4d0f474034
*: fix TestTransportErrorc
...
CI can be slow. We should just wait longer.
2016-05-11 10:09:40 -07:00
98dbdd5fbb
*: simplify network metrics
2016-05-09 22:37:12 -07:00
c5af1d7a88
rafthttp: fix race on peer status activeSince
2016-05-04 11:48:16 -07:00
8291110049
rafthttp: do not create new connections after stopping transport
2016-04-28 10:10:52 -07:00
67645095e9
*: add debugging metrics
2016-04-26 09:52:56 -07:00
fb137f11c5
rafthttp: make TestPipelineKeepSendingWhenPostError reliable
2016-04-20 09:38:47 -07:00
641a1a66e1
*: fix govet -shadow in go tip
2016-04-15 07:39:52 -07:00
f2d558644d
rafthttp: fix comment in msgappv2
2016-04-12 10:14:06 -07:00
6e6d64fb9b
*: clean up unused vars, functions
...
With help from https://github.com/dominikh/go-unused .
IsNetTimeoutError seems useful, so moved to pkg/netutil.
2016-04-06 21:33:55 -07:00
c09f23c46d
*: clean up bool comparison
2016-04-02 18:27:54 -07:00
a42d1dc1fe
*: drain http.Response.Body before closing
2016-03-30 09:35:47 -07:00
d533c14881
Merge pull request #4876 from heyitsanthony/integration-races
...
*: fix races from clientv3/integration tests
2016-03-29 09:10:53 -07:00
0c137b344b
rafthttp: do not block on proposal
2016-03-28 21:40:12 -07:00
813afc3d11
rafthttp: fix race between AddRemote and Send
2016-03-28 16:08:18 -07:00
70a9391378
*: enable v3 by default
2016-03-23 17:01:36 -07:00
900a61b023
*: http and https on the same port
2016-03-23 10:28:38 -07:00
bd832e5b0a
*: migrate Godeps to vendor/
2016-03-22 17:10:28 -07:00
21b33de810
rafthttp: drop go1.4 tests
2016-03-18 18:46:11 -07:00
5d256b7b86
rafthttp: pause peer should also pause its reader
2016-03-17 09:43:02 -07:00
afa0368dcc
*: fix godoc bugs in interfaces and slice fields
...
detected with goword
2016-02-24 00:45:40 -08:00
c5b51946eb
*: exported godoc fixups
2016-02-21 20:36:44 -08:00
2a3cacb60c
rafthttp: remove unncessary go routine
2016-02-18 07:57:58 -08:00
e4f22cd6d8
rafthtt: smart batching
...
Improved the overall performance more than 20% under heavyload
with little latency impact
heavy load
```
Requests/sec: ~23200 vs Requests/sec: ~31500
Latency distribution:
10% in 0.0883 secs.
25% in 0.1022 secs.
50% in 0.1207 secs.
75% in 0.1460 secs.
90% in 0.1647 secs.
95% in 0.1783 secs.
99% in 0.2223 secs.
vs
Latency distribution:
10% in 0.1119 secs.
25% in 0.1272 secs.
50% in 0.1469 secs.
75% in 0.1626 secs.
90% in 0.1765 secs.
95% in 0.1863 secs.
99% in 0.2276 secs.
```
Similar on light load too.
2016-02-17 13:17:12 -08:00
74382f56fb
rafthttp: handle short case in if statement
2016-02-16 19:26:51 -08:00
d393102e24
rafthttp: refactor
2016-02-16 19:21:53 -08:00
11d3e9ac69
rafthttp: better comment for streamWriter
2016-02-16 19:21:06 -08:00
56318f5433
rafthttp: add necessary locking
2016-02-16 19:18:05 -08:00
b72a0788ad
Merge pull request #4439 from xiang90/fix_tr
...
transport: make tr stop safe
2016-02-08 11:25:03 -08:00
ea688c1f06
transport: make tr stop safe
2016-02-08 11:13:52 -08:00
fe7cfe4d3d
rafthttp: plumb local peer URLs through transport
2016-02-05 16:44:41 -08:00
db0b505de5
rafthttp: add requester to transport if peer does not exist
...
cluster integration now supports adding members with stopped nodes, too
Fixes #3699
2016-02-03 14:16:46 -08:00
20461ab11a
*: fix many typos
2016-01-31 21:42:39 -08:00
96d82b40fb
rafthttp: fix typo in test comment
2016-01-29 15:59:36 +09:00
b6077f9d57
*: fix minor typos
2016-01-14 01:28:29 -08:00
54a45ba2f5
*: fix typo
2016-01-06 16:17:02 +08:00
e1bf726bc1
*: split out etcdserver's test mockup objects to live in interfaces' packages
2016-01-05 09:39:13 -08:00
d07a9cd893
Merge pull request #4059 from xiang90/snap_log
...
rafthttp: better snapshot sending logging
2015-12-28 10:05:35 -08:00
4be152bb4f
rework
2015-12-23 16:21:16 -08:00
ab31ba0d29
*: fix snapshot sending cycle
2015-12-23 13:58:57 -08:00
74dba2d4cf
rafthttp: better snapshot sending logging
...
snapshot sending is an important event. We should always log it explicitly.
2015-12-23 12:36:07 -08:00
94da4b9ee5
rafthttp: log before receiving snapshot
...
Database snapshot can be as large as 5GB. It is reasonable
to log before receiving it. Or the user might not know what
is happening and why etcd starts to use IO intensively.
2015-12-23 08:45:36 -08:00
80541d74d0
rafhttp: make TestStreamWriterAttachOutgoingConn more robust
2015-12-14 17:06:38 -08:00
7d78e0c85e
rafthttp: remove the unncessary TODO
...
The issue is not caused by this code, but by reading snapshot
from disk. etcd assumes the snapshot of v3 kv should live in
memory. If not, etcd does not work well anyway.
2015-12-11 09:37:41 -08:00
95c29838e3
rafthttp: move ReadCloser to ioutil
2015-12-11 09:37:41 -08:00
2fc3320e59
rafthttp: kill the receiving body timeout TODO in snapshot sender
2015-12-10 22:42:17 -08:00
23bd60ccce
*: rewrite snapshot sending
2015-12-08 18:21:21 -08:00
6aa559f93d
rafthttp: more detailed logs when filled-up buffer
...
This adds more detailed explanation about why some messages got dropped to help
users debug. This is for https://github.com/coreos/etcd/issues/3806 .
2015-11-18 14:17:43 -08:00
81229dbea9
*: add missing package descriptions
...
This adds and updates package descriptions in etcd projects.
And also deletes some duplicate LICENSE statements.
2015-11-17 20:54:10 -08:00
ff36b9d9bc
Merge pull request #3700 from xiang90/metrics_hi
...
Replace Summary with Histogram for all metrics
2015-11-10 10:06:45 -08:00
964f6050ee
raft: use HistogramVec for message_sent_latency
2015-11-10 10:05:32 -08:00
Xiang Li