810a5146dd
Merge pull request #1635 from jonboulle/doc
...
etcdserver: add docstrings for confchanges
2014-11-07 10:20:05 -08:00
5055863e09
etcdserver: add docstrings for confchanges
2014-11-07 10:19:55 -08:00
bf47fe7cac
Merge pull request #1647 from xiangli-cmu/force_cluster
...
etcdserver: force new cluster
2014-11-07 10:15:53 -08:00
9d19429993
Merge pull request #1609 from yichengq/202
...
etcdserver: refactor sender
2014-11-07 10:12:02 -08:00
142dfc7d88
etcdserver: add doc for getIDset
2014-11-07 09:00:58 -08:00
0a9c6164af
etcdserver: add support for force cluster
2014-11-07 08:49:01 -08:00
376268391b
Merge pull request #1646 from jonboulle/1536_disco_proxy
...
discovery: add command line flag for discovery-proxy
2014-11-07 08:32:23 -08:00
d6f37ec9ad
Merge pull request #1648 from jonboulle/delete_member_404
...
etcdhttp: return 404 when removing nonexistent member
2014-11-06 17:24:17 -08:00
ca1b30db10
etcdhttp: return 404 when removing nonexistent member
2014-11-06 16:59:40 -08:00
9454d30854
etcdserver: add sendHub tests
2014-11-06 16:49:13 -08:00
f75e56932a
Merge pull request #1643 from jonboulle/fix_flags
...
pkg: fix SetFlagsFromEnv behaviour
2014-11-06 16:43:21 -08:00
5604b4c57c
flag: split out SetFlagsFromEnvBad test; declare return error
2014-11-06 16:40:13 -08:00
8f1885a398
discovery: add command line flag for discovery-proxy
2014-11-06 16:35:24 -08:00
ccded6644a
Merge pull request #1505 from yichengq/193
...
etcdserver: refactor non-blocking check for sync tests
2014-11-06 15:48:18 -08:00
321d65c4ac
pkg: fix SetFlagsFromEnv behaviour
...
This function was fundamentally buggy, as a panic could be trivially
triggered by setting the wrong environment variable (e.g.
ETCD_BIND_ADDR=foo). Instead, let's propagate the error and present it
to the user in a cleaner way.
2014-11-06 14:39:30 -08:00
c5e6053fcd
Merge pull request #1638 from xiangli-cmu/better_logging
...
etcdserver: better logging for clusterFromPeerURLs
2014-11-06 14:33:53 -08:00
eb0d80767e
etcdserver: better logging for clusterFromPeerURLs
2014-11-06 14:28:07 -08:00
6fa031fa69
Merge pull request #1641 from bdarnell/remove-raftlog-reset
...
raft: remove raftLog.resetUnstable and resetNextEnts
2014-11-06 14:26:08 -08:00
21987c8701
raft: remove raftLog.resetUnstable and resetNextEnts
...
These methods are no longer used outside of tests and are redundant with
the new stableTo and appliedTo methods.
2014-11-06 17:18:00 -05:00
457b30e585
etcdserver: add/remove sender in sendhub explicitly
2014-11-06 14:04:14 -08:00
2138163c61
etcdserver: code clean on sender struct
2014-11-06 14:04:14 -08:00
211c5e3e29
etcdserver: fix data race in Cluster struct
2014-11-06 14:04:14 -08:00
c3b0de943c
etcdserver: discard messages if sender reaches max serving
...
It is the correct thing to do to ensure that the communication is full
of out-of-date messages.
It results in that integration testing is very easy to throw MsgProp away,
and makes client wait until 5 min timeout. Sync interval and heartbeat are
increased to alleviate the traffic.
2014-11-06 14:04:14 -08:00
1e05cd75c7
etcdserver: refactor sender
...
1. restrict the number of inflight connections to remote member
2. support stop
2014-11-06 14:04:14 -08:00
087e0e8b62
Merge pull request #1636 from xiangli-cmu/client
...
client: add error handling for addmember
2014-11-06 12:46:15 -08:00
b65dd84e1a
Merge pull request #1632 from jonboulle/cs_flag
...
etcdmain: use StringsFlag for initialclusterstate
2014-11-06 12:36:22 -08:00
66572561bf
client: add error handling for addmember
2014-11-06 12:31:24 -08:00
b53a98eb38
Merge pull request #1631 from xiangli-cmu/validate_doc
...
Validate doc
2014-11-06 11:45:00 -08:00
a1f5df22ad
doc: document conflict case when adding a member
2014-11-06 11:16:49 -08:00
04f6208ace
etcdmain: use StringsFlag for initialclusterstate
2014-11-06 11:13:24 -08:00
3cb885c6b2
etcdhttp: return 409 instead of 412 when there is a conflict when adding a member
2014-11-06 11:07:25 -08:00
9c8f9b3560
Merge pull request #1585 from coreos/clean-up-other-apis-docs
...
docs: clean up other apis
2014-11-06 10:48:02 -08:00
4ed60471fe
Merge pull request #1627 from xiangli-cmu/validate_peer_url
...
etcdserver: validate peerurl when adding members
2014-11-06 10:43:22 -08:00
7d28d80e5a
Merge pull request #1626 from jonboulle/proxy_stuff
...
discovery: simplify interface
2014-11-06 10:09:16 -08:00
45d7ef99c4
Merge pull request #1629 from asmundg/x-fix-typo
...
Fix typo in docs
2014-11-06 09:58:54 -08:00
0d8345e0c1
Fix typo in docs
...
Suggesting that users add two nodes with the same name is probably not a
good idea.
2014-11-06 10:49:40 +01:00
2760739ceb
Merge pull request #1625 from yichengq/205
...
docs: describe the lifetime of discovery url
2014-11-06 00:32:49 -08:00
5d755bd54a
docs: describe the lifetime of discovery url
2014-11-06 00:31:19 -08:00
bd2b18b6de
etcdserver: validate peerurl when adding members
2014-11-05 23:12:48 -08:00
68bca981de
discovery: simplify interface
...
There's no real need to expose a Discoverer interface/struct when the
only use of the interface (and indeed the module) is to invoke a single
function. This isn't Java, after all. So instead, simplify to Discovery
exposing just two functions: JoinCluster (i.e. what was formerly called
"discovery"), and GetCluster (hitherto "ProxyDiscovery")
2014-11-05 22:45:01 -08:00
6fdbb086f4
Merge pull request #1623 from xiangli-cmu/valid_configuration
...
Valid configuration
2014-11-05 18:13:04 -08:00
99b1af40c6
etcdserver: move config validation to cluster
2014-11-05 17:55:07 -08:00
99bb479a60
Merge pull request #1618 from yichengq/203
...
etcdserver: improve panic message in Cluster
2014-11-05 17:14:26 -08:00
98406af448
cluster: separate out membersFromStore from newClusterFromStore
2014-11-05 15:56:43 -08:00
6c9169b4f4
etcdserver: improve panic message in Cluster
2014-11-05 15:39:28 -08:00
3fc6f9c24f
Merge pull request #1586 from xiangli-cmu/fix_node
...
*: add Advance interface to raft.Node
2014-11-05 15:09:51 -08:00
0d7c43d885
*: add a Advance interface to raft.Node
...
Node set the applied to committed right after it sends out Ready to application. This is not
correct since the application has not actually applied the entries at that point. We add a
Advance interface to Node. Application needs to call Advance to tell raft Node its progress.
Also this change can avoid unnecessary copying when application is still applying entires but
there are more entries to be applied.
2014-11-05 15:04:14 -08:00
c5140d5c18
Merge pull request #1614 from yichengq/194
...
*: handle panic and fatal organizedly
2014-11-05 14:08:35 -08:00
fdb82718e0
Merge pull request #1612 from jonboulle/proxy
...
proxy: add docstrings
2014-11-05 13:56:24 -08:00
791b2fd503
*: handle panic and fatal more consistently
...
1. etcd fatals if there is critical error in the system and operator should
do something for it
2. etcd panics if there happens something unexpected, and it should be
reported to us to debug.
2014-11-05 13:53:24 -08:00
Jonathan Boulle