Yicheng Qin
fbaef05885
*: bump to v2.0.10
2015-04-22 15:21:38 -07:00
Brandon Philips
31a94d28e3
etcdctl: add extended as output format
...
extended wasn't documented in the help as one of the output formats, fix
this!
Conflicts:
etcdctl/main.go
2015-04-22 15:11:06 -07:00
Xiang Li
88660a303f
snap: load should only return ErrNoSnapshot
...
If there is no available snapshot, load should return
ErrNoSnapshot. etcdserver might recover from that error
if it still have complete WAL files.
2015-04-22 15:09:38 -07:00
Xiang Li
53c74dbd0b
etcdserver: prevExist=true + condition is compareAndSwap
...
PrevExist indicates the key should exist. Condition compares with
an existing key. So PrevExist+condition = CompareAndSwap not Update.
2015-04-22 15:09:28 -07:00
Xiang Li
8a8af60fad
etcdctl: backup tool should use the new layout
2015-04-22 15:09:15 -07:00
xiaost
7de19fefe8
etcdserver: fix minor bug in EtcdServer.send
...
it seems to nothing serious.
after deleted peers, the log may output:
"etcdserver: send message to unknown receiver %s"
2015-04-22 15:09:04 -07:00
Xiang Li
7750f387b0
wal: better log msg
2015-04-22 15:08:50 -07:00
Xiang Li
e33ab24442
wal: never leave a corrupted wal file
...
If the process dies during wal.cut(), it might leave a corrupted wal
file. This commit solves the problem by creating a temp wal file first,
then atomically rename it to a wal file when we are sure it is vaild.
Conflicts:
wal/wal.go
2015-04-22 15:08:42 -07:00
Xiang Li
fce2c1eeaf
discovery: drop trailing . from srv target
2015-04-22 15:06:20 -07:00
Xiang Li
6a3bb93305
discovery: add a test case for srv
...
During srv discovery, it should try to match local member with
resolved addr and return unresolved hostnames for the cluster.
Conflicts:
discovery/srv_test.go
2015-04-22 15:06:03 -07:00
Xiang Li
21455d2f3b
*: stop using resolved tcp addr
...
We start to resolve host into tcp addrs since we generate
tcp based initial-cluster during srv discovery. However it
creates problems around tls and cluster verification. The
srv discovery only needs to use resolved the tcp addr to
find the local node. It does not have to resolve everything
and use the resolved addrs.
This fixes #2488 and #2226
2015-04-22 14:59:07 -07:00
Brian Akins
51bb4220c5
Clarify that it is the proxy doing the shuffle.
2015-04-22 14:58:54 -07:00
Brian Akins
d8c506923f
proxy: shuffle endpoints
...
Shuffle endpoitns to avoid being "stuck" to a single cluster member.
2015-04-22 14:58:40 -07:00
Yicheng Qin
5d778f85ca
*: bump to v2.0.9+git
2015-04-07 15:18:50 -07:00
Yicheng Qin
02697ca725
*: bump to v2.0.9
2015-04-07 15:18:29 -07:00
Xiang Li
bd693c7069
etcdctl: refactor message in import command
2015-04-07 15:16:13 -07:00
Xiang Li
52c90cdcfb
etcdctl: import hidden keys
2015-04-07 14:49:40 -07:00
Xiang Li
a88b22ac0a
store: fix watcher removal
2015-04-07 14:46:10 -07:00
Yicheng Qin
e93f8b8a12
*: bump to v2.0.8+git
2015-03-31 14:29:38 -07:00
Yicheng Qin
86e616c6e9
*: bump to v2.0.8
2015-03-31 14:29:13 -07:00
Barak Michener
5ae55a2c0d
etcdctl: fix import typos
2015-03-31 13:48:18 -07:00
Xiang Li
62ce6eef7b
etcdctl: main routine of import command should wait for goroutine existing
2015-03-31 13:26:15 -07:00
Xiang Li
7df4f5c804
build: do not build internal debugging tool
...
We are still playing around with the dump-log tool.
Stop building it publicly until we are happy with its
ux and functionality.
2015-03-31 13:26:05 -07:00
Xiang Li
461c24e899
etcdct: adopt new client port by default
...
etcdserver uses both 4001 and 2379 for serving client requests by
default. etcdctl supports both ports by default.
2015-03-31 13:25:56 -07:00
Xiang Li
6d90d03bf0
etcdctl: add migratesnap command
2015-03-31 13:25:39 -07:00
Yicheng Qin
9995e80a2c
Revert "etcdhttp: add internalVersion"
...
This reverts commit a77bf97c147a93e5b78898b427d081ba11b8f118.
Conflicts:
version/version.go
Conflicts:
version/version.go
2015-03-31 13:25:22 -07:00
Xiang Li
229405f113
*: remove upgrading related stuff
2015-03-31 13:24:28 -07:00
Mateus Braga
b3f2a998d4
docs: add clarity about the 1000 events history
...
When talking about missing events on a particular key, the 1000 event history
limit can be understood as being per key, instead of etcd-wide events. Make it
clear that it is across all etcd keys.
2015-03-31 13:24:19 -07:00
Xiang Li
8436e901e9
etcdserver: loose member validation for joining existing cluster
2015-03-31 13:24:07 -07:00
Yicheng Qin
c03f5cb941
*: bump to v2.0.7+git
2015-03-24 23:14:38 -07:00
Yicheng Qin
0cb90e4bea
*: bump to v2.0.7
2015-03-24 23:07:57 -07:00
Yicheng Qin
df83b1b34e
wal: fix missing import
2015-03-24 23:00:04 -07:00
Xiang Li
f2bef04009
wal: releastTo should work with large release index
2015-03-24 22:51:02 -07:00
Yicheng Qin
02198336f6
version: not return err NotExist in Detect
2015-03-24 22:50:44 -07:00
Yicheng Qin
0c9a226e0e
etcdserver: print out extra files in data dir instead of erroring
2015-03-24 22:50:33 -07:00
Yicheng Qin
5bd1d420bb
etcdserver: add join-existing check
2015-03-24 22:49:41 -07:00
Yicheng Qin
a1cb5cb768
etcdmain: print error when non-flag args remain
2015-03-24 22:49:31 -07:00
Yicheng Qin
acba49fe81
*: bump to v2.0.6+git
2015-03-23 14:05:08 -07:00
Yicheng Qin
e3c902228b
*: bump to v2.0.6
2015-03-23 13:52:00 -07:00
Yicheng Qin
52a2d143d2
migrate: remove starter code
...
It has been moved to github.com/coreos/etcd-starter.
2015-03-21 11:15:26 -07:00
招牌疯子
f53d550a79
store: fixed clone error for store stats.
2015-03-21 11:14:06 -07:00
Brandon Philips
63b799b891
migrate: detect version 2.0.1
...
Without this code a second start will crash:
```
$ ./bin/etcd -name foobar --data-dir=foobar
2015/03/18 18:06:28 starter: detect etcd version 2.0.1 in foobar
2015/03/18 18:06:28 starter: unhandled etcd version in foobar
panic: starter: unhandled etcd version in foobar
goroutine 1 [running]:
log.Panicf(0x594770, 0x25, 0x208927c70, 0x1, 0x1)
/usr/local/go/src/log/log.go:314 +0xd0
github.com/coreos/etcd/migrate/starter.checkInternalVersion(0x20889a480, 0x0, 0x0)
/Users/philips/src/github.com/coreos/etcd/gopath/src/github.com/coreos/etcd/migrate/starter/starter.go:160 +0xf2f
github.com/coreos/etcd/migrate/starter.StartDesiredVersion(0x20884a010, 0x3, 0x3)
/Users/philips/src/github.com/coreos/etcd/gopath/src/github.com/coreos/etcd/migrate/starter/starter.go:77 +0x2a9
main.main()
/Users/philips/src/github.com/coreos/etcd/gopath/src/github.com/coreos/etcd/main.go:46 +0x25e
goroutine 9 [syscall]:
os/signal.loop()
/usr/local/go/src/os/signal/signal_unix.go:21 +0x1f
created by os/signal.init·1
/usr/local/go/src/os/signal/signal_unix.go:27 +0x35
```
2015-03-21 11:13:55 -07:00
Brandon Philips
697883fb8c
etcdmain: let user provide a name w/o initial-cluster update
...
Currently this doesn't work if a user wants to try out a single machine
cluster but change the name for whatever reason. This is because the
name is always "default" and the
```
./bin/etcd -name 'baz'
```
This solves our problem on CoreOS where the default is `ETCD_NAME=%m`.
2015-03-21 11:13:42 -07:00
Brandon Philips
f794f87f26
Documentation: fixup grammar around the unsafe flags
2015-03-21 11:13:28 -07:00
Xiang Li
0847986d4a
etcdmain: identify data dir type
2015-03-21 11:12:18 -07:00
funkygao
9ea80c6ac1
raft: fix godoc about starting a node
2015-03-21 11:11:21 -07:00
Xiang Li
02fb648abf
etcdmain: verify heartbeat and election flag
2015-03-21 11:11:09 -07:00
kmeaw
4c9e1686b1
pkg/flags: Add support for IPv6 addresses
...
Support IPv6 address for ETCD_ADDR and ETCD_PEER_ADDR
pkg/flags: Support IPv6 address for ETCD_ADDR and ETCD_PEER_ADDR
pkg/flags: tests for IPv6 addr and bind-addr flags
pkg/flags: IPAddressPort.Host: do not enclose IPv6 address in square brackets
pkg/flags: set default bind address to [::] instead of 0.0.0.0
pkg/flags: we don't need fmt any more
also, one minor fix: net.JoinHostPort takes string as a port value
pkg/flags: fix ipv6 tests
pkg/flags: test both IPv4 and IPv6 addresses in TestIPAddressPortString
etcdmain: test: use [::] instead of 0.0.0.0
2015-03-21 11:05:20 -07:00
Yicheng Qin
0fb9362c5c
*: bump to v2.0.5+git
2015-03-11 17:00:51 -07:00
Yicheng Qin
9481945228
*: bump to v2.0.5
2015-03-11 11:33:43 -07:00
