Xiang Li
d0dd205b0e
wal: add a bench for write entry
2014-11-20 12:07:17 -08:00
Xiang Li
b50f331558
Merge pull request #1744 from xiang90/next
...
raft: optimistically increase the next if the follower is already matched
2014-11-19 13:21:11 -08:00
Brandon Philips
12aaf046d7
Merge pull request #1752 from philips/include-etcd-migrate-in-releases
...
scripts: build-release add etcd-migrate
2014-11-19 12:13:21 -08:00
Brandon Philips
f08df9e0f3
Merge pull request #1753 from philips/fixup-build-docker
...
scripts: build-docker tag and use ENTRYPOINT
2014-11-19 12:13:08 -08:00
Brandon Philips
00df13138e
scripts: build-docker tag and use ENTRYPOINT
...
Use ENTRYPOINT so people can specify flags to etcd without providing the
binary.
Thanks to @hugod in IRC for pointing this out.
2014-11-19 12:11:27 -08:00
Brandon Philips
d2e36a9535
scripts: build-release add etcd-migrate
...
this tool is only temporary for the alphas but make sure it makes it in.
2014-11-19 12:10:06 -08:00
Brandon Philips
893fb3b890
Merge pull request #1747 from proppy/master
...
create .godir
2014-11-18 16:00:07 -08:00
Johan Euphrosine
03bacc1984
create .godir
2014-11-18 15:01:57 -08:00
Xiang Li
4c1fd07311
raft: optimistically increase the next if the follower is already matched
...
This is useful since we want to pipeline the appendEntry requests. Without
enabling optimistic increasing, the second pipelining appendEntry request
will include the entries the first one has already sent out. We decrease
the next directly to match if the leader receives a rejection for a matched
follower. This happens if one pipelining request get lost and following ones
arrives at the follower.
2014-11-18 13:41:38 -08:00
Xiang Li
e07ef6991c
Merge pull request #1740 from xiang90/handleheartbeat
...
raft: add handleHeartbeat
2014-11-18 13:25:12 -08:00
Yicheng Qin
f94ff96c69
Merge pull request #1739 from yichengq/230
...
rafthttp: send takes raft message instead of bytes
2014-11-18 10:11:05 -08:00
Xiang Li
bd4cfa2a07
raft: add handleHeartbeat
...
handleHeartbeat commits to the commit index in the message. It never decreases the
commit index of the raft state machine.
2014-11-18 08:34:06 -08:00
Yicheng Qin
1a72143ecb
rafthttp: send takes raft message instead of bytes
...
This gives streaming mechanism the chance to assemble and disassemble
raft messages.
2014-11-17 22:39:53 -08:00
Xiang Li
6cac631a0d
Merge pull request #1738 from yichengq/231
...
*: add rafthttp pkg into test list
2014-11-17 17:14:40 -08:00
Yicheng Qin
04d416291a
*: add rafthttp pkg into test list
2014-11-17 16:45:54 -08:00
Xiang Li
1635844ebd
Merge pull request #1737 from xiang90/include_commit
...
raft: include commitIndex in heartbeat
2014-11-17 16:26:07 -08:00
Xiang Li
b93d87f17f
raft: include commitIndex in heartbeat
2014-11-17 16:19:28 -08:00
Yicheng Qin
a2c568a144
Merge pull request #1669 from yichengq/215
...
*: add rafthttp as a separate package
2014-11-17 16:14:59 -08:00
Yicheng Qin
f24e214ee5
rafthttp: move server stats in raftHandler to etcdserver
2014-11-17 16:02:20 -08:00
Yicheng Qin
5dc5f8145c
*: etcdhttp.raftHandler -> rafthttp.RaftHandler
2014-11-17 15:52:24 -08:00
Yicheng Qin
3fcc011717
etcdserver: rename sender.go -> sendhub.go
2014-11-17 15:35:15 -08:00
Yicheng Qin
84fbf7aab5
*: etcdserver.sender -> rafthttp.Sender
2014-11-17 15:35:10 -08:00
Xiang Li
e04e4632b3
Merge pull request #1736 from xiang90/verify
...
etcdserver: refactor ValidateClusterAndAssignIDs
2014-11-17 14:34:59 -08:00
Xiang Li
0541f0afa0
etcdserver: refactor ValidateClusterAndAssignIDs
2014-11-17 14:23:37 -08:00
Xiang Li
144db790ae
Merge pull request #1729 from xiang90/test_rm
...
integration: add integration test for remove member
2014-11-17 13:35:26 -08:00
Xiang Li
c26de66262
integration: add integration test for remove member
2014-11-17 13:28:09 -08:00
Xiang Li
ee8fbee7ab
Merge pull request #1733 from xiang90/config
...
integration: add test for member restart
2014-11-17 13:19:59 -08:00
Xiang Li
800747e1cf
integration: add test for member restart
2014-11-17 11:28:18 -08:00
Xiang Li
a817ca705b
Merge pull request #1726 from xiang90/fix_sender
...
etcdserver: add buffer to the sender queue
2014-11-14 16:13:32 -08:00
Yicheng Qin
fd512ffe6d
version: bump to alpha.3
2014-11-14 15:51:01 -08:00
Xiang Li
7c4b84a6cd
etcdserver: add buffer to the sender queue
2014-11-14 15:18:16 -08:00
Xiang Li
ac5a282003
Merge pull request #1720 from xiang90/sender_stop
...
*: gracefully stop etcdserver
2014-11-14 14:16:39 -08:00
Xiang Li
8bf71d796e
*: gracefully stop etcdserver
2014-11-14 14:12:24 -08:00
Barak Michener
4e251f8624
Merge pull request #1578 from barakmich/bcm_migrate
...
migrate: Add a migration tool to go from etcd v0.4 -> v0.5 data directories
2014-11-14 16:58:22 -05:00
Barak Michener
192f200d9e
Fix up migration tool, add snapshot migration
...
Fixes all updates since bcwaldon sketched the original, with cleanup and
into an acutal working state. The commit log follows:
fix pb reference and remove unused file post rebase
unbreak the migrate folder
correctly detect node IDs
fix snapshotting
Fix previous broken snapshot
Add raft log entries to the translation; fix test for all timezones. (Still in progress, but passing)
Fix etcd:join and etcd:remove
print more data when dumping the log
Cleanup based on yichengq's comments
more comments
Fix the commited index based on the snapshot, if one exists
detect nodeIDs from snapshot
add initial tool documentation and match the semantics in the build script and main
formalize migration doc
rename function and clarify docs
fix nil pointer
fix the record conversion test
add migration to test suite and fix govet
2014-11-14 16:46:08 -05:00
Brian Waldon
5ea1f2d96f
etcd4: migration from v0.4 -> v0.5
2014-11-14 15:57:26 -05:00
Brian Waldon
c36abeabd1
etcdserver: export Member.StoreKey
2014-11-14 15:57:26 -05:00
Yicheng Qin
b6887e4a83
Merge pull request #1719 from yichengq/228
...
etcdserver: recover snapshot before applying committed entries
2014-11-14 12:18:41 -08:00
Yicheng Qin
77433ff6da
etcdserver: recover cluster when receiving newer snapshot
2014-11-14 12:11:21 -08:00
Yicheng Qin
dfaa7290c4
etcdserver: check and select committed entries to apply
2014-11-14 12:11:16 -08:00
Yicheng Qin
f6a7f96967
etcdserver: recover from snapshot before applying requests
2014-11-14 12:08:39 -08:00
Yicheng Qin
7d0ffb3f12
raft: not set applied when restored from snapshot
...
applied is only updated by application level through Advance.
2014-11-14 12:08:39 -08:00
Xiang Li
1f7198855b
Merge pull request #1703 from jonboulle/proc
...
*: fix Procfile
2014-11-14 08:39:08 -08:00
Xiang Li
6f7fd89ba2
Merge pull request #1718 from yichengq/226
...
sender: support elegant stop
2014-11-13 21:29:02 -08:00
Yicheng Qin
12dba7d413
sender: support elegant stop
2014-11-13 17:44:36 -08:00
Xiang Li
e66bda957b
Merge pull request #1714 from xiang90/stop
...
StopNotify
2014-11-13 15:16:52 -08:00
Xiang Li
6a1fe00615
Merge pull request #1704 from xiang90/print_config
...
*: print out configuration when necessary
2014-11-13 14:35:50 -08:00
Yicheng Qin
11f392bdc8
Merge pull request #1708 from yichengq/223
...
etcdserver: validate new node is not registered before in best effort
2014-11-13 14:30:40 -08:00
Xiang Li
b5d480f17a
etcdserver: add StopNotify
2014-11-13 14:16:48 -08:00
Xiang Li
978d0f1ca1
etcdserver: fix TestDoProposalStopped test
...
We start etcd server in this test without the cluster. Sometimes it panics when
accessing the cluster. Most of the time it does not panic, since we can stop the
server fast enough before applying the first configuration change entry.
2014-11-13 14:08:59 -08:00
