a1ce07a321
etcdserver: reject member removal that breaks the current active quorum
2016-08-10 17:00:39 -07:00
9063ce5e3f
etcdserver, embed: stricter reconfig checking
...
Make --strict-reconfig-check a default and check if cluster is healthy when
adding a member.
2016-08-05 16:59:25 -07:00
d69d438289
*: minor cleanup for lease
2016-08-04 20:39:32 -07:00
29a077bdbe
etcdserver: always recover lessor first
2016-08-04 08:06:19 -07:00
bf71497537
etcdserver, lease: tie lease min ttl to election timeout
2016-08-02 13:06:57 -07:00
06da46c4ee
etcdserver: apply serialized requests outside auth apply lock
...
Fixes #6010
2016-07-30 22:00:49 -07:00
13c2d32061
Merge pull request #6045 from heyitsanthony/fix-version-race
...
etcdserver, api, membership: don't race on setting version
2016-07-27 08:56:39 -07:00
0090573749
etcdserver: skip range requests in txn if the result is needless
...
If a server isn't serving txn requests from a client, the server
doesn't need the result of range requests in the txn.
This is a succeeding commit of
https://github.com/coreos/etcd/pull/5689
2016-07-26 19:49:07 -07:00
de2c3ec3db
etcdserver, api, membership: don't race on setting version
...
Fixes #6029
2016-07-26 18:21:40 -07:00
2d761d64a4
etcdserver: set applied index correctly
2016-07-16 11:44:18 -07:00
27b03f0ed5
*: deny proposals when there is a huge gap between apply/commit
2016-07-14 10:02:55 -07:00
f65e75e4b3
*: remove unnecessary data upgrade code
2016-07-11 15:11:56 -07:00
c47689d98f
Merge pull request #5689 from mitake/skip-apply
...
RFC: etcdserver, pkg: skip needless log entry applying
2016-07-10 01:23:35 +09:00
f78d4713ea
etcdserver: atomic access alignment
...
Most fields accessed with sync/atomic functions are 64bit aligned, but a couple
are not. This makes comments out of date and therefore misleading.
Affected fields reordered, comments scrubbed and updated.
2016-07-08 11:20:47 -07:00
abb20ec51f
etcdserver, pkg: skip needless log entry applying
...
This commit lets etcdserver skip needless log entry applying. If the
result of log applying isn't required by the node (client that issued
the request isn't talking with the node) and the operation has no side
effects, applying can be skipped.
It would contribute to reduce disk I/O on followers and be useful for
a cluster that processes much serializable get.
2016-07-08 15:16:45 +09:00
8a8a8253fa
etcdserver: commit before sending snapshot
2016-07-03 13:54:05 -07:00
b7f5f8fc99
etcdserver: exit on missing backend only if semver is >= 3.0.0
2016-07-01 09:10:01 -07:00
9614dc6e71
etcdserver: check index of the kv when restarting
2016-06-27 10:27:27 -07:00
891ddcba6e
etcdserver: refuse to restart if backend file is missing
2016-06-26 21:16:51 -07:00
d37e564eaa
etcdserver: use TouchDirAll
2016-06-19 11:26:52 -07:00
57474697af
etcdserver: add applied metrics
2016-06-17 11:52:50 -07:00
699e76b631
etcdserver: only pause compaction when sending snapshot
2016-06-16 08:57:02 -07:00
16db9e68a2
auth, etcdserver: separate auth checking apply from core apply
2016-06-15 09:03:27 -07:00
65ff76882b
Merge pull request #5624 from xiang90/warn_apply
...
etcdserver: warn heavy apply
2016-06-10 15:28:27 -07:00
64eccd519d
etcdserver: warn heavy apply
2016-06-10 14:43:34 -07:00
ead5096fa9
auth, etcdserver: make auth tokens consistent for all nodes
...
Currently auth tokens are generated in the replicated state machine
layer randomly. It means one auth token generated in node A cannot be
used for node B. It is problematic for load balancing and fail
over. This commit moves the token generation logic from the state
machine to API layer (before raft) and let all nodes share a single
token.
Log index of Raft is also added to a token for ensuring uniqueness of
the token and detecting activation of the token in the cluster (some
nodes can receive the token before generating and installing the token
in its state machine).
This commit also lets authStore have simple token related things. It
is required because of unit test. The test requires cleaning of the
state of the simple token things after one test (succeeding test can
create duplicated token and it causes panic).
2016-06-10 13:55:37 -07:00
8528c8c599
*: more logging on critical state change
...
Add more logging for better debugging purpose.
2016-05-31 23:31:03 -07:00
815bc5307f
Merge pull request #5489 from linuxcer/master
...
etcdserver: fix typo in server.go
2016-05-30 15:20:02 -07:00
29cc568659
etcdserver: fix typo in server.go
2016-05-31 05:54:30 +08:00
8e821cdc70
*: do permission check in raft log apply phase
...
This commit lets etcdserver check permission during its log applying
phase. With this change, permission checking of operations is
supported.
Currently, put and range are supported. In addition, multi key
permission check of range isn't supported yet.
2016-05-29 00:05:48 +09:00
6f2e7875aa
etcdctl: add migrate command
...
Migrate command accepts a datadir and an optional user-provided
transformer function that transform v2 keys to v2 keys.
Migrate command then builds a v3 backend state based on the existing
v2 keys and the output of the transformer function.
2016-05-19 12:17:15 -07:00
73204e9637
etcdserver: wait for snapshots before closing raft
...
Fixes #5374
2016-05-17 15:04:25 -07:00
abb4cd5646
etcdserver: update LICENSE header
2016-05-12 20:49:40 -07:00
9c103dd0de
*: cancel required leader streams when memeber lost its leader
2016-05-12 19:42:21 -07:00
9bf601a921
etcdserver: log timeout
2016-05-03 13:39:31 -07:00
434f2c356d
etcdserver: do not serve requests before finish the first internal proposal
2016-04-27 15:46:31 -07:00
b7ac758969
*: rename storage package to mvcc
2016-04-25 15:25:51 -07:00
08d879341d
etcdserver: stop raft after stopping apply scheduler
...
Was causing a pipeline leak.
2016-04-22 17:15:13 -07:00
4991cda202
etcdsever: fix the leaky snashot routine issue
2016-04-21 08:48:11 -07:00
ebace2eb1b
etcdserver: split out v2 Do() API from core server code
2016-04-20 10:29:22 -07:00
41382bc3f0
etcdserver: split out v2 raft apply interface
2016-04-20 10:29:22 -07:00
641a1a66e1
*: fix govet -shadow in go tip
2016-04-15 07:39:52 -07:00
e9735b7bd0
etcdserver: save cluster version into backend
2016-04-12 11:37:22 -07:00
0a684c10ad
Merge pull request #5025 from xiang90/no_dup_resp
...
etcdserver: do not send out out of date appResp
2016-04-11 14:41:52 -07:00
3bad47d691
Merge pull request #5018 from xiang90/b
...
etcdserver: set backend to cluster
2016-04-11 13:02:57 -07:00
130567832f
Merge pull request #4734 from luxas/32bit_alignments
...
etcdserver: align 64-bit atomics on 8-byte boundary
2016-04-10 19:18:15 -07:00
de7692b2b2
etcdserver: do not send out out of date appResp
2016-04-09 23:30:00 -07:00
3c0ac9d600
etcdserver: set backend to cluster
2016-04-08 21:46:45 -07:00
b13b77f362
membership: update attr in membership pkg
2016-04-07 21:25:32 -07:00
bf2289ae00
etcdserver: move membership related code to membership pkg
2016-04-07 14:21:37 -07:00
030865abe3
*: move Cluster interface to api
2016-04-07 08:05:47 -07:00
2c50eb240e
*: mv etcdhttp into api pkg
2016-04-04 16:31:35 -07:00
7ce5c2b9ff
Merge pull request #4902 from heyitsanthony/alarm-ctl
...
etcdctl: alarm command
2016-03-30 13:55:29 -07:00
eddc741b5e
*: simplify consistent index handling
2016-03-30 13:38:28 -07:00
96ee00a322
etcdserverpb: make alarm memberId uint64
...
To be consistent with Cluster API
2016-03-29 20:15:39 -07:00
3fbacf4be2
v3rpc: move Hash RPC to Maintenance service
2016-03-28 17:15:58 -07:00
a403a94d7b
etcdserver: cap new keys on space alarm
2016-03-28 14:56:26 -07:00
9e7f47c490
etcdserver: Alarm RPC
...
Alarms are events that nodes can use to relay health information to
the rest of the cluster. A node may Activate an alarm and that alarm
will stay set until Deactivated.
2016-03-28 14:56:26 -07:00
9c8253c543
etcdserver, v3rpc: space quotas
2016-03-28 14:56:26 -07:00
94e77cfa5d
etcdserver: move v3 raft apply functions to interface
2016-03-28 13:16:21 -07:00
70a9391378
*: enable v3 by default
2016-03-23 17:01:36 -07:00
bd832e5b0a
*: migrate Godeps to vendor/
2016-03-22 17:10:28 -07:00
afb1bc242b
Merge pull request #4822 from mitake/auth-backend
...
auth, etcdserver: add a method for updating backend during apply snap…
2016-03-21 23:34:46 -07:00
4e39f690f2
auth, etcdserver: add a method for recoverying from backend during apply snapshot
...
This commit adds a new method Recovery() to auth.AuthStore for
recoverying auth state from backend during apply snapshot. It follows
a manner of the lessor.
2016-03-22 15:17:40 +09:00
67c2384bdf
*: replace '-' with '--' in doc
...
Fix https://github.com/coreos/etcd/issues/4595 .
2016-03-21 11:12:43 -07:00
4eb1cfd658
etcdserver, auth: new package auth for the auth feature
...
This commit adds a new package auth. Its role is persisting auth
related metadata. This commit also connects its main interface
AuthStore and v3 server.
2016-03-14 13:57:41 +09:00
c6192d1d7a
etcdserver: remove todo
2016-03-10 19:10:20 -08:00
adcba975cb
*: recover lessor when applying snapshot
2016-03-10 17:06:58 -08:00
2f12ea893b
etcdserver: add maintain service to support defrag
2016-03-09 22:29:21 -08:00
a44645e13d
etcdserver: align 64-bit atomics on 8-byte boundary
2016-03-10 07:24:33 +02:00
f0dbd0b856
Merge pull request #4646 from xiang90/starvation
...
etcdserver: detect raft stravation
2016-02-29 19:58:17 -08:00
1d6ebdd35c
pkg/timeutil: removal
...
Overkill of a package for three lines of code.
2016-02-29 17:07:24 -08:00
d6520303c6
etcdserver: detect raft starvation caused by contention
2016-02-29 17:06:57 -08:00
d265fe000c
*: support time based auto compaction.
...
Fix https://github.com/coreos/etcd/issues/3906 .
We will have extensive doc to talk about what is compaction
and what is auto compaction soon.
2016-02-25 16:02:03 -08:00
c5b51946eb
*: exported godoc fixups
2016-02-21 20:36:44 -08:00
616f395920
etcdserver: use fifo scheduler for applier
2016-02-11 19:21:30 -08:00
82778ed478
Add refresh parameter to allow TTL refreshes without firing watch/wait responses
2016-02-08 10:37:37 -07:00
fe7cfe4d3d
rafthttp: plumb local peer URLs through transport
2016-02-05 16:44:41 -08:00
e44e753e66
pkg/idutil: reduce conflict rate from 1% to 0.005%
...
Perviously, we only use 8bits from member identification
in id generation. The conflict rate is A(256,3)/256^3, which
is around 1%. Now we use 16bites to reduce the rate to 0.005%.
We can attach the full member id into id generation if needed...
2016-02-04 13:25:18 -08:00
9ae8d85049
integration: decrease timeout for isMemberBootstrapped
...
Spending seconds(!) when it would fail anyway.
integration/TestV3 (before): 100.670
integration/TestV3 (after): 29.571
2016-02-02 14:34:58 -08:00
20461ab11a
*: fix many typos
2016-01-31 21:42:39 -08:00
c4a0159601
documentation: fix typo "a etcd" -> "an etcd"
...
"a" is not a correct article for "etcd".
2016-01-28 19:19:32 +09:00
bd02d668c8
etcdserver: don't try to apply empty message list
...
If all messages have been applied, don't apply an empty messages list;
otherwise appliedi will update to 0 and etcd will panic.
Fixes #4278
2016-01-26 11:56:37 -08:00
12f6b8e72d
etcdserver: complete stopWithDelay on server shutdown
...
Was causing goroutine leaks on my machine.
2016-01-25 19:45:29 -08:00
ef6320e638
etcdserver: make cluster checking interval shorter
2016-01-25 08:16:05 +08:00
1aa312fcce
*: lease forwarding should resue transport
2016-01-25 06:56:07 +08:00
2e157530a0
etcdhttp, lease, v3api: forward keepalives to leader
...
keepalives don't go through raft so let follower peers announce
keepalives to the leader through the peer http handler
2016-01-22 12:40:40 -08:00
9113a27bde
lease: grant consistent lease IDs
...
When raft broadcasts a Grant to all nodes, all nodes must
agree on the same lease ID. Otherwise, attaching a key to
a lease will fail since the lease ID is node-dependent.
2016-01-22 09:43:39 -08:00
588f655b4e
etcdmain: add an option for pprof
...
This commit adds a new option for activating profiling based on pprof
in etcd process.
- -enable-pprof: boolean type option which activates profiling
For example, if a client URL is http://localhost:12379 , users and
developers access heap profiler with this URL:
http://localhost:12379/debug/pprof/heap
2016-01-13 16:12:26 +09:00
6b1d9fb7ce
*: stop lessor when etcdserver is stopped
2016-01-12 10:06:11 -08:00
f5753f2f51
*: support lease Attach
...
Now we can attach keys to leases. And revoking the lease removes all
the attached keys of that lease.
2016-01-09 11:01:58 -08:00
2566699a48
*: revoke expired leases
2016-01-08 13:37:58 -08:00
f76166a041
*: fix minor typos
2016-01-08 00:21:19 -08:00
d9ca929a33
*: add support for lease create and revoke
...
Basic support for lease operations like create and revoke.
We still need to:
1. attach keys to leases in KV implmentation if lease field is set
2. leader periodically removes expired leases
3. leader serves keepAlive requests and follower forwards keepAlive
requests to leader.
2016-01-07 16:39:39 -08:00
1714290f4e
storage: support recovering from backend
...
We want the KV to support recovering from backend to avoid
additional pointer swap. Or we have to do coordination between
etcdserver and API layer, since API layer might have access to
kv pointer and use a closed kv.
2016-01-06 21:16:55 -08:00
5dd3f91903
*: make backend outside kv
...
KV and lease will share the same backend. Thus we need to make
backend outside KV.
2016-01-05 19:55:29 -08:00
8346a7c052
Merge pull request #4094 from heyitsanthony/send-merged-done-nowait
...
etcdserver: respect done channel when sleeping for snapshot backoff
2015-12-29 20:18:08 -08:00
942b5570bd
Merge pull request #4096 from heyitsanthony/serialize-applier-snapmerge
...
etcdserver: serialize snapshot merger with applier
2015-12-29 19:37:11 -08:00
4cd86ae1ef
etcdserver: serialize snapshot merger with applier
...
Avoids inconsistent snapshotting by only attempting to
create a snapshot after an apply completes.
Fixes #4061
2015-12-29 18:38:39 -08:00
c7c3bda8b9
etcdserver: respect done channel when sleeping for snapshot backoff
2015-12-29 13:23:41 -08:00
Anthony Romano