Skip to content
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

reload region cache when store is resolved from invalid status #843

Merged
merged 14 commits into from
Jul 14, 2023

Conversation

you06
Copy link
Contributor

@you06 you06 commented Jun 16, 2023

close #841

Before this PR:
image

image

With this PR:
image

image

@you06 you06 changed the title reload region cache when store is resolved from invalid reload region cache when store is resolved from invalid status Jun 16, 2023
@@ -573,18 +573,22 @@ func (state *accessFollower) next(bo *retry.Backoffer, selector *replicaSelector
if state.option.preferLeader {
state.lastIdx = state.leaderIdx
}
offset := rand.Intn(replicaSize)
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Under PreferLeader mode, offset should be set to 0 to make the selection starting from leaderIdx.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yeah, I changed the code and not to use offset in the first selection.

@cfzjywxk
Copy link
Contributor

/cc @crazycs520

// as candidates to serve the Read request. Meanwhile, we should make the choice of next() meet Uniform Distribution.
for cnt := 0; cnt < replicaSize && !state.isCandidate(idx, selector.replicas[idx]); cnt++ {
idx = AccessIndex((int(idx) + rand.Intn(replicaSize)) % replicaSize)
var idx AccessIndex
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Still recommend to revert to the original one, current implementation is bit more complex and hard to read.

Signed-off-by: you06 <[email protected]>
// async reload triggered by other thread.
return
}
go func() {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It can be risky to spawn a reload task for each region, especially if many regions become invalid due to the related store being marked as such. To mitigate this risk, we should implement a refreshing strategy in the region cache-related workers and consider incorporating the store's "health checker."

@crazycs520 is now considering refactoring the region cache component, he could give some advice about it.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

For the v6.5.x release version, a quick fix may still be needed.

@@ -573,18 +573,36 @@ func (state *accessFollower) next(bo *retry.Backoffer, selector *replicaSelector
if state.option.preferLeader {
state.lastIdx = state.leaderIdx
}
var offset int
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Peer selection can become quite complex when the accessFollower is used for all follower read and stale read processing, with various configurations. To prevent any potential issues, it may be beneficial to separate them into different categories.

ReplicaSelector itself is quite complex too..

Copy link
Contributor

@cfzjywxk cfzjywxk Jun 26, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

After discussing with @you06, we could do some refactoring on the master branch like

match access_type {
     leade_only =>
     follower_read_random =>
     follower_read_replica_only =>
     follower_read_closet => 
     stale_read_random => 
     stale_read_closest =>
     stale_read_retry =>
}

while keeping the change on release-6.5 as simple as possible to reduce the risk.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Maybe the refactor can be in another PR?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Ok let's merge this first

cfzjywxk pushed a commit that referenced this pull request Jun 28, 2023
#846)

* reload region cache when store is resolved from invalid

Signed-off-by: you06 <[email protected]>

* reload region once

Signed-off-by: you06 <[email protected]>

* build in go1.18

Signed-off-by: you06 <[email protected]>

* build in go1.18

Signed-off-by: you06 <[email protected]>

* handle region reload in resolve goroutine

Signed-off-by: you06 <[email protected]>

* retest

Signed-off-by: you06 <[email protected]>

* fix data race (#736)

Signed-off-by: Smityz <[email protected]>
Co-authored-by: disksing <[email protected]>
Signed-off-by: you06 <[email protected]>

* build with go 1.18

Signed-off-by: you06 <[email protected]>

* fix integration test (#673)

Signed-off-by: disksing <[email protected]>
Signed-off-by: you06 <[email protected]>

* Update internal/locate/region_cache.go

Co-authored-by: crazycs <[email protected]>
Signed-off-by: you06 <[email protected]>

* address comment

Signed-off-by: you06 <[email protected]>

* address comment

Signed-off-by: you06 <[email protected]>

---------

Signed-off-by: you06 <[email protected]>
Signed-off-by: Smityz <[email protected]>
Signed-off-by: disksing <[email protected]>
Co-authored-by: Smilencer <[email protected]>
Co-authored-by: disksing <[email protected]>
Co-authored-by: crazycs <[email protected]>
internal/locate/region_cache.go Show resolved Hide resolved
internal/locate/region_cache.go Outdated Show resolved Hide resolved
@cfzjywxk cfzjywxk requested a review from ekexium June 29, 2023 07:18
Signed-off-by: you06 <[email protected]>
@you06 you06 mentioned this pull request Jul 5, 2023
@disksing disksing merged commit 85fc8f3 into tikv:master Jul 14, 2023
iosmanthus added a commit that referenced this pull request Aug 11, 2023
* client-go: add some key range info to error when PD returned no region (#862)

Signed-off-by: Chao Wang <[email protected]>

* *: refine non-global stale-read request retry logic (#863)

Signed-off-by: crazycs520 <[email protected]>

* Fix the issue that primary pessimistic lock may be left not cleared after GC (#866)

* Fix the issue that primary pessimistic lock may be left not cleared after GC

Signed-off-by: MyonKeminta <[email protected]>

* Fix mysteriously shown up thing that makes compilation failed

Signed-off-by: MyonKeminta <[email protected]>

* Fix test effectiveness (forgot to set txn2 to pessimistic txn); add more strict checks

Signed-off-by: MyonKeminta <[email protected]>

* Address comments

Signed-off-by: MyonKeminta <[email protected]>

---------

Signed-off-by: MyonKeminta <[email protected]>
Co-authored-by: MyonKeminta <[email protected]>

* add explicit request source type to label the external request like lightning/br (#868)

Signed-off-by: nolouch <[email protected]>

* use '%d' instead of '%q' for some int values in error message (#875)

Signed-off-by: Chao Wang <[email protected]>

* format key in error message in method `scanRegions` (#876)

Signed-off-by: Chao Wang <[email protected]>

* make cop request timeout a config paramter (#865)

* update

Signed-off-by: Spade A <[email protected]>

* update

Signed-off-by: Spade A <[email protected]>

* update

Signed-off-by: Spade A <[email protected]>

* update

Signed-off-by: Spade A <[email protected]>

---------

Signed-off-by: Spade A <[email protected]>

* region_cache: support check pending tiflash peer (#821)

Signed-off-by: guo-shaoge <[email protected]>
Co-authored-by: disksing <[email protected]>

* *: add `SnapshotIterReverse` and make `iterReverse` supports `lowerBound` (#883)

Signed-off-by: Jason Mo <[email protected]>

* *: fix stale read ops metric (#878) (#889)

Signed-off-by: crazycs520 <[email protected]>
Co-authored-by: disksing <[email protected]>

* add gc options (#828)

Signed-off-by: weedge <[email protected]>
Co-authored-by: disksing <[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]>

* ci: update setup-go action (#904)

Signed-off-by: disksing <[email protected]>

* fix unexpected slow query during GC running after stop 1 tikv-server (#899) (#909)

* fix unexpected slow query during GC running after stop 1 tikv-server

Signed-off-by: crazycs520 <[email protected]>

* fix test

Signed-off-by: crazycs520 <[email protected]>

---------

Signed-off-by: crazycs520 <[email protected]>

* resource_manager: ignore ru metrics for background request (#872)

Signed-off-by: husharp <[email protected]>
Co-authored-by: disksing <[email protected]>

* add more log for diagnose (#915)

* add more log for diagnose

Signed-off-by: crazycs520 <[email protected]>

* fix

Signed-off-by: crazycs520 <[email protected]>

* add more log for diagnose

Signed-off-by: crazycs520 <[email protected]>

* add more log

Signed-off-by: crazycs520 <[email protected]>

* address comment

Signed-off-by: crazycs520 <[email protected]>

---------

Signed-off-by: crazycs520 <[email protected]>

* use context logger as much as possible (#908)

* use context logger as much as possible

Signed-off-by: crazycs520 <[email protected]>

* refine

Signed-off-by: crazycs520 <[email protected]>

---------

Signed-off-by: crazycs520 <[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]>

* request_source: remove default label (#890)

* request_source: remove default label

Signed-off-by: nolouch <[email protected]>

* add a function to set request source task type (#925)

* add a function to set request source task type

Signed-off-by: glorv <[email protected]>

* ci: update go version (#936)

* ci: update go version

Signed-off-by: crazycs520 <[email protected]>

* fix test

Signed-off-by: crazycs520 <[email protected]>

---------

Signed-off-by: crazycs520 <[email protected]>

* use tidb_kv_read_timeout as first kv request timeout (#919)

* support tidb_kv_read_timeout as first round kv request timeout

Signed-off-by: crazycs520 <[email protected]>

* fix ci

Signed-off-by: crazycs520 <[email protected]>

* fix ci

Signed-off-by: crazycs520 <[email protected]>

* fix ci

Signed-off-by: crazycs520 <[email protected]>

* fix ci

Signed-off-by: crazycs520 <[email protected]>

* fix ci

Signed-off-by: crazycs520 <[email protected]>

* update comment

Signed-off-by: crazycs520 <[email protected]>

* refine test

Signed-off-by: crazycs520 <[email protected]>

---------

Signed-off-by: crazycs520 <[email protected]>

* [pick] resource_control: bypass some internal urgent request (#938)

* resource_control: bypass some internal urgent request (#884)

Signed-off-by: nolouch <[email protected]>

* resourcecontrol: fix nil pointer (#900)

Signed-off-by: nolouch <[email protected]>

---------

Signed-off-by: nolouch <[email protected]>

---------

Signed-off-by: Chao Wang <[email protected]>
Signed-off-by: crazycs520 <[email protected]>
Signed-off-by: MyonKeminta <[email protected]>
Signed-off-by: nolouch <[email protected]>
Signed-off-by: Spade A <[email protected]>
Signed-off-by: guo-shaoge <[email protected]>
Signed-off-by: Jason Mo <[email protected]>
Signed-off-by: weedge <[email protected]>
Signed-off-by: you06 <[email protected]>
Signed-off-by: disksing <[email protected]>
Signed-off-by: husharp <[email protected]>
Signed-off-by: cfzjywxk <[email protected]>
Signed-off-by: glorv <[email protected]>
Signed-off-by: iosmanthus <[email protected]>
Co-authored-by: 王超 <[email protected]>
Co-authored-by: crazycs <[email protected]>
Co-authored-by: MyonKeminta <[email protected]>
Co-authored-by: MyonKeminta <[email protected]>
Co-authored-by: ShuNing <[email protected]>
Co-authored-by: Spade  A <[email protected]>
Co-authored-by: guo-shaoge <[email protected]>
Co-authored-by: disksing <[email protected]>
Co-authored-by: Hangjie Mo <[email protected]>
Co-authored-by: weedge <[email protected]>
Co-authored-by: you06 <[email protected]>
Co-authored-by: Hu# <[email protected]>
Co-authored-by: cfzjywxk <[email protected]>
Co-authored-by: glorv <[email protected]>
you06 added a commit to you06/client-go that referenced this pull request Aug 11, 2023
you06 added a commit to you06/client-go that referenced this pull request Aug 15, 2023
you06 added a commit to you06/client-go that referenced this pull request Aug 15, 2023
cfzjywxk added a commit that referenced this pull request Sep 26, 2023
* 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]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

need to reload region cache when invalid store become health
5 participants