storage: range all unsynced at once #4043

gyuho · 2015-12-22T17:32:28Z

This is for #3848
to batch RangeHistory for all watchings at once.

xiang90 · 2015-12-22T19:49:48Z

storage/watchable_store.go

+	// s.store.compactMainRev < startRev <= s.store.currentRev.main
+	startRev := minRev
+
+	if startRev > 0 && startRev <= s.store.compactMainRev {


we should not simply return an error here. We might not be able to send an event to a watching due to the compaction, but for other watchers it might be just fine. So when ranging the unsynced watchers, we need to identify the watchers that cannot be caught up due to compaction, send error to them and remove them.

Ok I will just log and removed from unsynced

xiang90 · 2015-12-23T01:10:19Z

storage/watchable_store.go

+	fmt.Println("keys:", keys)
+	fmt.Println("vals:", vals)
+	/*
+	   min: foo1


min and max should be just rev not key. The key in boltdb is just rev like 1, 2, 3... not foo, bar.

Got it. I will try using rev. Thanks Xiang.

Sincerely,
Gyu-Ho Lee
On Dec 22, 2015 5:10 PM, "Xiang Li" [email protected] wrote:

In storage/watchable_store.go
#4043 (comment):

default:

s.store.kvindex.Restore(kv.Key, revision{kv.CreateRevision, 0}, rev, kv.Version)

}

// update revision

s.store.currentRev = rev

}

tx.Unlock()

fmt.Println("min:", string(min))

fmt.Println("max:", string(max))

fmt.Println("keys:", keys)

fmt.Println("vals:", vals)

/*

min: foo1

min and max should be just rev not key. The key in boltdb is just rev like
1, 2, 3... not foo, bar.

—
Reply to this email directly or view it on GitHub
https://github.com/coreos/etcd/pull/4043/files#r48315420.

xiang90 · 2015-12-23T20:19:12Z

@gyuho I think this is a good way to get all kv events that falls into the range of [min, max]. However, we probably want to send these events to watchers directly rather than keeping them in memory. Or it is very memory inefficient.

Basically, as you range over the keys you can decide what watching you should send the events to.

xiang90 · 2015-12-23T20:20:15Z

storage/watchable_store.go

+			delete(s.unsynced, w)
+			continue
+		}
+		totalLimit += cap(w.ch) - len(w.ch)


i am not sure why do we need to calculate the totalLimit. Can you please explain this?

It's legacy arg from RangeHistory but just confirmed that we can just pass 0 to get all.

https://github.com/coreos/etcd/blob/master/storage/backend/batch_tx.go#L88-L90

I will change. Thanks,

xiang90 · 2015-12-27T18:35:02Z

storage/watchable_store.go

+		if _, ok := keyToWatchings[string(kv.Key)]; !ok {
+			continue
+		}
+		revbs = append(revbs, ks[i])


i think we should move line 340 - 369 right here. Then we do not need to append these key-value pairs to a slice. And the code would be easier to read.

This is for etcd-io#3848. It replaces RangeHistory method for more efficient event sending.

This has been replaced by operations inside `syncWatchings`.

xiang90 · 2015-12-28T20:12:24Z

LGTM

storage: range all unsynced at once

Fix panic when gRPC proxy leader watcher is restored: ``` go test -v -tags cluster_proxy -cpu 4 -race -run TestV3WatchRestoreSnapshotUnsync === RUN TestV3WatchRestoreSnapshotUnsync panic: watcher minimum revision 9223372036854775805 should not exceed current revision 16 goroutine 156 [running]: github.com/coreos/etcd/mvcc.(*watcherGroup).chooseAll(0xc4202b8720, 0x10, 0xffffffffffffffff, 0x1) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watcher_group.go:242 +0x3b5 github.com/coreos/etcd/mvcc.(*watcherGroup).choose(0xc4202b8720, 0x200, 0x10, 0xffffffffffffffff, 0xc420253378, 0xc420253378) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watcher_group.go:225 +0x289 github.com/coreos/etcd/mvcc.(*watchableStore).syncWatchers(0xc4202b86e0, 0x0) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:340 +0x237 github.com/coreos/etcd/mvcc.(*watchableStore).syncWatchersLoop(0xc4202b86e0) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:214 +0x280 created by github.com/coreos/etcd/mvcc.newWatchableStore /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:90 +0x477 exit status 2 FAIL github.com/coreos/etcd/integration 2.551s ``` gRPC proxy spawns a watcher with a key "proxy-namespace__lostleader" and watch revision "int64(math.MaxInt64 - 2)" to detect leader loss. But, when the partitioned node restores, this watcher triggers panic with "watcher minimum revision ... should not exceed current ...". This check was added a long time ago, by my PR: etcd-io#4043 (comment) > we can remove this checking actually. it is impossible for a unsynced watching to have a future rev. or we should just panic here. However, now it's possible that a unsynced watcher has a future revision, when it was moved from a synced watcher group through restore operation. And there was no gRPC proxy, when we added this check. This PR adds "restore" flag to indicate that a watcher was moved from the synced watcher group with restore operation. Otherwise, the watcher with future revision in an unsynced watcher group would still panic. Example logs with future revision watcher from restore operation: ``` {"level":"info","ts":1527196358.9057755,"caller":"mvcc/watcher_group.go:261","msg":"choosing future revision watcher from restore operation","watch-key":"proxy-namespace__lostleader","watch-revision":9223372036854775805,"current-revision":16} {"level":"info","ts":1527196358.910349,"caller":"mvcc/watcher_group.go:261","msg":"choosing future revision watcher from restore operation","watch-key":"proxy-namespace__lostleader","watch-revision":9223372036854775805,"current-revision":16} ``` Signed-off-by: Gyuho Lee <[email protected]>

…ation Fix panic when gRPC proxy leader watcher is restored: ``` go test -v -tags cluster_proxy -cpu 4 -race -run TestV3WatchRestoreSnapshotUnsync === RUN TestV3WatchRestoreSnapshotUnsync panic: watcher minimum revision 9223372036854775805 should not exceed current revision 16 goroutine 156 [running]: github.com/coreos/etcd/mvcc.(*watcherGroup).chooseAll(0xc4202b8720, 0x10, 0xffffffffffffffff, 0x1) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watcher_group.go:242 +0x3b5 github.com/coreos/etcd/mvcc.(*watcherGroup).choose(0xc4202b8720, 0x200, 0x10, 0xffffffffffffffff, 0xc420253378, 0xc420253378) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watcher_group.go:225 +0x289 github.com/coreos/etcd/mvcc.(*watchableStore).syncWatchers(0xc4202b86e0, 0x0) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:340 +0x237 github.com/coreos/etcd/mvcc.(*watchableStore).syncWatchersLoop(0xc4202b86e0) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:214 +0x280 created by github.com/coreos/etcd/mvcc.newWatchableStore /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:90 +0x477 exit status 2 FAIL github.com/coreos/etcd/integration 2.551s ``` gRPC proxy spawns a watcher with a key "proxy-namespace__lostleader" and watch revision "int64(math.MaxInt64 - 2)" to detect leader loss. But, when the partitioned node restores, this watcher triggers panic with "watcher minimum revision ... should not exceed current ...". This check was added a long time ago, by my PR, when there was no gRPC proxy: etcd-io#4043 (comment) > we can remove this checking actually. it is impossible for a unsynced watching to have a future rev. or we should just panic here. However, now it's possible that a unsynced watcher has a future revision, when it was moved from a synced watcher group through restore operation. This PR adds "restore" flag to indicate that a watcher was moved from the synced watcher group with restore operation. Otherwise, the watcher with future revision in an unsynced watcher group would still panic. Example logs with future revision watcher from restore operation: ``` {"level":"info","ts":1527196358.9057755,"caller":"mvcc/watcher_group.go:261","msg":"choosing future revision watcher from restore operation","watch-key":"proxy-namespace__lostleader","watch-revision":9223372036854775805,"current-revision":16} {"level":"info","ts":1527196358.910349,"caller":"mvcc/watcher_group.go:261","msg":"choosing future revision watcher from restore operation","watch-key":"proxy-namespace__lostleader","watch-revision":9223372036854775805,"current-revision":16} ``` Signed-off-by: Gyuho Lee <[email protected]>

…ation This also happens without gRPC proxy. Fix panic when gRPC proxy leader watcher is restored: ``` go test -v -tags cluster_proxy -cpu 4 -race -run TestV3WatchRestoreSnapshotUnsync === RUN TestV3WatchRestoreSnapshotUnsync panic: watcher minimum revision 9223372036854775805 should not exceed current revision 16 goroutine 156 [running]: github.com/coreos/etcd/mvcc.(*watcherGroup).chooseAll(0xc4202b8720, 0x10, 0xffffffffffffffff, 0x1) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watcher_group.go:242 +0x3b5 github.com/coreos/etcd/mvcc.(*watcherGroup).choose(0xc4202b8720, 0x200, 0x10, 0xffffffffffffffff, 0xc420253378, 0xc420253378) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watcher_group.go:225 +0x289 github.com/coreos/etcd/mvcc.(*watchableStore).syncWatchers(0xc4202b86e0, 0x0) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:340 +0x237 github.com/coreos/etcd/mvcc.(*watchableStore).syncWatchersLoop(0xc4202b86e0) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:214 +0x280 created by github.com/coreos/etcd/mvcc.newWatchableStore /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:90 +0x477 exit status 2 FAIL github.com/coreos/etcd/integration 2.551s ``` gRPC proxy spawns a watcher with a key "proxy-namespace__lostleader" and watch revision "int64(math.MaxInt64 - 2)" to detect leader loss. But, when the partitioned node restores, this watcher triggers panic with "watcher minimum revision ... should not exceed current ...". This check was added a long time ago, by my PR, when there was no gRPC proxy: etcd-io#4043 (comment) > we can remove this checking actually. it is impossible for a unsynced watching to have a future rev. or we should just panic here. However, now it's possible that a unsynced watcher has a future revision, when it was moved from a synced watcher group through restore operation. This PR adds "restore" flag to indicate that a watcher was moved from the synced watcher group with restore operation. Otherwise, the watcher with future revision in an unsynced watcher group would still panic. Example logs with future revision watcher from restore operation: ``` {"level":"info","ts":1527196358.9057755,"caller":"mvcc/watcher_group.go:261","msg":"choosing future revision watcher from restore operation","watch-key":"proxy-namespace__lostleader","watch-revision":9223372036854775805,"current-revision":16} {"level":"info","ts":1527196358.910349,"caller":"mvcc/watcher_group.go:261","msg":"choosing future revision watcher from restore operation","watch-key":"proxy-namespace__lostleader","watch-revision":9223372036854775805,"current-revision":16} ``` Signed-off-by: Gyuho Lee <[email protected]>

…ation This also happens without gRPC proxy. Fix panic when gRPC proxy leader watcher is restored: ``` go test -v -tags cluster_proxy -cpu 4 -race -run TestV3WatchRestoreSnapshotUnsync === RUN TestV3WatchRestoreSnapshotUnsync panic: watcher minimum revision 9223372036854775805 should not exceed current revision 16 goroutine 156 [running]: github.com/coreos/etcd/mvcc.(*watcherGroup).chooseAll(0xc4202b8720, 0x10, 0xffffffffffffffff, 0x1) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watcher_group.go:242 +0x3b5 github.com/coreos/etcd/mvcc.(*watcherGroup).choose(0xc4202b8720, 0x200, 0x10, 0xffffffffffffffff, 0xc420253378, 0xc420253378) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watcher_group.go:225 +0x289 github.com/coreos/etcd/mvcc.(*watchableStore).syncWatchers(0xc4202b86e0, 0x0) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:340 +0x237 github.com/coreos/etcd/mvcc.(*watchableStore).syncWatchersLoop(0xc4202b86e0) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:214 +0x280 created by github.com/coreos/etcd/mvcc.newWatchableStore /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:90 +0x477 exit status 2 FAIL github.com/coreos/etcd/integration 2.551s ``` gRPC proxy spawns a watcher with a key "proxy-namespace__lostleader" and watch revision "int64(math.MaxInt64 - 2)" to detect leader loss. But, when the partitioned node restores, this watcher triggers panic with "watcher minimum revision ... should not exceed current ...". This check was added a long time ago, by my PR, when there was no gRPC proxy: #4043 (comment) > we can remove this checking actually. it is impossible for a unsynced watching to have a future rev. or we should just panic here. However, now it's possible that a unsynced watcher has a future revision, when it was moved from a synced watcher group through restore operation. This PR adds "restore" flag to indicate that a watcher was moved from the synced watcher group with restore operation. Otherwise, the watcher with future revision in an unsynced watcher group would still panic. Example logs with future revision watcher from restore operation: ``` {"level":"info","ts":1527196358.9057755,"caller":"mvcc/watcher_group.go:261","msg":"choosing future revision watcher from restore operation","watch-key":"proxy-namespace__lostleader","watch-revision":9223372036854775805,"current-revision":16} {"level":"info","ts":1527196358.910349,"caller":"mvcc/watcher_group.go:261","msg":"choosing future revision watcher from restore operation","watch-key":"proxy-namespace__lostleader","watch-revision":9223372036854775805,"current-revision":16} ``` Signed-off-by: Gyuho Lee <[email protected]>

…ation This also happens without gRPC proxy. Fix panic when gRPC proxy leader watcher is restored: ``` go test -v -tags cluster_proxy -cpu 4 -race -run TestV3WatchRestoreSnapshotUnsync === RUN TestV3WatchRestoreSnapshotUnsync panic: watcher minimum revision 9223372036854775805 should not exceed current revision 16 goroutine 156 [running]: github.com/coreos/etcd/mvcc.(*watcherGroup).chooseAll(0xc4202b8720, 0x10, 0xffffffffffffffff, 0x1) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watcher_group.go:242 +0x3b5 github.com/coreos/etcd/mvcc.(*watcherGroup).choose(0xc4202b8720, 0x200, 0x10, 0xffffffffffffffff, 0xc420253378, 0xc420253378) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watcher_group.go:225 +0x289 github.com/coreos/etcd/mvcc.(*watchableStore).syncWatchers(0xc4202b86e0, 0x0) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:340 +0x237 github.com/coreos/etcd/mvcc.(*watchableStore).syncWatchersLoop(0xc4202b86e0) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:214 +0x280 created by github.com/coreos/etcd/mvcc.newWatchableStore /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:90 +0x477 exit status 2 FAIL github.com/coreos/etcd/integration 2.551s ``` gRPC proxy spawns a watcher with a key "proxy-namespace__lostleader" and watch revision "int64(math.MaxInt64 - 2)" to detect leader loss. But, when the partitioned node restores, this watcher triggers panic with "watcher minimum revision ... should not exceed current ...". This check was added a long time ago, by my PR, when there was no gRPC proxy: #4043 (comment) > we can remove this checking actually. it is impossible for a unsynced watching to have a future rev. or we should just panic here. However, now it's possible that a unsynced watcher has a future revision, when it was moved from a synced watcher group through restore operation. This PR adds "restore" flag to indicate that a watcher was moved from the synced watcher group with restore operation. Otherwise, the watcher with future revision in an unsynced watcher group would still panic. Example logs with future revision watcher from restore operation: ``` {"level":"info","ts":1527196358.9057755,"caller":"mvcc/watcher_group.go:261","msg":"choosing future revision watcher from restore operation","watch-key":"proxy-namespace__lostleader","watch-revision":9223372036854775805,"current-revision":16} {"level":"info","ts":1527196358.910349,"caller":"mvcc/watcher_group.go:261","msg":"choosing future revision watcher from restore operation","watch-key":"proxy-namespace__lostleader","watch-revision":9223372036854775805,"current-revision":16} ``` Signed-off-by: Gyuho Lee <[email protected]> iii Signed-off-by: Gyuho Lee <[email protected]>

…ation This also happens without gRPC proxy. Fix panic when gRPC proxy leader watcher is restored: ``` go test -v -tags cluster_proxy -cpu 4 -race -run TestV3WatchRestoreSnapshotUnsync === RUN TestV3WatchRestoreSnapshotUnsync panic: watcher minimum revision 9223372036854775805 should not exceed current revision 16 goroutine 156 [running]: github.com/coreos/etcd/mvcc.(*watcherGroup).chooseAll(0xc4202b8720, 0x10, 0xffffffffffffffff, 0x1) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watcher_group.go:242 +0x3b5 github.com/coreos/etcd/mvcc.(*watcherGroup).choose(0xc4202b8720, 0x200, 0x10, 0xffffffffffffffff, 0xc420253378, 0xc420253378) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watcher_group.go:225 +0x289 github.com/coreos/etcd/mvcc.(*watchableStore).syncWatchers(0xc4202b86e0, 0x0) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:340 +0x237 github.com/coreos/etcd/mvcc.(*watchableStore).syncWatchersLoop(0xc4202b86e0) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:214 +0x280 created by github.com/coreos/etcd/mvcc.newWatchableStore /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:90 +0x477 exit status 2 FAIL github.com/coreos/etcd/integration 2.551s ``` gRPC proxy spawns a watcher with a key "proxy-namespace__lostleader" and watch revision "int64(math.MaxInt64 - 2)" to detect leader loss. But, when the partitioned node restores, this watcher triggers panic with "watcher minimum revision ... should not exceed current ...". This check was added a long time ago, by my PR, when there was no gRPC proxy: #4043 (comment) > we can remove this checking actually. it is impossible for a unsynced watching to have a future rev. or we should just panic here. However, now it's possible that a unsynced watcher has a future revision, when it was moved from a synced watcher group through restore operation. This PR adds "restore" flag to indicate that a watcher was moved from the synced watcher group with restore operation. Otherwise, the watcher with future revision in an unsynced watcher group would still panic. Example logs with future revision watcher from restore operation: ``` {"level":"info","ts":1527196358.9057755,"caller":"mvcc/watcher_group.go:261","msg":"choosing future revision watcher from restore operation","watch-key":"proxy-namespace__lostleader","watch-revision":9223372036854775805,"current-revision":16} {"level":"info","ts":1527196358.910349,"caller":"mvcc/watcher_group.go:261","msg":"choosing future revision watcher from restore operation","watch-key":"proxy-namespace__lostleader","watch-revision":9223372036854775805,"current-revision":16} ``` Signed-off-by: Gyuho Lee <[email protected]>

…ation This also happens without gRPC proxy. Fix panic when gRPC proxy leader watcher is restored: ``` go test -v -tags cluster_proxy -cpu 4 -race -run TestV3WatchRestoreSnapshotUnsync === RUN TestV3WatchRestoreSnapshotUnsync panic: watcher minimum revision 9223372036854775805 should not exceed current revision 16 goroutine 156 [running]: github.com/coreos/etcd/mvcc.(*watcherGroup).chooseAll(0xc4202b8720, 0x10, 0xffffffffffffffff, 0x1) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watcher_group.go:242 +0x3b5 github.com/coreos/etcd/mvcc.(*watcherGroup).choose(0xc4202b8720, 0x200, 0x10, 0xffffffffffffffff, 0xc420253378, 0xc420253378) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watcher_group.go:225 +0x289 github.com/coreos/etcd/mvcc.(*watchableStore).syncWatchers(0xc4202b86e0, 0x0) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:340 +0x237 github.com/coreos/etcd/mvcc.(*watchableStore).syncWatchersLoop(0xc4202b86e0) /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:214 +0x280 created by github.com/coreos/etcd/mvcc.newWatchableStore /home/gyuho/go/src/github.com/coreos/etcd/mvcc/watchable_store.go:90 +0x477 exit status 2 FAIL github.com/coreos/etcd/integration 2.551s ``` gRPC proxy spawns a watcher with a key "proxy-namespace__lostleader" and watch revision "int64(math.MaxInt64 - 2)" to detect leader loss. But, when the partitioned node restores, this watcher triggers panic with "watcher minimum revision ... should not exceed current ...". This check was added a long time ago, by my PR, when there was no gRPC proxy: etcd-io#4043 (comment) > we can remove this checking actually. it is impossible for a unsynced watching to have a future rev. or we should just panic here. However, now it's possible that a unsynced watcher has a future revision, when it was moved from a synced watcher group through restore operation. This PR adds "restore" flag to indicate that a watcher was moved from the synced watcher group with restore operation. Otherwise, the watcher with future revision in an unsynced watcher group would still panic. Example logs with future revision watcher from restore operation: ``` {"level":"info","ts":1527196358.9057755,"caller":"mvcc/watcher_group.go:261","msg":"choosing future revision watcher from restore operation","watch-key":"proxy-namespace__lostleader","watch-revision":9223372036854775805,"current-revision":16} {"level":"info","ts":1527196358.910349,"caller":"mvcc/watcher_group.go:261","msg":"choosing future revision watcher from restore operation","watch-key":"proxy-namespace__lostleader","watch-revision":9223372036854775805,"current-revision":16} ``` Signed-off-by: Gyuho Lee <[email protected]>

xiang90 reviewed Dec 22, 2015
View reviewed changes

gyuho force-pushed the storage_range_all_unsynced branch from 3849241 to 1945eb6 Compare December 23, 2015 00:13

xiang90 reviewed Dec 23, 2015
View reviewed changes

gyuho force-pushed the storage_range_all_unsynced branch from 1206087 to d5ce3fd Compare December 23, 2015 18:09

gyuho changed the title ~~[WIP] storage: range all unsynced at once~~ storage: range all unsynced at once Dec 23, 2015

gyuho force-pushed the storage_range_all_unsynced branch from d5ce3fd to 4be489b Compare December 23, 2015 18:13

xiang90 reviewed Dec 23, 2015
View reviewed changes

gyuho force-pushed the storage_range_all_unsynced branch 3 times, most recently from f01f143 to 41d8b62 Compare December 24, 2015 00:48

gyuho mentioned this pull request Dec 24, 2015

Enhancement: Allow watch from "first available index" vs. new updates or updates from a valid start index #1666

Closed

xiang90 reviewed Dec 27, 2015
View reviewed changes

gyuho force-pushed the storage_range_all_unsynced branch 5 times, most recently from 2f4a90a to 53e90ed Compare December 28, 2015 19:08

gyuho force-pushed the storage_range_all_unsynced branch 3 times, most recently from 03e1ce6 to 58c8124 Compare December 28, 2015 19:30

gyuho added 2 commits December 28, 2015 11:37

storage: range all unsynced at once

78b0b8a

This is for etcd-io#3848. It replaces RangeHistory method for more efficient event sending.

storage: delete RangeHistory

ecc3e15

This has been replaced by operations inside `syncWatchings`.

gyuho force-pushed the storage_range_all_unsynced branch from 58c8124 to ecc3e15 Compare December 28, 2015 19:37

gyuho added a commit that referenced this pull request Dec 28, 2015

Merge pull request #4043 from gyuho/storage_range_all_unsynced

6c5dc28

storage: range all unsynced at once

gyuho merged commit 6c5dc28 into etcd-io:master Dec 28, 2015

gyuho deleted the storage_range_all_unsynced branch January 31, 2016 00:44

gyuho mentioned this pull request May 24, 2018

mvcc: fix panic by allowing future revision watcher from restore operation #9775

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

storage: range all unsynced at once #4043

storage: range all unsynced at once #4043

gyuho commented Dec 22, 2015

xiang90 Dec 22, 2015

gyuho Dec 22, 2015

xiang90 Dec 23, 2015

gyuho Dec 23, 2015

xiang90 commented Dec 23, 2015

xiang90 Dec 23, 2015

gyuho Dec 23, 2015

xiang90 Dec 27, 2015

xiang90 commented Dec 28, 2015

storage: range all unsynced at once #4043

storage: range all unsynced at once #4043

Conversation

gyuho commented Dec 22, 2015

xiang90 Dec 22, 2015

Choose a reason for hiding this comment

gyuho Dec 22, 2015

Choose a reason for hiding this comment

xiang90 Dec 23, 2015

Choose a reason for hiding this comment

gyuho Dec 23, 2015

Choose a reason for hiding this comment

xiang90 commented Dec 23, 2015

xiang90 Dec 23, 2015

Choose a reason for hiding this comment

gyuho Dec 23, 2015

Choose a reason for hiding this comment

xiang90 Dec 27, 2015

Choose a reason for hiding this comment

xiang90 commented Dec 28, 2015