99dcc8c | Brandon Philips | 02 June 2014, 22:25:03 UTC | chore(server): bump back to 0.4.2 | 02 June 2014, 22:25:03 UTC |
3d2523e | Yicheng Qin | 02 June 2014, 22:12:05 UTC | Merge pull request #825 from unihorn/98 fix(multi_node_kill_all_and_recovery_test): ensure cluster is up | 02 June 2014, 22:12:05 UTC |
25e69d9 | Yicheng Qin | 02 June 2014, 21:43:51 UTC | fix(multi_node_kill_all_and_recovery_test): ensure cluster is up | 02 June 2014, 21:43:51 UTC |
707174b | Brandon Philips | 02 June 2014, 21:19:52 UTC | chore(server): bump to 0.4.2+git | 02 June 2014, 21:19:52 UTC |
ce92cc3 | Brandon Philips | 02 June 2014, 21:17:38 UTC | feat(CHANGELOG): bump to v0.4.2 | 02 June 2014, 21:17:38 UTC |
5bfbf3a | Yicheng Qin | 02 June 2014, 21:12:08 UTC | Merge pull request #824 from unihorn/97 fix(remove_node_test): remove unnecessary cluster configuration | 02 June 2014, 21:12:08 UTC |
e04a188 | Yicheng Qin | 02 June 2014, 19:00:29 UTC | fix(remove_node_test): remove unnecessary cluster configuration The cluster configuration operation is originally to make sure the instance won't be added back automatically between removal and check for the number of existing peer-mode instances. But this could make some node removed before the removal command. Use longer sync interval instead to avoid this problem. | 02 June 2014, 20:30:19 UTC |
a51fda3 | Brandon Philips | 02 June 2014, 19:06:00 UTC | Merge pull request #822 from philips/add-notes-about-discovery docs(cluster-discovery): add caution to use old discovery endpoint | 02 June 2014, 19:06:00 UTC |
ca44801 | Yicheng Qin | 06 May 2014, 21:33:48 UTC | docs(cluster-discovery): add caution to use old discovery endpoint | 02 June 2014, 18:34:56 UTC |
2387ef3 | Yicheng Qin | 02 June 2014, 18:04:07 UTC | Merge pull request #819 from unihorn/97 fix(server): joinIndex is not set after recovery from full outage | 02 June 2014, 18:04:07 UTC |
d5bfca9 | Yicheng Qin | 02 June 2014, 17:38:36 UTC | Merge pull request #814 from unihorn/91 fix(server/v2): set correct content-type for etcdError response | 02 June 2014, 17:38:36 UTC |
7cb1269 | Yicheng Qin | 31 May 2014, 17:42:31 UTC | fix(simple_snapshot_test): enlarge reasonable index range | 31 May 2014, 17:42:31 UTC |
444e017 | Yicheng Qin | 31 May 2014, 17:32:03 UTC | fix(remove_node_test): ensure cluster config is activated | 31 May 2014, 17:32:03 UTC |
356675b | Yicheng Qin | 31 May 2014, 17:15:03 UTC | fix(multi_node_kill_all_and_recovery_test): ensure cluster running | 31 May 2014, 17:15:03 UTC |
d776863 | Yicheng Qin | 31 May 2014, 17:03:39 UTC | fix(server): set joinIndex when recovered | 31 May 2014, 17:03:39 UTC |
37796ed | Yicheng Qin | 31 May 2014, 01:37:36 UTC | tests: add TestMultiNodeKillAllAndRecorveryAndRemoveLeader This one breaks because it doesn't set joinIndex correctly. | 31 May 2014, 17:01:45 UTC |
f007cf3 | Yicheng Qin | 31 May 2014, 01:36:58 UTC | Merge pull request #818 from unihorn/96 fix(standby_server): able to join the cluster containing itself | 31 May 2014, 01:36:58 UTC |
ca29691 | Yicheng Qin | 31 May 2014, 01:36:23 UTC | tests(standby_test): comments | 31 May 2014, 01:36:23 UTC |
4bebb53 | Yicheng Qin | 30 May 2014, 20:53:41 UTC | fix(standby_server): able to join the cluster containing itself Standby server will switch to peer server if it finds that it has been contained in the cluster. | 30 May 2014, 21:03:49 UTC |
c27db1e | Brandon Philips | 30 May 2014, 20:45:12 UTC | Merge pull request #816 from unihorn/95 docs(clustering): limit for peer-address changing | 30 May 2014, 20:45:12 UTC |
a5fc1d2 | Brandon Philips | 30 May 2014, 20:41:32 UTC | Merge pull request #817 from cholcombe973/master Adding autodock into the libraries and tools section | 30 May 2014, 20:41:32 UTC |
1df0b94 | Chris Holcombe | 30 May 2014, 20:20:28 UTC | Adding autodock into the libraries and tools section | 30 May 2014, 20:20:28 UTC |
3a71eb9 | Rob Szumski | 30 May 2014, 19:26:07 UTC | Merge pull request #808 from robszumski/update-optimal-size fix(docs): add information about standbys | 30 May 2014, 19:26:07 UTC |
001cceb | Rob Szumski | 21 May 2014, 18:44:30 UTC | fix(docs): update doc with standby info | 30 May 2014, 19:23:22 UTC |
98ff4af | Yicheng Qin | 30 May 2014, 15:50:16 UTC | docs(clustering): limit for peer-address changing | 30 May 2014, 15:50:16 UTC |
db4c5e0 | Yicheng Qin | 29 May 2014, 21:18:50 UTC | fix(server/v2): set correct content-type for etcdError response "net/http".Error reset the content type, so we get rid of it and write our own one. | 29 May 2014, 21:18:50 UTC |
b3c5ed6 | Brandon Philips | 22 May 2014, 16:50:40 UTC | chore(pkg/btrfs): remove accidental swp file. | 22 May 2014, 16:50:40 UTC |
22c944d | Brandon Philips | 21 May 2014, 03:55:57 UTC | chore(server): bump 0.4.0+git | 21 May 2014, 03:55:57 UTC |
a2d16b5 | Brandon Philips | 21 May 2014, 03:46:46 UTC | chore(server): bump to 0.4.1 | 21 May 2014, 03:46:46 UTC |
b637b3a | Brandon Philips | 21 May 2014, 03:41:21 UTC | Merge pull request #806 from philips/add-machine-api-docs fix(server): add user facing remove API | 21 May 2014, 03:41:21 UTC |
0eba3c9 | Brandon Philips | 21 May 2014, 03:25:34 UTC | feat(Documentation): document the entire admin machines API Flesh out this document a bit more completely. | 21 May 2014, 03:25:34 UTC |
c3aab42 | Brandon Philips | 21 May 2014, 03:07:47 UTC | fix(Documentation): update based on standby refactor These docs were not updated after the refactoring of the standy mode. Fix that now. | 21 May 2014, 03:07:47 UTC |
62560f9 | Brandon Philips | 21 May 2014, 03:01:10 UTC | fix(server): add user facing remove API This was accidently removed as we refactored the standy stuff. Re-add this user facing remove endpoint that matches the config endpoints. | 21 May 2014, 03:01:10 UTC |
3c04f8b | Brandon Philips | 21 May 2014, 00:52:48 UTC | fix(scripts/build-release): put everything into a release dir | 21 May 2014, 00:53:17 UTC |
cc37c58 | Brandon Philips | 21 May 2014, 00:10:28 UTC | chore(server): bump to 0.4.0+git | 21 May 2014, 00:10:28 UTC |
07d1eb0 | Brandon Philips | 21 May 2014, 00:08:17 UTC | chore(server): bump to 0.4.0 | 21 May 2014, 00:09:22 UTC |
9a006d6 | Blake Mizerany | 20 May 2014, 23:23:33 UTC | doc(README): add note about 32bit systems - fixes 358 | 20 May 2014, 23:23:33 UTC |
6860bfd | Xiang Li | 20 May 2014, 21:37:53 UTC | Merge pull request #805 from xiangli-cmu/fix_test fix(test/remove_node_test.go) fix a deadlock in the test | 20 May 2014, 21:37:53 UTC |
aaedf32 | Xiang Li | 20 May 2014, 21:34:59 UTC | fix(test/remove_node_test.go) fix a deadlock in the test The go-etcd client waits for the response from the paused node. And the test waits for the reponse to continue. Actually we do not even need that small test, since we will check the machine status afterwards. | 20 May 2014, 21:34:59 UTC |
16d89c9 | Xiang Li | 20 May 2014, 21:32:45 UTC | Merge pull request #804 from xiangli-cmu/fix_curr_index_race fix(store): synchronize access to CurrentIndex | 20 May 2014, 21:32:45 UTC |
516ebdb | Xiang Li | 20 May 2014, 20:53:47 UTC | fix(store): synchronize access to CurrentIndex | 20 May 2014, 20:53:47 UTC |
1e7a7b1 | Xiang Li | 20 May 2014, 20:37:14 UTC | Merge pull request #799 from xiangli-cmu/deny_unknow_peer hack(server): notify removed peers when they try to become candidates | 20 May 2014, 20:37:14 UTC |
4c953a7 | Xiang Li | 20 May 2014, 20:36:16 UTC | Merge pull request #803 from unihorn/94 fix(peer_server): set store and registry when setting raft server | 20 May 2014, 20:36:16 UTC |
934c28d | Yicheng Qin | 20 May 2014, 20:12:12 UTC | fix(peer_server): set store and registry when setting raft server New raft server needs new store and registry. | 20 May 2014, 20:12:12 UTC |
0c8b13c | Xiang Li | 20 May 2014, 18:06:00 UTC | Merge pull request #19 from unihorn/92 tests(remove_node): add TestRemovePausedNode | 20 May 2014, 18:06:00 UTC |
9e5b12f | Yicheng Qin | 20 May 2014, 17:35:43 UTC | tests(remove_node): add TestRemovePausedNode | 20 May 2014, 18:01:14 UTC |
189fece | Xiang Li | 20 May 2014, 17:11:04 UTC | hack(server): notify removed peers when they try to become candidates A peer might be removed during a network partiton. When it comes back it will not have received any of the log entries that would have notified it of its removal and go onto propose a vote. This will disrupt the cluster and the cluster should give the machine feedback that it is no longer a member. The term of a denied vote is MaxUint64. The notification of the removal is a raft event. These two modification are quick heck. In reaction to this notification the machine should shutdown. In this case the shutdown just moves it towards becoming a standby server. | 20 May 2014, 17:17:32 UTC |
ad27aa0 | Brandon Philips | 20 May 2014, 16:37:28 UTC | Merge pull request #791 from yifan-gu/ipv6_zone_fix Ipv6 zone fix | 20 May 2014, 16:37:28 UTC |
dffe460 | Brandon Philips | 20 May 2014, 16:30:11 UTC | Merge pull request #793 from binocarlos/patch-1 Added the -peer-heartbeat-interval option | 20 May 2014, 16:30:11 UTC |
b4e4bf4 | Yifan Gu | 18 May 2014, 08:38:06 UTC | fix(config.go) return a valid URL struct from sanitizeURL() pass the URL struct above to sanitizeBindAddr() Since url.Parse() will return an error when parsing an already-parsed ipv6 url string, (e.g. [http://[fe80::6203:8ff:fe9e:ace%25eth0]:7001), so I just return the valid URL struct from sanitizeURL() and send it to sanitizeBindAddr(), then there is no need to parse it again in sanitizeBindAddr(). Besides, for IPV6 url, the percent sign should be escaped, see: http://en.wikipedia.org/wiki/IPv6_address#Link-local_addresses_and_zone_indices | 18 May 2014, 19:42:47 UTC |
e948216 | Kai Davenport | 18 May 2014, 18:22:48 UTC | Added the -peer-heartbeat-interval option | 18 May 2014, 18:22:48 UTC |
cbcf55d | Brandon Philips | 18 May 2014, 17:09:07 UTC | Merge pull request #744 from unihorn/71 docs(api): add Cluster Config API | 18 May 2014, 17:09:07 UTC |
1084e51 | Brandon Philips | 18 May 2014, 17:08:52 UTC | Merge pull request #786 from unihorn/91 feat(standby_server): write cluster info to disk | 18 May 2014, 17:08:52 UTC |
ad9155c | Brandon Philips | 17 May 2014, 23:15:35 UTC | Merge pull request #789 from bass3m/master Add Julia etcd client | 17 May 2014, 23:15:35 UTC |
84f71b6 | Yicheng Qin | 16 May 2014, 22:07:49 UTC | chore(standby_server): remove error return because standby server should be started in best efforts. | 16 May 2014, 22:07:49 UTC |
d2d68f9 | Bassem Youssef | 16 May 2014, 17:11:31 UTC | Add Julia etcd client | 16 May 2014, 17:11:31 UTC |
71679bc | Yicheng Qin | 16 May 2014, 05:00:07 UTC | feat(standby_server): make atomic move for file to avoid the risk of writing out a corrupted file. | 16 May 2014, 05:00:07 UTC |
a824be4 | Yicheng Qin | 16 May 2014, 04:10:15 UTC | feat(standby_server): save/load Running into disk | 16 May 2014, 04:10:15 UTC |
35cc81e | Yicheng Qin | 16 May 2014, 03:57:58 UTC | feat(standby_server): save/load syncInterval to disk | 16 May 2014, 03:57:58 UTC |
716496e | Yicheng Qin | 16 May 2014, 03:18:59 UTC | chore(standby_server): still sleep for the first time | 16 May 2014, 03:18:59 UTC |
e5ce4fc | Yicheng Qin | 16 May 2014, 02:47:12 UTC | docs(standbys): clarify pseudocode | 16 May 2014, 02:47:12 UTC |
9940347 | Jonathan Boulle | 16 May 2014, 01:30:15 UTC | Merge pull request #788 from alex-sherwin/patch-1 fix(doc): fixed typo for etcd lock examples | 16 May 2014, 01:30:15 UTC |
d2a2cc1 | Alex Sherwin | 16 May 2014, 01:27:50 UTC | fixed typo for etcd lock examples | 16 May 2014, 01:27:50 UTC |
b7d9fdb | Yicheng Qin | 15 May 2014, 06:38:22 UTC | feat(standby_server): write cluster info to disk For better fault tolerance and availability. | 15 May 2014, 11:47:15 UTC |
7cf8a4a | Brandon Philips | 14 May 2014, 17:03:03 UTC | Merge pull request #779 from unihorn/89 feat: implement standby mode | 14 May 2014, 17:03:03 UTC |
8510263 | Yicheng Qin | 14 May 2014, 14:13:05 UTC | chore(standby_server): let syncInterval represent in second unit This is done to keep consistency with other namings. | 14 May 2014, 14:13:05 UTC |
fc77b3e | Yicheng Qin | 14 May 2014, 02:28:28 UTC | fix(simple_snapshot_test): enlarge reasonable index range | 14 May 2014, 02:28:28 UTC |
f6591b9 | Yicheng Qin | 14 May 2014, 02:16:45 UTC | chore(standby): minor changes based on comments | 14 May 2014, 02:19:52 UTC |
403f709 | Yicheng Qin | 13 May 2014, 16:45:13 UTC | chore(cluster_config): set default timeout to 5s Or the leader death could let the standbys down for a rather long time. | 13 May 2014, 20:13:44 UTC |
cbb706c | Yicheng Qin | 13 May 2014, 15:08:03 UTC | bump(goraft/raft): c76c5d95 | 13 May 2014, 15:08:03 UTC |
c0027bf | Yicheng Qin | 13 May 2014, 02:42:18 UTC | feat(cluster_config): change field from int to float64 This is modified for better flexibility, especially for testing. | 13 May 2014, 02:42:18 UTC |
735ffb3 | Yicheng Qin | 09 May 2014, 22:51:48 UTC | Merge pull request #782 from unihorn/90 fix(TestV1Watch): ensure server has started | 09 May 2014, 22:51:48 UTC |
6a64141 | Yicheng Qin | 09 May 2014, 22:42:18 UTC | fix(TestV1Watch): ensure server has started | 09 May 2014, 22:42:18 UTC |
5367c1c | Yicheng Qin | 09 May 2014, 22:38:03 UTC | chore(standby): minor changes based on comments | 09 May 2014, 22:38:03 UTC |
c6b1a73 | Yicheng Qin | 09 May 2014, 21:55:16 UTC | feat(option): add cluster config option It will be used when creating a brand-new cluster. | 09 May 2014, 22:22:11 UTC |
6d4f018 | Yicheng Qin | 09 May 2014, 20:28:21 UTC | chore(cluster_config): rename SyncClusterInterval to SyncInterval for better naming | 09 May 2014, 20:28:21 UTC |
765cd5d | Yicheng Qin | 09 May 2014, 09:08:50 UTC | refactor(find_cluster): make it simpler | 09 May 2014, 09:27:04 UTC |
baadf63 | Yicheng Qin | 09 May 2014, 02:47:19 UTC | feat: implement standby mode Change log: 1. PeerServer - estimate initial mode from its log through removedInLog variable - refactor FindCluster to return the estimation - refactor Start to call FindCluster explicitly - move raftServer start and cluster init from FindCluster to Start - remove stopNotify from PeerServer because it is not used anymore 2. Etcd - refactor Run logic to fit the specification 3. ClusterConfig - rename promoteDelay to removeDelay for better naming - add SyncClusterInterval field to ClusterConfig - commit command to set default cluster config when cluster is created - store cluster config info into key space for consistency - reload cluster config when reboot 4. add StandbyServer 5. Error - remove unused EcodePromoteError | 09 May 2014, 08:56:55 UTC |
5bd08a3 | Yicheng Qin | 30 April 2014, 02:40:40 UTC | docs(standbys): specification for standby module | 08 May 2014, 22:02:55 UTC |
f1c13e2 | Yicheng Qin | 08 May 2014, 21:08:38 UTC | Merge pull request #774 from unihorn/83 feat(join): check cluster conditions before join | 08 May 2014, 21:08:38 UTC |
6c950ea | Yicheng Qin | 08 May 2014, 21:02:09 UTC | Merge pull request #772 from unihorn/81 feat(peer_server): stop service when removed | 08 May 2014, 21:02:09 UTC |
5c7a963 | Yicheng Qin | 08 May 2014, 20:20:46 UTC | chore(peer_server): adjust code to make it more clear | 08 May 2014, 20:20:46 UTC |
c92231c | Yicheng Qin | 08 May 2014, 20:17:51 UTC | Merge branch 'master' of github.com:coreos/etcd Conflicts: server/peer_server_handlers.go | 08 May 2014, 20:17:51 UTC |
e960a0e | Yicheng Qin | 08 May 2014, 20:15:10 UTC | chore(client): minor changes based on comments The changes are made on error handling, comments and constant. | 08 May 2014, 20:15:10 UTC |
015d228 | Yicheng Qin | 08 May 2014, 19:28:44 UTC | Merge pull request #763 from unihorn/77 fix(raft_server_stats): set startTime when init | 08 May 2014, 19:28:44 UTC |
e31249f | Yicheng Qin | 08 May 2014, 19:26:56 UTC | Merge pull request #778 from unihorn/88 fix(TestV2Watch): ensure server has started | 08 May 2014, 19:26:56 UTC |
b3e66ee | Yicheng Qin | 08 May 2014, 19:12:26 UTC | fix(TestV2Watch): ensure server has started | 08 May 2014, 19:18:08 UTC |
af33d61 | Yicheng Qin | 08 May 2014, 18:53:46 UTC | Merge pull request #775 from unihorn/84 refactor(tests/server_utils): use etcd instance | 08 May 2014, 18:53:46 UTC |
bc4a98c | Yicheng Qin | 08 May 2014, 18:53:26 UTC | Merge pull request #776 from unihorn/85 feat(peer_server): add State field to machineMessage | 08 May 2014, 18:53:26 UTC |
fb17ccf | Yicheng Qin | 08 May 2014, 17:54:50 UTC | chore(cors): more explanation for flush in ServeHTTP | 08 May 2014, 17:54:50 UTC |
04f09d2 | Yicheng Qin | 07 May 2014, 20:09:25 UTC | feat(peer_server): add State field to machineMessage State field indicates the state of each machine. For now, its value could be follower or leader. | 08 May 2014, 17:25:39 UTC |
6450d6d | Yicheng Qin | 08 May 2014, 16:11:04 UTC | Merge pull request #777 from unihorn/86 fix(registry): fetch peers from store instead of cache | 08 May 2014, 16:11:04 UTC |
0558b54 | Yicheng Qin | 07 May 2014, 21:33:30 UTC | fix(registry): fetch peers from store instead of cache The current cache implmentation may contain removed machines, so we fetch peers from store for correctness. | 08 May 2014, 15:44:32 UTC |
5465201 | Yicheng Qin | 07 May 2014, 23:31:17 UTC | chore(peer_server): more explanation for asyncRemove | 07 May 2014, 23:31:17 UTC |
ae81f84 | Yicheng Qin | 07 May 2014, 23:09:08 UTC | refactor(client): remove useless logic in redirection | 07 May 2014, 23:09:08 UTC |
c9ce14c | Yicheng Qin | 07 May 2014, 19:18:32 UTC | chore(peer_server): set client transporter separately It also moves the hack on timeout from raft transporter to client transporter. | 07 May 2014, 20:26:05 UTC |
bed20b7 | Yicheng Qin | 07 May 2014, 19:51:41 UTC | chore(peer_server): add more function description | 07 May 2014, 19:51:41 UTC |
206881b | Yicheng Qin | 07 May 2014, 19:44:48 UTC | fix(peer_server): check running status before start/stop This makes peer server more robust. | 07 May 2014, 19:44:48 UTC |
cf25650 | Yicheng Qin | 07 May 2014, 19:39:07 UTC | fix(etcd): exit 0 when removed | 07 May 2014, 19:39:07 UTC |
7dce4c8 | Yicheng Qin | 02 May 2014, 00:59:35 UTC | refactor(tests/server_utils): use etcd instance Remove duplicated etcd start code. | 07 May 2014, 18:49:03 UTC |