-
Notifications
You must be signed in to change notification settings - Fork 225
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
add region cache state test & fix some issues of replica selector #910
Conversation
9bbea85
to
533d7eb
Compare
Signed-off-by: you06 <[email protected]>
e607df2
to
5efef85
Compare
Signed-off-by: you06 <[email protected]>
Signed-off-by: you06 <[email protected]>
/cc @crazycs520 |
// In stale-read, the request will fallback to leader after the local follower failure. | ||
// If the leader is also unavailable, we can fallback to the follower and use replica-read flag again, | ||
// The remote follower not tried yet, and the local follower can retry without stale-read flag. | ||
if state.isStaleRead { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This could be printed in the above log so we know if tryFollower
happens in current turn, so is the fallbackFromLeader
state.
@@ -609,14 +615,30 @@ func (state *accessFollower) next(bo *retry.Backoffer, selector *replicaSelector | |||
// If there is no candidate, fallback to the leader. | |||
if selector.targetIdx < 0 { | |||
leader := selector.replicas[state.leaderIdx] | |||
leaderInvalid := leader.isEpochStale() || state.IsLeaderExhausted(leader) | |||
leaderEpochStale := leader.isEpochStale() | |||
leaderInvalid := leaderEpochStale || state.IsLeaderExhausted(leader) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Maybe we could refactor the implememtation of state.IsLeaderExhausted
, cheking if the leader is exausted by the current states instead of leader.isExhausted(2)
.
|
||
resp, _, err := s.regionRequestSender.SendReqCtx(bo, req, regionLoc.Region, time.Second, tikvrpc.TiKV, ops...) | ||
if !available { | ||
if err != nil { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Is this err
not checked if it's expected or not?
…kv#910) Signed-off-by: you06 <[email protected]>
…kv#910) Signed-off-by: you06 <[email protected]> remove duplicate code Signed-off-by: you06 <[email protected]>
…kv#910) Signed-off-by: you06 <[email protected]> remove duplicate code Signed-off-by: you06 <[email protected]>
…kv#910) Signed-off-by: you06 <[email protected]> remove duplicate code Signed-off-by: you06 <[email protected]>
…kv#910) Signed-off-by: you06 <[email protected]> remove duplicate code Signed-off-by: you06 <[email protected]>
…) (#942) * add region cache state test & fix some issues of replica selector (#910) Signed-off-by: you06 <[email protected]> remove duplicate code Signed-off-by: you06 <[email protected]> * remove comment Signed-off-by: you06 <[email protected]> * lint Signed-off-by: you06 <[email protected]> * fix flaky test Signed-off-by: you06 <[email protected]> --------- Signed-off-by: you06 <[email protected]>
* reload region cache when store is resolved from invalid status (#843) Signed-off-by: you06 <[email protected]> Co-authored-by: disksing <[email protected]> * fallback to follower when leader is busy (#916) (#923) * fallback to follower when leader is busy Signed-off-by: you06 <[email protected]> Co-authored-by: cfzjywxk <[email protected]> Co-authored-by: cfzjywxk <[email protected]> * Resume max retry time check for stale read retry with leader option(#903) (#911) * Resume max retry time check for stale read retry with leader option Signed-off-by: cfzjywxk <[email protected]> * add cancel Signed-off-by: cfzjywxk <[email protected]> --------- Signed-off-by: cfzjywxk <[email protected]> * add region cache state test & fix some issues of replica selector (#910) Signed-off-by: you06 <[email protected]> remove duplicate code Signed-off-by: you06 <[email protected]> * enable workflow for tidb-7.1 Signed-off-by: you06 <[email protected]> * update Signed-off-by: you06 <[email protected]> update Signed-off-by: you06 <[email protected]> fix test Signed-off-by: you06 <[email protected]> fix test Signed-off-by: you06 <[email protected]> * lint Signed-off-by: you06 <[email protected]> * lint Signed-off-by: you06 <[email protected]> * fix flaky test Signed-off-by: you06 <[email protected]> --------- Signed-off-by: you06 <[email protected]> Signed-off-by: cfzjywxk <[email protected]> Co-authored-by: disksing <[email protected]> Co-authored-by: cfzjywxk <[email protected]> Co-authored-by: cfzjywxk <[email protected]>
Co-authored-by: cfzjywxk <[email protected]> Co-authored-by: cfzjywxk <[email protected]> Co-authored-by: disksing <[email protected]> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: zzm <[email protected]> Co-authored-by: husharp <[email protected]> Co-authored-by: you06 <[email protected]> Co-authored-by: buffer <[email protected]> Co-authored-by: 3pointer <[email protected]> Co-authored-by: buffer <[email protected]> Co-authored-by: husharp <[email protected]> Co-authored-by: crazycs520 <[email protected]> Co-authored-by: Smilencer <[email protected]> Co-authored-by: ShuNing <[email protected]> Co-authored-by: zyguan <[email protected]> Co-authored-by: Jack Yu <[email protected]> Co-authored-by: Weizhen Wang <[email protected]> Co-authored-by: lucasliang <[email protected]> Co-authored-by: healthwaite <[email protected]> Co-authored-by: xufei <[email protected]> Co-authored-by: JmPotato <[email protected]> Co-authored-by: ekexium <[email protected]> Co-authored-by: 山岚 <[email protected]> Co-authored-by: glorv <[email protected]> Co-authored-by: Yongbo Jiang <[email protected]> resolve locks interface for tidb gc_worker (#945) fix some issues of replica selector (#910) (#942) fix some issues of replica selector (#910) fix issue of configure kv timeout not work when disable batch client (#980) fix batch-client wait too long and add some metrics (#973) fix batch-client wait too long and add some metrics (#973)" (#984) fix data race at the aggressiveLockingDirty (#913) fix MinSafeTS might be set to MaxUint64 permanently (#994) fix: fix invalid nil pointer when trying to record Store.SlownessStat. (#1017) Fix batch client batchSendLoop panic (#1021) fix request source tag unset (#1025) Fix comment of `SuspendTime` (#1057)
This PR add the selector replica test for stale read, cover the most possible errors, and fix the bug the test finds:
Fixes: