Hitoshi Mitake
ca91f898a2
auth, e2e, clientv3: the root role should be granted access to every key
...
This commit changes the semantics of the root role. The role should be
able to access to every key.
Partially fixes https://github.com/coreos/etcd/issues/6355
2016-10-11 12:19:46 -07:00
Gyu-Ho Lee
fcbada7798
Merge pull request #6622 from luxas/backport_arm_fixes
...
Backport arm fixes
2016-10-11 12:15:58 -07:00
Jared Hulbert
fad9bdc3e1
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 This makes comments out of date and therefore misleading.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.
2016-10-11 11:48:43 +03:00
Jared Hulbert
198ccb8b7b
raftpb: atomic access alignment
...
The Entry struct has misaligned fields that are accessed atomically. The
misalignment is caused by the EntryType enum which the Protocol Buffers
spec forces to be a 32bit int.
Moving the order of the fields without renumbering them in the .proto file
seems to align the go structure without changing the wire format.
2016-10-11 11:48:43 +03:00
Jared Hulbert
dc5d5c6ac8
raft: atomic access alignment
...
The relevant structures are properly aligned, however, there is no comment
highlighting the need to keep it aligned as is present elsewhere in the
codebase.
Adding note to keep alignment, in line with similar comments in the codebase.
2016-10-11 11:48:43 +03:00
Gyu-Ho Lee
f771eaca47
version: bump to v3.0.12+git
2016-10-07 16:42:12 -07:00
Gyu-Ho Lee
2d1e2e8e64
version: bump to v3.0.12
2016-10-07 15:14:25 -07:00
Gyu-Ho Lee
6412758177
v3rpc: remove redundant locks
2016-10-07 15:13:56 -07:00
Xiang Li
836c8159f6
v3rpc: lock progress and prevKV map correctly
2016-10-07 15:13:12 -07:00
Gyu-Ho Lee
e406e6e8f4
etcdctl/ctlv3: add 'prev-kv' flag to watch command
2016-10-07 14:23:09 -07:00
Gyu-Ho Lee
2fa2c6284e
clientv3: add 'prevKV' field to watch request
2016-10-07 14:22:58 -07:00
Gyu-Ho Lee
2862c4fa12
v3rpc: implement 'prev-kv' watch
2016-10-07 14:22:19 -07:00
Gyu-Ho Lee
6f89fbf8b5
etcdserver: use mvcc.WatchableKV for prev-kv watch
2016-10-07 14:22:00 -07:00
Gyu-Ho Lee
6ae7ec9a3f
*: regenerate proto
2016-10-07 14:21:19 -07:00
Gyu-Ho Lee
4a35b1b20a
etcdserverpb: add 'prev_kb' to WatchCreateRequest
2016-10-07 14:20:46 -07:00
Gyu-Ho Lee
c859c97ee2
mvccpb: add 'prev_kv' field
2016-10-07 14:19:59 -07:00
Gyu-Ho Lee
a091c629e1
version: bump to v3.0.11+git
2016-10-07 13:25:21 -07:00
Gyu-Ho Lee
96de94a584
version: bump to v3.0.11
2016-10-07 11:27:48 -07:00
Gyu-Ho Lee
e9cd8410d7
integration: add 'prevKV' to TestV3DeleteRange
2016-10-07 11:03:19 -07:00
Gyu-Ho Lee
e37ede1d2e
etcdserver: handle 'PrevKV'
2016-10-07 11:00:48 -07:00
Gyu-Ho Lee
4420a29ac4
etcdctl/ctlv3: add 'prev-kv' flag
2016-10-07 10:56:06 -07:00
Gyu-Ho Lee
0544d4bfd0
clientv3: add WithPrevKV OpOption
2016-10-07 10:54:45 -07:00
Gyu-Ho Lee
fe7379f102
clientv3: add Op.prevKV
2016-10-07 10:51:01 -07:00
Gyu-Ho Lee
c76df5052b
*: update proto to add 'prev_kv'
2016-10-07 10:47:47 -07:00
Xiang Li
3299cad1c3
*: add put prevkv
2016-10-07 10:39:08 -07:00
Anthony Romano
d9ab018c49
integration: test a canceled watch won't return a closing error
2016-10-05 14:19:36 -07:00
Anthony Romano
e853451cd2
clientv3: only return closing error to watcher if context is not canceled
...
Fixes #6503
2016-10-05 14:19:32 -07:00
Anthony Romano
1becf9d2f5
clientv3: fix race on watch initial revision
...
The initial revision was being updated in the substream goroutine defer;
this was racing with the resume path fetching the initial revision when
the substream closes during resume. Instead, update the initial revision
whenever the substream processes a new watch response. Since the substream
cannot receive a watch response while it is resuming, the write to the
initial revision is ordered to always happen after the resume read.
Fixes #6586
2016-10-05 10:56:36 -07:00
Anthony Romano
1a712cf187
clientv3: make IsProgressNotify() false on compact event and closed channel
...
Fixes #6549
2016-10-04 15:13:02 -07:00
Gyu-Ho Lee
023f335f67
wal: set PageWriter offset in file encoder
2016-10-04 15:12:47 -07:00
Gyu-Ho Lee
bf0da78b63
pkg/ioutil: configure pageOffset in NewPageWriter
2016-10-04 15:12:46 -07:00
Anthony Romano
e8473850a2
integration: test canceling watchers when disconnected
2016-10-04 15:12:37 -07:00
Anthony Romano
b836d187fd
clientv3: simplify watch synchronization
...
Was more complicated than it needed to be and didn't really work in the
first place. Restructured watcher registation to use a queue.
2016-10-04 15:12:18 -07:00
Gyu-Ho Lee
9b09229c4d
version: bump to v3.0.10+git
2016-09-23 11:13:45 -07:00
Gyu-Ho Lee
546c0f7ed6
version: bump to v3.0.10
2016-09-23 10:49:03 -07:00
sharat
adbad1c9b5
ctlv3: close snapshot file before rename (Windows)
2016-09-23 09:11:02 -07:00
Anthony Romano
273b986751
clientv3: process closed watcherStreams in watcherGrpcStream run loop
...
Was racing with Watch() when closing the grpc stream on no watchers.
Fixes #6476
2016-09-21 15:52:20 -07:00
Gyu-Ho Lee
5b205729b9
rafthttp: add v3.0.0 to supported streams
2016-09-16 21:54:55 +09:00
Anthony Romano
fe900b09dd
version: bump to v3.0.9+git
2016-09-15 15:10:23 -07:00
Anthony Romano
494c012659
version: bump to v3.0.9
2016-09-15 12:56:33 -07:00
Anthony Romano
4abc381ebe
clientv3: drain buffered WatchResponses before resuming
...
Otherwise, the watcherStream can receive WatchResponses in the
middle of a resume, corrupting the stream.
Fixes #6364
2016-09-15 12:38:15 -07:00
Anthony Romano
73c8fdac53
integration: fix compilation for backported Election test
2016-09-15 11:45:37 -07:00
sharat
ee2717493a
ctlv3: fix line parsing for Windows
2016-09-15 11:25:53 -07:00
Xiang Li
2435eb9ecd
clientv3: balancer panics when call up after close
...
Fix the issue by adding a simple guard varable.
2016-09-15 18:46:26 +09:00
Anthony Romano
8fb533dabe
embed: warn on domain name in listener
2016-09-15 18:46:19 +09:00
Anthony Romano
2f0f5ac504
Revert "Merge pull request #6365 from heyitsanthony/fix-dns-bind"
...
This reverts commit af5ab7b351da6a0f0594
2016-09-15 18:43:46 +09:00
Jason E. Aten
9ab811d478
auth: fix range handling bugs.
...
Test 15, counting from zero, in TestGetMergedPerms
in etcd/auth/range_perm_cache_test.go, was trying
incorrectly assert that [a, b) merged with [b, "")
should be [a, b). Added a test specifically for
this. This patch fixes the incorrect larger test
and the bugs in the code that it was hiding.
Fixes #6359
2016-09-15 18:41:56 +09:00
Anthony Romano
e0a99fb4ba
version: bump to v3.0.8+git
2016-09-09 15:56:31 -07:00
Anthony Romano
d40982fc91
version: bump to v3.0.8
2016-09-09 13:14:44 -07:00
Gyu-Ho Lee
fe3a1cc31b
wal: fix error type
2016-09-09 09:11:25 +09:00
