-07-13 00:56:08 -04:00
Eliot Horowitz
b19fb22ca1 Merge branch 'master' of github.com:mongodb/mongo 2010-07-08 18:46:01 -04:00
Eliot Horowitz
7a43996200 todo comment 2010-07-08 18:45:53 -04:00
Alberto Lerner
62d7b02c1e Better balancer messages 2010-07-08 12:09:05 -04:00
Alberto Lerner
4717034108 SERVER-1280 Allow balancer to be turned off 2010-07-06 19:58:37 -04:00
Alberto Lerner
e3e227686b SERVER-940 A bit more cleanup after shard was removed 2010-06-30 13:50:16 -04:00
Alberto Lerner
30b1524a77 SERVER-940 Balancer now drains chunks off a shard that's being removed 2010-06-23 19:25:03 -07:00
Eliot Horowitz
e1e111d05e didn't mean to compile that 2010-06-21 18:28:52 -04:00
Eliot Horowitz
a71cc3cd4b cleaning exceptions 2010-06-21 13:41:34 -04:00
Alberto Lerner
21bb8b439c Don't wait for an incoming query to load the shards info 2010-06-08 16:24:39 -04:00
Alberto Lerner
88c7f16767 Better balancer messages 2010-06-08 16:01:52 -04:00
Alberto Lerner
9f22a44e05 SERVER-950 addshard's maxSize option is in effect (carefull, no way to undo it yet). 2010-06-07 15:15:06 -04:00
Alberto Lerner
ef6a47ad09 SERVER-950 Balancer policy now picks up size limits (but balancer is not passing it yet) 2010-06-05 17:54:49 -04:00
Alberto Lerner
e9ac0d3a9f SERVER-950 Pass space usage and limits information the to balancer policy 2010-06-05 16:17:59 -04:00
Alberto Lerner
efe90867ca SERVER-950 Separate the space-based policy from the balancing mechanism [2/2] 2010-06-04 16:14:34 -04:00
Alberto Lerner
d7a8bfbf3a SERVER-950 Separate the space-based policy from the balancing mechanism [1/2] 2010-06-04 14:39:12 -04:00