Hawkin
/
arangodb
mirror of https://gitee.com/bigwinds/arangodb
1
0
Fork 0
Code Issues Projects Releases Wiki Activity
33,720 Commits 358 Branches 495 Tags 994 MiB
Commit Graph

102 Commits

Author SHA1 Message Date
Michael Hackstein a3ba4b249d DBServer collections now use a cluster-wide unique id for their cids. Better handling of waiting for replication. 2016-09-08 16:26:17 +02:00
jsteemann 2df53c92ce Merge branch 'devel' of https://github.com/arangodb/arangodb into generic-col-types 2016-09-08 09:51:24 +02:00
Kaveh Vahedipour 4e9f455094 too verbose messages from cluster.js 2016-09-08 07:38:10 +00:00
Max Neunhoeffer 549367a15e Fix error check in tryLaunchTask. 2016-09-07 09:46:59 +02:00
Max Neunhoeffer c83c5200cf Fix behaviour of tryLaunchJob when database is dropped. 2016-09-06 16:29:52 +02:00
Max Neunhoeffer c083693ef0 Better error reporting in tryLaunchTask. 2016-09-06 14:28:00 +02:00
Michael Hackstein 513f6ce560 Now all AGENTs do re port that they are not part of a cluster. This did confuse the statistics.js 2016-08-31 17:20:28 +02:00
Andreas Streichardt 5579eca2df Change debug levels for shard sync...no more spam 2016-08-23 10:53:14 +02:00
Andreas Streichardt 6ee1f6d14a Reformat code 2016-06-30 16:57:17 +02:00
Max Neunhoeffer 3ee7ff4aa9 Fix for synchronisation of shards logic. 2016-06-29 11:59:26 +02:00
Max Neunhoeffer 50f325b4cf Fix typos. 2016-06-18 00:02:02 -07:00
Max Neunhoeffer 67bd33152b Fix typo. 2016-06-17 23:58:58 -07:00
Max Neunhoeffer aae9e92dc0 Change logic when a sync shard job is scheduled. 2016-06-17 23:57:15 -07:00
Max Neunhoeffer 7563cf56d0 Fix sync repl fix. 2016-06-17 23:18:28 -07:00
Max Neunhoeffer fe0cb0a958 Another fix for shard synchronisation. 2016-06-17 23:16:21 -07:00
hkernbach 70ad0867d4 shardDistribution api, shards view 2016-06-17 11:10:11 +02:00
Max Neunhoeffer 2bb8e4441d Sort out waiting for sync. repl. to settle at bootstrap and create DB. 
This makes startup and creating databases slower, but leads to
a lot less trouble later on. It can now no longer happen that
a collection or database is dropped before synchronous replication
has settled.
 2016-06-15 15:25:05 +02:00
Max Neunhoeffer 5153b230b8 Wait for sync repl to settle after db create. 2016-06-14 16:20:18 +02:00
Max Neunhoeffer 85d64d0eae Do not retry for too long to synchronize a shard. 2016-06-14 13:54:53 +02:00
Max Neunhoeffer b6ba8f841c Lower info level of some warning. 2016-06-14 10:51:15 +02:00
Max Neunhoeffer 9c74d6a076 Properly protect traffic in Sync Keyspace. 2016-06-14 10:49:28 +02:00
Max Neunhoeffer 79025546f7 Improve moveShard tests and testing infrastructure. 2016-06-13 23:28:32 +02:00
Max Neunhoeffer 1f926fc24b Fix synchronous replication: use shard name instead of ID. 
Explanation: Different replicas of the same shard will have different
local collection IDs. Therefore we have to use the shard name
(collection name) to indentify collections in synchronous replication.
 2016-06-13 13:54:32 +02:00
Max Neunhoeffer 794a08dd86 Finish first version of rebalanceShards. 2016-06-10 15:56:25 +02:00
Max Neunhoeffer 3a8412090d rebalanceShards. 2016-06-10 14:41:46 +02:00
Max Neunhoeffer f9fa8ee39f Please jslint. 2016-06-10 12:14:15 +02:00
Max Neunhoeffer f2dfbf7448 Change format of /_admin/cluster/shardDistribution. 2016-06-10 12:12:55 +02:00
Max Neunhoeffer 6b7b54cf03 Better handling of shutdown situation for shard sync. 2016-06-10 12:04:09 +02:00
Max Neunhoeffer 059b24d7e7 Start work on rebalancing of shards. 2016-06-10 00:13:42 +02:00
Max Neunhoeffer ebe0b5c64a Report result of DBSyncServerJob back to the heartbeat thread. 2016-06-09 12:03:34 +02:00
jsteemann 0d5e91e804 fixed jslint 2016-06-08 16:58:15 +02:00
Max Neunhoeffer 49cd6ca9f4 Fix leaderResign (no double resign) and leader change. 2016-06-08 15:54:00 +02:00
Max Neunhoeffer d48f5506f2 Fix shard removeShardFollower functionality. 2016-06-08 14:59:39 +02:00
Max Neunhoeffer 659a06765d Let a follower unfollow a leader. 2016-06-08 10:00:19 +02:00
Max Neunhoeffer c85bd2fdbe Implement leader resignation. 2016-06-07 23:26:34 +02:00
Max Neunhoeffer fe5b9f9c23 Remove shard entries in Agency less eagerly. 
This fixes a bug about collections vanishing in Current.
 2016-06-07 23:01:56 +02:00
Max Neunhoeffer 5f14a4eaa5 Fix JS code using asyncRequest. 2016-06-06 15:21:36 +02:00
Max Neunhoeffer 19747c8b6c Fix crash due to wrong usage of asyncRequest. 2016-06-06 11:30:39 +02:00
Jan Steemann 0b17568365 fix jslint 2016-05-31 10:17:05 +02:00
Max Neunhoeffer 1538bb7cb0 Bugfix for sync. repl. 2016-05-31 01:18:05 -06:00
Max Neunhoeffer eaacae8c2f Take out a debugging message. 2016-05-25 22:29:12 +02:00
Max Neunhoeffer 4a4472e31e Repair DBServerAgencySync in background. 
Also: Notice role change follower->leader.
 2016-05-25 17:27:27 +02:00
Max Neunhoeffer 0227e47153 Fix shard sync jobs in background. 
Also fix bug that indexes and properties in followers are updated.
 2016-05-25 11:55:07 +02:00
Max Neunhoeffer 37851116ae Better error reporting in syncShards. 2016-05-25 08:05:28 +02:00
Max Neunhoeffer d801510ac5 Run shard sync in background task. 2016-05-24 23:28:22 +02:00
Max Neunhoeffer 94cd868deb Revert "Run shard sync in background task." 
This reverts commit a8b68b79eb.
 2016-05-24 13:16:51 +02:00
Max Neunhoeffer a8b68b79eb Run shard sync in background task. 2016-05-24 12:01:06 +02:00
Max Neunhoeffer 31bff28489 Factor out synchronizeOneShard. 2016-05-24 09:43:59 +02:00
Max Neunhoeffer 717d110a1d More fixes in DBServerAgencySync js part. 2016-05-23 22:54:21 +02:00
Max Neunhoeffer 7dec299c3a Improve error messages. 2016-05-23 20:40:48 +02:00