Commit Graph

1813 Commits (9ab35726622c1babbe95e458f5f561b6be8a1113)

Author SHA1 Message Date
Sam Batschelet 1411c585be etcdserver: fix typo in log message
Signed-off-by: Sam Batschelet <sbatsche@redhat.com>
2019-05-10 09:54:00 -04:00
shivaramr 9150bf52d6 go modules: Fix module path version to include version number 2019-04-26 15:29:50 -07:00
Jingyi Hu cca0d5c1be
Merge pull request #10672 from nolouch/fix-probing-log
api/rafthttp: fix the probing status log print
2019-04-24 23:05:23 -07:00
nolouch decc0d5f43 api/rafthttp: fix the probing status print
Signed-off-by: nolouch <nolouch@gmail.com>
2019-04-23 19:51:34 +08:00
Gyuho Lee 877f11bed8 etcdserver: improve heartbeat send failures logging
Signed-off-by: Gyuho Lee <leegyuho@amazon.com>
2019-04-19 10:58:17 -07:00
Sam Batschelet 9915d02022 *: Change gRPC proxy to expose etcd server endpoint /metrics
This PR resolves an issue where the `/metrics` endpoints exposed by the proxy were not returning metrics of the etcd members servers but of the proxy itself.

Signed-off-by: Sam Batschelet <sbatsche@redhat.com>
2019-04-10 16:09:32 -04:00
James Shubin 368f70a37c etcdserver: Use panic instead of fatal on no space left error
When using the embed package to embed etcd, sometimes the storage prefix
being used might be full. In this case, this code path triggers, causing
an: `etcdserver: create wal error: no space left on device` error, which
causes a fatal. A fatal differs from a panic in that it also calls
os.Exit(1). In this situation, the calling program that embeds the etcd
server will be abruptly killed, which prevents it from cleaning up
safely, and giving a proper error message. Depending on what the calling
program is, this can cause corruption and data loss.

This patch switches the fatal to a panic. Ideally this would be a
regular error which would get propagated upwards to the StartEtcd
command, but in the meantime at least this can be caught with recover().

This fixes the most common fatal that I've experienced, but there are
surely more that need looking into. If possible, the errors should be
threaded down into the code path so that embedding etcd can be more
robust.

Fixes: https://github.com/etcd-io/etcd/issues/10588
2019-03-27 15:24:33 -04:00
johncming bd41f74168 etcdserver/api/rafthttp: fix the location of close http body. 2019-03-11 22:20:38 +08:00
zhoulin xie a943ad0ee4 client/keys_bench_test.go: Fix some misspells
Signed-off-by: zhoulin xie <zhoulin.xie@daocloud.io>
2019-02-28 14:36:06 -05:00
Gyuho Lee 8d1a62e7ef *: use default log configuration for server
Signed-off-by: Gyuho Lee <leegyuho@amazon.com>
2019-02-21 10:57:26 -08:00
WizardCXY e6c6d8492e *: add flag to let etcd use the new boltdb freelistType feature 2019-02-14 11:07:08 +08:00
Hitoshi Mitake 72dd4a18c5 *: add a new option --enable-grpc-gateway for enabling/disabling grpc gateway 2019-01-23 03:26:34 +09:00
Xiang Li 2063b358c8
Merge pull request #10218 from mailgun/maxim/develop
Remove infinite loop in doSerialize
2019-01-09 10:38:25 -08:00
johncming e8f46ce341 etcdserver: add a test to verify not to send duplicated append responses 2019-01-09 10:37:43 +08:00
Sam Batschelet 577d7c0df2 e2e: update test to reflect (ST1005) update.
Signed-off-by: Sam Batschelet <sbatsche@redhat.com>
2019-01-08 21:04:20 -05:00
Sam Batschelet a82703b69e *: error strings should not end with punctuation or a newline (ST1005)
Signed-off-by: Sam Batschelet <sbatsche@redhat.com>
2019-01-08 21:04:20 -05:00
Xiang Li 6511829d1f
Merge pull request #10374 from johncming/deprecated
api/rafthttp: remove deprecated req.Cancel.
2019-01-08 14:33:25 -08:00
Gyuho Lee 442c863413
Merge pull request #10377 from johncming/cancel-pos
api/v2auth: remove defer in loop.
2019-01-08 09:43:06 -08:00
Xiang Li b04633fd8e
Merge pull request #10375 from johncming/redundant-parentheses
etcdserver: remove redundant parentheses.
2019-01-07 18:38:26 -08:00
caoming e96dbfb973 api/v2auth: remove defer in loop. 2019-01-08 08:56:55 +08:00
caoming 5060560f92 api/v2store: use camel case instead of snake case. 2019-01-07 10:35:23 +08:00
caoming 802e2aaadd etcdserver: remove redundant parentheses. 2019-01-07 10:27:52 +08:00
caoming 4651f49a5c api/rafthttp: remove deprecated req.Cancel. 2019-01-07 10:12:47 +08:00
caoming b2e0e760a0 etcdserver: add missing lg assignment. 2019-01-05 09:24:48 +08:00
lsytj0413 792aad932f refactor(*): fix golint warning 2018-12-24 11:43:10 +08:00
Xiang Li 3faed211e5 *: add flags to setup backend related config 2018-11-26 15:50:26 -08:00
Gyuho Lee 291768af0f etcdserver/*: add "etcd_cluster_version" metric
Signed-off-by: Gyuho Lee <leegyuho@amazon.com>
2018-11-13 21:49:12 -08:00
Maxim Vladimirskiy 91e583cba6 etcdserver: Remove infinite loop in doSerialize
Once chk(ai) fails with auth.ErrAuthOldRevision it will always do,
regardless how many times you retry. So the error is better be returned
to fail the pending request and make the client re-authenticate.
2018-11-12 23:28:24 +03:00
Shin'ya Ueoka aa4313a55a *: fix github links 2018-11-10 11:14:18 +09:00
Gyuho Lee 0f0919c19c
Merge pull request #10159 from gyuho/version-log
etcdserver: clear message in cluster version decision
2018-10-09 18:10:14 -07:00
Gyuho Lee d2a0f17b82
Merge pull request #10155 from gyuho/metrics-messages
rafthttp: probe all raft transports
2018-10-09 11:18:31 -07:00
Gyuho Lee 59dd78dde8 etcdserver: clear message in cluster version decision
Only leader can decide cluster version.
Clarify the logging that this local node is the leader.

Signed-off-by: Gyuho Lee <leegyuho@amazon.com>
2018-10-08 16:05:10 -07:00
Gyuho Lee 601d8b4677 etcdserver/api/etcdhttp: remove unused "HandleHealth" function
Signed-off-by: Gyuho Lee <leegyuho@amazon.com>
2018-10-07 17:16:18 -07:00
Gyuho Lee 004e04a1d1 etcdserver/api/etcdhttp: add "etcd_server_health_success/failures"
Signed-off-by: Gyuho Lee <leegyuho@amazon.com>
2018-10-07 17:15:12 -07:00
Gyuho Lee 884a8bd36b etcdserver/api/rafthttp: configure "streamProber" in tests
Signed-off-by: Gyuho Lee <leegyuho@amazon.com>
2018-10-07 03:32:05 -07:00
Gyuho Lee 7b1ef37054 etcdserver/api/rafthttp: probe all Raft messages' RTT
This PR adds another probing routine to monitor the connection
for Raft message transports. Previously, we only monitored
snapshot transports.

In our production cluster, we found one TCP connection had >8-sec
latencies to a remote peer, but "etcd_network_peer_round_trip_time_seconds"
metrics shows <1-sec latency distribution, which means etcd server
was not sampling enough while such latency spikes happen
outside of snapshot pipeline connection.

Signed-off-by: Gyuho Lee <leegyuho@amazon.com>
2018-10-07 03:28:54 -07:00
Gyuho Lee 4a239070c8 etcdserver/api/rafthttp: display roundtripper name in warnings
Signed-off-by: Gyuho Lee <leegyuho@amazon.com>
2018-10-07 03:14:42 -07:00
Gyuho Lee 47cff4dfe5 etcdserver/api/rafthttp: rename to "pipelineProber"
Preliminary work to add prober to "streamRt"

Signed-off-by: Gyuho Lee <leegyuho@amazon.com>
2018-10-07 03:13:10 -07:00
nolouch 6ea54195a6 client/integration: try to fix tests 2018-09-18 01:44:57 +08:00
nolouch c15fb607f6 server: broadcast leader changed 2018-09-17 14:15:04 +08:00
nolouch f3f6427586 server: prevent blocking 2018-09-14 16:08:29 +08:00
nolouch 4de27039cb server: drop read request if found leader changed 2018-09-14 15:58:35 +08:00
Gyuho Lee 8560221091 etcdserver: fix gofmt warnings with Go 1.11
Signed-off-by: Gyuho Lee <leegyuho@amazon.com>
2018-08-28 21:45:12 -07:00
Gyuho Lee 0ef9ef3c74 *: rerun "gofmt"
Signed-off-by: Gyuho Lee <leegyuho@amazon.com>
2018-08-28 18:25:39 -07:00
Gyuho Lee 1399bc69ce etcdserver: update import paths to "go.etcd.io/etcd"
Signed-off-by: Gyuho Lee <leegyuho@amazon.com>
2018-08-28 17:47:55 -07:00
Sam Batschelet af85949b41
Merge pull request #10024 from gyuho/became-inactive
etcdserver/api/rafthttp: clarify "became inactive" warning
2018-08-24 22:12:16 -04:00
Sam Batschelet 24ee22ab48
Merge pull request #10026 from gyuho/read-index
etcdserver: clarify read index wait timeout warnings
2018-08-24 22:11:58 -04:00
Gyuho Lee 38711761a1 etcdserver: clarify read index wait timeout warnings
"read index" doesn't tell much about the root cause.
Most likely, the local follower node is having slow
network, thus timing out waiting to receive read
index response from leader.

Signed-off-by: Gyuho Lee <leegyuho@amazon.com>
2018-08-17 17:59:41 -07:00
Gyuho Lee 156ff6461d etcdserver/api/rafthttp: clarify "became inactive" warning
Signed-off-by: Gyuho Lee <leegyuho@amazon.com>
2018-08-17 17:45:53 -07:00
Jingyi Hu 8d85259b56 etcdserver/api/v3rpc/interceptor: add log level checking
Check log level before generating and writing log info.
2018-08-17 16:12:05 -07:00