Remove stale-read flag after resolving lock #792
Merged
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
Signed-off-by: you06 <you1474600@gmail.com>
Signed-off-by: you06 <you1474600@gmail.com>
Signed-off-by: you06 <you1474600@gmail.com>
Signed-off-by: you06 <you1474600@gmail.com>
Signed-off-by: you06 <you1474600@gmail.com>
cfzjywxk
approved these changes
May 25, 2023
tikvrpc/tikvrpc.go
Outdated
| @@ -275,6 +275,12 @@ func (req *Request) EnableStaleRead() { | |||
| req.ReplicaRead = false | |||
| } | |||
|
|
|||
| // DisableStaleRead is called when stale-read fallbacks to leader read after meeting key-is-locked error. | |||
| func (req *Request) DisableStaleRead() { | |||
There was a problem hiding this comment.
How about DisableStaleReadAfterFallback
There was a problem hiding this comment.
Rename it with DisableStaleReadMeetLock because it should only take effect when key is locked, other rollback won't affect this flag
Signed-off-by: you06 <you1474600@gmail.com>
…disable-stale-read
zyguan
approved these changes
May 26, 2023
|
/test |
Signed-off-by: you06 <you1474600@gmail.com>
you06
added a commit
to you06/client-go
that referenced
this pull request
May 26, 2023
* remove stale-read when key-is-locked Signed-off-by: you06 <you1474600@gmail.com> * disable follower read for batch get Signed-off-by: you06 <you1474600@gmail.com> * the stale-read flag may already set to false Signed-off-by: you06 <you1474600@gmail.com> * fix batchget Signed-off-by: you06 <you1474600@gmail.com> * Reset busy-threshold when stale read fallback to leader read Signed-off-by: you06 <you1474600@gmail.com> * address comment Signed-off-by: you06 <you1474600@gmail.com> * trigger CI Signed-off-by: you06 <you1474600@gmail.com> --------- Signed-off-by: you06 <you1474600@gmail.com>
This was referenced May 26, 2023
cfzjywxk
pushed a commit
that referenced
this pull request
May 29, 2023
* add metrics for stale-read traffic (#776) Signed-off-by: you06 <you1474600@gmail.com> * Remove stale-read flag after resolving lock (#792) * remove stale-read when key-is-locked Signed-off-by: you06 <you1474600@gmail.com> * disable follower read for batch get Signed-off-by: you06 <you1474600@gmail.com> * the stale-read flag may already set to false Signed-off-by: you06 <you1474600@gmail.com> * fix batchget Signed-off-by: you06 <you1474600@gmail.com> * Reset busy-threshold when stale read fallback to leader read Signed-off-by: you06 <you1474600@gmail.com> * address comment Signed-off-by: you06 <you1474600@gmail.com> * trigger CI Signed-off-by: you06 <you1474600@gmail.com> --------- Signed-off-by: you06 <you1474600@gmail.com> --------- Signed-off-by: you06 <you1474600@gmail.com>
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.
Ref pingcap/tidb#43659
Remove the stale-read flag after resolving lock.
Before this PR:
After this PR:
Because the request struct is reused when retrying in get, it'll suffer not leader error, because the stale read field is already reset from the context patcher. But the request is regenerated in batch get, so it'll meet data is not ready instead. For both cases, this PR reduces unnecessary RPC.