3829 Commits

Author SHA1 Message Date
Xiang Li
e23f9e76d1 raft: do not applysnapshot in raft 2014-11-26 10:59:13 -08:00
Xiang Li
8de98d4903 raft: clean up 2014-11-25 16:21:50 -08:00
Xiang Li
9bd1786fe4 raft: memory storage does not append out of date entries 2014-11-25 15:18:40 -08:00
Xiang Li
9df0e7715d raft: do not panic on out of date compaction 2014-11-25 15:14:39 -08:00
Xiang Li
01cbcce8ba etcdserver: do not applySnapshot twice 2014-11-25 14:53:49 -08:00
Xiang Li
74d8c7f457 etcdserver: cleanup main loop 2014-11-25 14:38:18 -08:00
Yicheng Qin
7e6e305c4f Merge branch 'log_interface'
Conflicts:
	raft/raft.go
2014-11-25 14:22:11 -08:00
Yicheng Qin
a13d5a70ff etcdserver: save snapshot before entries 2014-11-25 12:39:15 -08:00
Yicheng Qin
dd57c1f189 Merge pull request #1784 from yichengq/234
raft: not compact log if the compact index < first index of the log
2014-11-25 11:52:01 -08:00
Yicheng Qin
4b43824be9 raft: not compact log if the compact index < first index of the log
It should ignore the compact operation instead of panic because the case that
the log is restored from snapshot before executing compact is reasonable.
2014-11-25 11:51:20 -08:00
Yicheng Qin
1a5333e51d Merge pull request #1778 from yichengq/232
docs: add configuration doc
2014-11-25 11:16:44 -08:00
Yicheng Qin
07ca99f4d6 docs: add configuration doc 2014-11-25 11:09:29 -08:00
Xiang Li
aa2721e31d Merge pull request #1781 from orls/fix-backup-snapshots
Fix backup snapshots
2014-11-25 08:59:16 -08:00
Xiang Li
c46e30412e Merge pull request #1782 from orls/fix-forcenew-panic
etcdserver: Fix panic when restoring from backup (-force-new-cluster)
2014-11-25 08:58:32 -08:00
Owen Smith
dbb6a75e3f backup: Ensure dest snapshot dir always exists
Otherwise the SaveSnap operation fails
2014-11-25 16:57:18 +00:00
Owen Smith
c67b937d62 etcdserver: truncate WAL from correct index when forcing new cluster
When loading from a backup with a snapshot and WAL, the length of WAL entries
can be lower than the current index integer value, causing a panic when
slicing off uncommitted entries. This looks for WAL entries higher than
the current index before slicing.
2014-11-25 16:46:56 +00:00
Owen Smith
9974bf0291 backup: Add fatal if dest snapshot failed
Errors are currently silently swallowed
2014-11-25 14:08:20 +00:00
Yicheng Qin
8aa89dba3d raft: make if checking match the error in storage.Term 2014-11-25 00:52:13 -08:00
Yicheng Qin
8ee1bf31d6 raft: use IsEmptySnap to check the empty snapshot 2014-11-25 00:37:21 -08:00
Yicheng Qin
e466126510 raft: set snapshot to nil when it is saved 2014-11-25 00:30:22 -08:00
Yicheng Qin
e17bcd8932 raft: remove wont-fix TODO in ApplyConfChange 2014-11-25 00:10:44 -08:00
Yicheng Qin
85d0e2f130 raft: remove unused raftLog.isOutOfAppliedBounds 2014-11-25 00:07:55 -08:00
Yicheng Qin
1e0f87df8c raft: stricter checking in raftLog.slice 2014-11-25 00:05:00 -08:00
Yicheng Qin
1d01c8aa2d raft: remove unused raftLog.at function 2014-11-24 23:52:28 -08:00
Yicheng Qin
2c06a1d815 raft: not set applied when restore log from snapshot
applied is only updated by application level through Advance.
2014-11-24 23:37:47 -08:00
Yicheng Qin
0d200baf72 raft: refine raftLog.term 2014-11-24 23:27:57 -08:00
Yicheng Qin
7fcaca6d18 raft: simplify raftLog.lastIndex 2014-11-24 23:08:51 -08:00
Yicheng Qin
8670f4012b raft: remove useless line in raftLog.append 2014-11-24 22:42:55 -08:00
Yicheng Qin
239c8dd479 raft: add comment to newLog 2014-11-24 21:47:12 -08:00
Yicheng Qin
54e1237271 etcdserver: panic when snapshot on raft storage
Snapshot on raft storage should always succeed. If there is an error, it must
be internal fault and needs stack info to debug.
2014-11-24 21:22:49 -08:00
Yicheng Qin
1b038da18a etcdserver: init snapi when init appliedi 2014-11-24 21:19:30 -08:00
Yicheng Qin
bd9e93eeea etcdserver: remove finished TODO for raftStorage.Compact 2014-11-24 21:10:53 -08:00
Yicheng Qin
185d37c333 etcdserver: not load dummy entry from the wal 2014-11-24 20:51:04 -08:00
Xiang Li
e3cb3d640b Merge pull request #1780 from xiang90/fix_raft_lead
raft: always check leader changes in node run loop
2014-11-24 19:29:53 -08:00
Xiang Li
9455119968 raft: always check leader changes in node run loop 2014-11-24 19:07:10 -08:00
Yicheng Qin
d67eea4a7d Merge pull request #1779 from yichengq/233
rafthttp: limit the buffer for every read correctly
2014-11-24 17:44:25 -08:00
Yicheng Qin
61ce494386 rafthttp: limit the buffer for every read correctly 2014-11-24 17:39:39 -08:00
Xiang Li
65ad1f6ffd raft: attach Index to Entry in all tests 2014-11-24 17:13:47 -08:00
Xiang Li
10ebf1a335 raft: fix memoryStorage append 2014-11-24 16:36:59 -08:00
Xiang Li
2876c652ab raft: fix for go vet 2014-11-24 15:00:38 -08:00
Xiang Li
d69e4dbe6d etcdserver: initial index to 1 2014-11-24 14:57:08 -08:00
Xiang Li
453133977d etcdserver: save snapshot only if the index is greater than previous snap index 2014-11-24 14:47:59 -08:00
Xiang Li
4b7af29c37 etcdserver: fix TriggerSnap test.
Sleep for millisecond to allow the server to apply the first nop and
first put separately.
2014-11-24 14:47:49 -08:00
Yicheng Qin
cfb96de413 Merge pull request #1770 from yichengq/230
*: set read/write timeout for raft transport and listener
2014-11-24 14:01:57 -08:00
Yicheng Qin
1e797c1e38 rafthttp: limit the data size read from connection each time 2014-11-24 13:54:45 -08:00
Yicheng Qin
3e55834c38 *: set read/write timeout for raft transport and listener 2014-11-24 13:46:44 -08:00
Yicheng Qin
ad58122e3c pkg/transport: fix dialer typo 2014-11-24 13:35:46 -08:00
Yicheng Qin
400e573013 rafthttp: log start and stop of streaming 2014-11-24 13:35:46 -08:00
Xiang Li
62a8df304a raft: fix error message in TestLogRestore 2014-11-24 11:10:02 -08:00
Xiang Li
e8afdcfe0a raft: refactor testUnstableEnts 2014-11-24 10:40:38 -08:00