server: try to replace the down replicas instead of removing directly #1222
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
disksing
reviewed
Aug 30, 2018
| if down && sum < 1200 { | ||
| // only need to print once | ||
| down = false | ||
| simutil.Logger.Error("making up replicas don't start immediately") |
There was a problem hiding this comment.
Can we abstract a statistic to evaluate different strategies instead of just print a log message? For example, we may count the number of regions with insufficient replicas in the scheduling process.
server/schedule/replica_checker.go
Outdated
| @@ -170,7 +172,8 @@ func (r *ReplicaChecker) checkDownPeer(region *core.RegionInfo) *Operator { | |||
| if stats.GetDownSeconds() < uint64(r.cluster.GetMaxStoreDownTime().Seconds()) { | |||
| continue | |||
| } | |||
| return CreateRemovePeerOperator("removeDownReplica", r.cluster, OpReplica, region, peer.GetStoreId()) | |||
|
|
|||
| return r.handleReplica(region, peer, "Down") | |||
There was a problem hiding this comment.
Maybe replacePeer or fixPeer.
server/schedule/replica_checker.go
Outdated
| @@ -251,3 +232,33 @@ func (r *ReplicaChecker) checkBestReplacement(region *core.RegionInfo) *Operator | |||
| checkerCounter.WithLabelValues("replica_checker", "new_operator").Inc() | |||
| return CreateMovePeerOperator("moveToBetterLocation", r.cluster, region, OpReplica, oldPeer.GetStoreId(), newPeer.GetStoreId(), newPeer.GetId()) | |||
| } | |||
|
|
|||
| func (r *ReplicaChecker) handleReplica(region *core.RegionInfo, peer *metapb.Peer, status string) *Operator { | |||
| removeExtra := fmt.Sprintf("%s%s%s", "removeExtra", status, "Replica") | |||
There was a problem hiding this comment.
How about fmt.Sprintf("removeExtra%sReplica", status)?
siddontang
reviewed
Aug 31, 2018
| @@ -251,3 +232,33 @@ func (r *ReplicaChecker) checkBestReplacement(region *core.RegionInfo) *Operator | |||
| checkerCounter.WithLabelValues("replica_checker", "new_operator").Inc() | |||
| return CreateMovePeerOperator("moveToBetterLocation", r.cluster, region, OpReplica, oldPeer.GetStoreId(), newPeer.GetStoreId(), newPeer.GetId()) | |||
| } | |||
|
|
|||
| func (r *ReplicaChecker) fixPeer(region *core.RegionInfo, peer *metapb.Peer, status string) *Operator { | |||
There was a problem hiding this comment.
what does fixPeer mean?
There was a problem hiding this comment.
It means this peer needs to be removed in order to be made up in another store or just simply be replaced. The name before addressing the comment is handleReplica. There could be a better naming about this.
rleungx
force-pushed
the
down-replica
branch
from
9878633
to
7381274
Compare
disksing
approved these changes
Sep 4, 2018
|
PTAL @nolouch @siddontang |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What problem does this PR solve?
At present, if a store is down, it will remove peers in the store directly, and then make up replicas. It usually needs some time because it will traverse all the regions.
What is changed and how it works?
This PR tries to move the peers to another store if the original store is down.
Besides, this PR adds a case in the simulator to simulate this kind of situation. It will print an error log without these changes:
making up replicas don't start immediately.Check List
Tests