cherrypick data affinity feature to release-8.5 by lhy1024 · Pull Request #10061 · tikv/pd

lhy1024 · 2025-12-14T11:04:14Z

Conflict Resolution Details

Commit1: PR #9993 - Affinity Storage & Config

Conflict 1: pkg/slice/slice.go and pkg/slice/slice_test.go

Resolution: Accept both

<<<<<<< HEAD
=======
// HasDupInSorted takes a sorted slice and checks whether it contains any duplicate elements.
func HasDupInSorted[T comparable](sortedSlice []T) bool { ... }
>>>>>>> 12d3952d4

Conflict 2: pkg/storage/storage.go

Resolution: Remove MaintenanceStorage

<<<<<<< HEAD
=======
endpoint.MaintenanceStorage
endpoint.AffinityStorage
>>>>>>> 12d3952d4

Conflict 3: Keypath Architecture ⚠️ Critical

Resolution: Refactor keypath implementation

AffinityGroupIDPath returns relative path rather than full path (consistent with RegionLabelKeyPath)
LoadAllAffinityGroups uses keypath.AffinityGroupPath+"/" because loadRangeByPrefix will add PDRootPath() prefix

Master branch implementation:

package keypath

import (
	"path"
)

// AffinityGroupIDPath returns the path to save a specific affinity group object.
// Its format is: "/pd/{cluster_id}/affinity_groups/{group_id}"
func AffinityGroupIDPath(groupID string) string {
	return path.Join(PDRootPath(), AffinityGroupPath, groupID)
}

// AffinityGroupsPathPrefix returns the path prefix for all affinity groups.
// Its format is: "/pd/{cluster_id}/affinity_groups/"
// This is used for loading all groups via a range read.
func AffinityGroupsPathPrefix() string {
	return path.Join(PDRootPath(), AffinityGroupPath)
}

Release-8.5 implementation:

// AffinityGroupIDPath returns the path to save a specific affinity group object.
// Its format is: "/pd/{cluster_id}/affinity_groups/{group_id}"
func AffinityGroupIDPath(groupID string) string {
	return path.Join(AffinityGroupPath, groupID)
}

// AffinityGroupsPathPrefix returns the path prefix for all affinity groups.
// Its format is: "/pd/{cluster_id}/affinity_groups/"
// This is used for loading all groups via a range read.
func AffinityGroupsPathPrefix() string {
	return path.Join(PDRootPath(), AffinityGroupPath)
}

I checked it in etcd-ctl. They are the same.
8.5

master

Commit2-4: PR #9997, #9998, #9999

✅ No conflicts

Commit5: PR #10038 - Affinity Filter & kvproto

Conflict 1: go.sum and go.mod

Resolution: Update kvproto version

Conflict 2: pkg/schedule/core/cluster_informer.go

Resolution:

Kept old signature: AllocID() (uint64, error)
Added new method: GetAffinityManager() *affinity.Manager

Conflict 3: pkg/mock/mockcluster/mockcluster.go, pkg/schedule/schedulers/balance_leader_test.go, pkg/schedule/schedulers/balance_region_test.go

Resolution: Kept private field keyRangeManager with getter method

Conflict 4: hot_region_solver.go ⚠️ File Migration (Critical)

Resolution: Migrated affinity filter logic to hot_region.go:622-640

affinityFilter := filter.NewAffinityFilter(bs.SchedulerCluster)
if !affinityFilter.Select(bs.cur.region).IsOK() {
	hotSchedulerIgnoredAffinity.Inc()
	continue
}

Conflict 5: server/server.go:2155-2172

Resolution: Kept both methods

GetGlobalTSOAllocator()
GetAffinityManager()

Conflict 6: pkg/schedule/labeler/labeler.go:436-445 and pkg/core/store_option.go:101-108

Resolution: Copied function implementation from master

Commit6: PR #10050 - Refactor

✅ No conflicts

Commit7: PR #10041 - API & Client Support

Conflict 1: client/http/api.go and server/apiv2/router.go

Resolution:

Only add affinity-related APIs
Discard GetKeyspaceMetaByID and RegisterMaintenance

Conflict 2: tests/integrations/client/http_client_test.go

Resolution: Remove unrelated TestGetSiblingsRegions

Conflict 3: tests/server/apiv2/handlers/affinity_test.go

Resolution: Replace test framework functions

Function Type	Master (Microservice)	Release-8.5 (API Mode)
Both modes	`RunTest()`	`RunTestBasedOnMode()`
PD mode only	`RunTestInNonMicroserviceEnv()`	`RunTestInPDMode()`
MCS/API mode	`RunTestInMicroserviceEnv()`	`RunTestInAPIMode()`

Conflict 4: Naming Convention

Resolution: Use MicroService (capital S) rather than Microservice

Commit8: PR #10040 - Affinity Checker

Conflict 1: pkg/schedule/checker/metrics.go

Incoming changes:

<<<<<<< HEAD
=======
	affinityChecker   = "affinity_checker"
)

const (
	// patrol phases
	phaseWaitForChannel = "wait_for_channel"
	phaseCheckPriority  = "check_priority"
	phaseCheckPending   = "check_pending"
	phaseScanRegions    = "scan_regions"
	phaseUpdateLabel    = "update_label_stats"
)

// checkerControllerMetrics contains pre-created Prometheus metrics for the checker controller.
type checkerControllerMetrics struct {
	patrolPhaseHistograms map[string]prometheus.Observer
	checkRegionHistograms map[string]prometheus.Observer
	patrolRegionChannelSize prometheus.Gauge
}

func newCheckerControllerMetrics() *checkerControllerMetrics { ... }
>>>>>>> 4a0b4051a

Resolution: Only keep affinity-related metrics, discard patrol phases and controller metrics

Conflict 2: pkg/schedule/checker/checker_controller.go

Incoming changes:

<<<<<<< HEAD
	return &Controller{
=======
	ruleManager := cluster.GetRuleManager()
	c := &Controller{
>>>>>>> 4a0b4051a

Resolution: Accept the ruleManager initialization

Conflict 3: Checker Integration - measureChecker wrapper ⚠️ Critical

Incoming changes:

<<<<<<< HEAD
	if c.mergeChecker != nil {
		allowed := opController.OperatorCount(operator.OpMerge) < c.conf.GetMergeScheduleLimit()
		if !allowed {
			operator.IncOperatorLimitCounter(c.mergeChecker.GetType(), operator.OpMerge)
		} else if ops := c.mergeChecker.Check(region); ops != nil {
			// It makes sure that two operators can be added successfully altogether.
			return ops
=======
	if ops := measureChecker(c.metrics.checkRegionHistograms[affinityChecker], func() []*operator.Operator {
		if opController.OperatorCount(operator.OpAffinity) < c.conf.GetAffinityScheduleLimit() {
			// It makes sure that two affinity merge operators can be added successfully altogether.
			return c.affinityChecker.Check(region)
>>>>>>> fdc1cdf23 (schedule: add affinity checker (#10040))

Resolution: Remove measureChecker wrapper, keep release-8.5 style

Final implementation:

// Check affinity checker (after rule/replica checker, before merge checker)
if opController.OperatorCount(operator.OpAffinity) < c.conf.GetAffinityScheduleLimit() {
	// It makes sure that two affinity merge operators can be added successfully altogether.
	if ops := c.affinityChecker.Check(region); len(ops) > 0 {
		return ops
	}
} else {
	operator.IncOperatorLimitCounter(c.affinityChecker.GetType(), operator.OpAffinity)
}

// Keep release-8.5 merge checker style (no measureChecker)
if c.mergeChecker != nil {
	allowed := opController.OperatorCount(operator.OpMerge) < c.conf.GetMergeScheduleLimit()
	if !allowed {
		operator.IncOperatorLimitCounter(c.mergeChecker.GetType(), operator.OpMerge)
	} else if ops := c.mergeChecker.Check(region); ops != nil {
		return ops
	}
}
return nil

Conflict 4: API Method

Resolution: Replace AllowLeaderTransferIn() with AllowLeaderTransfer()

Commit9: PR #10043 - pd-ctl Support

✅ No conflicts

Commit9: PR #10081 - merge fix

replace suite.newRegionInfor with RegionInfo

Commit10: PR #10080 - other fix

Conflict 1: Function Name

Resolution: Replace AllowLeaderTransferIn() with AllowLeaderTransfer()

<<<<<<< HEAD
		case !store.AllowLeaderTransfer() || m.conf.CheckLabelProperty(config.RejectLeader, store.GetLabels()):
			unavailableStores[store.GetID()] = storeEvictLeader
=======
		case !store.AllowLeaderTransferIn() || m.conf.CheckLabelProperty(config.RejectLeader, store.GetLabels()) ||
			store.EvictedAsSlowStore() || store.EvictedAsStoppingStore() || store.IsEvictedAsSlowTrend():
			unavailableStores[store.GetID()] = storeLeaderEvicted
		case store.IsBusy():
			unavailableStores[store.GetID()] = storeBusy
>>>>>>> b53de7a81 (affinity: add scatter filter, add more evict check and avoid statistic miss (#10080))

Conflict 2: Other PR

Resolution: Keep old

<<<<<<< HEAD
	// If there is an old operator, replace it. The priority should be checked
	// already.
	if oldi, ok := oc.operators.Load(regionID); ok {
		old := oldi.(*Operator)
		_ = oc.removeOperatorInner(old)
		_ = old.Replace()
		oc.buryOperator(old)
=======
	old, loaded := oc.operators.LoadOrStore(regionID, op)
	if loaded {
		// If there is an old operator and it has lower priority, replace it
		oldOp := old.(*Operator)
		if !isHigherPriorityOperator(op, oldOp) {
			log.Debug("operator already exists with higher or equal priority",
				zap.Uint64("region-id", regionID),
				zap.Reflect("old", oldOp),
				zap.Reflect("new", op))
			_ = op.Cancel(AlreadyExist)
			oc.buryOperator(op)
			operatorCounter.WithLabelValues(op.Desc(), "redundant").Inc()
			return false
		}
		// replace old operator
		if !oc.operators.CompareAndSwap(regionID, oldOp, op) {
			_ = op.Cancel()
			oc.buryOperator(op)
			log.Debug("operator changed during replace, skip this add",
				zap.Uint64("region-id", regionID),
				zap.Reflect("old", oldOp),
				zap.Reflect("new", op))
			return false
		}
		oc.counts.dec(oldOp.SchedulerKind())
		oc.ack(oldOp)
		if oldOp.HasRelatedMergeRegion() {
			oc.removeRelatedMergeOperator(oldOp)
		}
		_ = oldOp.Replace()
		oc.buryOperator(oldOp)
>>>>>>> b53de7a81 (affinity: add scatter filter, add more evict check and avoid statistic miss (#10080))

<<<<<<< HEAD
=======
	"github.com/tikv/pd/pkg/utils/keyutil"
	"github.com/tikv/pd/pkg/utils/testutil"
>>>>>>> b53de7a81 (affinity: add scatter filter, add more evict check and avoid statistic mis

<<<<<<< HEAD
func (suite *httpClientTestSuite) TestGetSiblingsRegions() {
========
.....
>>>>>>>>

Conflict 3: OpAdmin⚠️ Critical

Keep OpAdmin|OpMerge, it changed to only OpAdmin in master.
But it still be OpAdmin|OpMerge in release 8.5.

Conflict 4: Rename
replace PauseLeaderTransferIn with PauseLeaderTransfer

What problem does this PR solve?

Issue Number: Close #9764

What is changed and how does it work?

Check List

Tests

Unit test
Integration test
Manual test (add detailed scripts or steps below)

Release note

None.

ti-chi-bot · 2025-12-14T11:04:17Z

Skipping CI for Draft Pull Request.
If you want CI signal for your change, please convert it to an actual PR.
You can still manually trigger a test run with /test all

ti-chi-bot · 2025-12-14T11:04:17Z

This cherry pick PR is for a release branch and has not yet been approved by triage owners.
Adding the do-not-merge/cherry-pick-not-approved label.

To merge this cherry pick:

It must be approved by the approvers firstly.
AFTER it has been approved by approvers, please wait for the cherry-pick merging approval from triage owners.

Details

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

codecov · 2025-12-14T11:10:17Z

Codecov Report

❌ Patch coverage is 75.70742% with 455 lines in your changes missing coverage. Please review.
✅ Project coverage is 77.86%. Comparing base (c5362f0) to head (cad0ed5).
⚠️ Report is 2 commits behind head on release-8.5.

Additional details and impacted files

@@               Coverage Diff               @@
##           release-8.5   #10061      +/-   ##
===============================================
- Coverage        77.95%   77.86%   -0.09%     
===============================================
  Files              471      482      +11     
  Lines            63502    65343    +1841     
===============================================
+ Hits             49505    50882    +1377     
- Misses           10370    10725     +355     
- Partials          3627     3736     +109

Flag	Coverage Δ
unittests	`77.86% <75.70%> (-0.09%)`	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

…tikv#9993) ref tikv#9764 Signed-off-by: lhy1024 <admin@liudos.us>

ref tikv#9764 Signed-off-by: lhy1024 <admin@liudos.us>

ref tikv#9764 Signed-off-by: lhy1024 <admin@liudos.us> Co-authored-by: 混沌DM <hundundm@gmail.com> Signed-off-by: lhy1024 <admin@liudos.us>

…anager (tikv#10038) ref tikv#9764 Signed-off-by: lhy1024 <admin@liudos.us>

close tikv#9764 Signed-off-by: lhy1024 <admin@liudos.us>

ref tikv#9764 Signed-off-by: lhy1024 <admin@liudos.us>

bufferflies · 2025-12-19T02:01:50Z

+	// UpdateAffinityGroupPeers updates the leader and voter stores of an affinity group.
+	UpdateAffinityGroupPeers(ctx context.Context, groupID string, leaderStoreID uint64, voterStoreIDs []uint64) (*AffinityGroupState, error)
+	// DeleteAffinityGroup deletes an affinity group by group ID.
+	DeleteAffinityGroup(ctx context.Context, groupID string, force bool) error


pls note when we should force or not.

we should force when the groups contain key range. I think we can add it in another pr.

bufferflies · 2025-12-19T02:11:06Z

+	var state AffinityGroupState
+	err = c.request(ctx, newRequestInfo().
+		WithName("UpdateAffinityGroupPeers").
+		WithURI(fmt.Sprintf(AffinityGroupByID, groupID)).


How about extracting a new function, such as WithURI(RegionByID(regionID))?

Ok, But it doesn't affect correctness. I've documented it, and I'll complete them in another PR.

bufferflies · 2025-12-19T02:12:43Z

+		return errors.Trace(err)
+	}
+	// Use POST with ?delete query parameter for batch deletion
+	url := AffinityGroups + "?delete"


bufferflies · 2025-12-19T02:50:14Z

 	}

+	// Check if region is in an affinity group that doesn't allow regular scheduling
+	if !r.affinityFilter.Select(region).IsOK() {


If it returns an error, the client will retry the scatter, but it always fails. So we should return nil for the affinity region

ok, I fixed it in #10091 (comment)

bufferflies · 2025-12-19T03:05:23Z

+			Subsystem: "affinity",
+			Name:      "status",
+			Help:      "Status of the affinity manager.",
+		}, []string{"type"})


We need to add a group label to check which group has problems.

I worry that there are lots of group metrics. We can query them in SQL

okJiang · 2025-12-19T02:59:07Z

 	defaultReplicaScheduleLimit   = 64
 	defaultMergeScheduleLimit     = 8
 	defaultHotRegionScheduleLimit = 4
+	defaultAffinityScheduleLimit  = 0 // default to disable


During normal use, what value will this be set to? In addition, what is the specific function of this parameter? Controlling the rate?

It controls affinity-move-peer and affinity-merge-region. We can use the value from leader-schedule-limit and merge-schedule-limit.

So I think 4 or 8 is ok. It also performed well in local testing.

When defaultAffinityScheduleLimit is 0, it means affinity scheduling is disabled.

When there is already a switch to control the feature, do we still need to set this default to 0?

I think it is safe to set defaultAffinityScheduleLimit to 0. Because we need to cherry pick it to 8.5.5.

And we now use it more in poc scenario. We should avoid to affect other users.

there is already a switch to control the feature

We use affinity-schedule-limit as a switch, and there is no other config to control it.

In checker PR, we replace affinity-schedule-enable with affinity-schedule-limit

ref tikv#9764 Signed-off-by: lhy1024 <admin@liudos.us>

okJiang · 2025-12-19T04:06:35Z

+// AffinityGroupsPathPrefix returns the path prefix for all affinity groups.
+// Its format is: "/pd/{cluster_id}/affinity_groups/"
+// This is used for loading all groups via a range read.
+func AffinityGroupsPathPrefix() string {


We can add this when using it.

In the furture, we will pick mcs for it. We will use AffinityGroupsPathPrefix in mcs.

I think we could keep it. And it has no any side effects

rleungx · 2025-12-19T06:17:18Z

 }

+// SetNodeState sets the node state for the store.
+func SetNodeState(nodeState metapb.NodeState) StoreCreateOption {


Can we directly use SetStoreState?

SetStoreState cannot set NodeState_Preparing, and it only be used in test file.

Please also add a comment.

ti-chi-bot · 2025-12-19T07:49:08Z

[LGTM Timeline notifier]

Timeline:

2025-12-19 03:07:04.740380974 +0000 UTC m=+1787969.554158536: ☑️ agreed by bufferflies.
2025-12-19 07:49:07.044806781 +0000 UTC m=+1804891.858584343: ☑️ agreed by rleungx.

lhy1024 · 2025-12-19T08:20:35Z

@okJiang @niubell PTAL

ti-chi-bot · 2025-12-19T08:39:00Z

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: bufferflies, HunDunDM, niubell, okJiang, rleungx

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Details

Needs approval from an approver in each of these files:

~~OWNERS~~ [HunDunDM,bufferflies,rleungx]
~~pkg/mcs/scheduling/server/config/OWNERS~~ [niubell]
~~pkg/schedule/config/OWNERS~~ [niubell]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

lhy1024 force-pushed the release-8.5-affinity branch 6 times, most recently from 7f1d976 to b3248da Compare December 16, 2025 12:52

lhy1024 and others added 7 commits December 16, 2025 20:52

affinity: add affinity storage, new config, label plan and error code (…

cc80c65

…tikv#9993) ref tikv#9764 Signed-off-by: lhy1024 <admin@liudos.us>

affinity: add base models and manager skeleton (tikv#9997)

111360b

ref tikv#9764 Signed-off-by: lhy1024 <admin@liudos.us>

affinity: add txn and key range operations (tikv#9998)

903d010

ref tikv#9764 Signed-off-by: lhy1024 <admin@liudos.us>

affinity: add policy, metrics collection and tests (tikv#9999)

4349bda

ref tikv#9764 Signed-off-by: lhy1024 <admin@liudos.us> Co-authored-by: 混沌DM <hundundm@gmail.com> Signed-off-by: lhy1024 <admin@liudos.us>

schedule: add affinity filter and kvproto update and use affinity man…

79a8237

…anager (tikv#10038) ref tikv#9764 Signed-off-by: lhy1024 <admin@liudos.us>

affinity: refactor and address comments (tikv#10050)

535ef70

close tikv#9764 Signed-off-by: lhy1024 <admin@liudos.us>

api: add affinity groups api and client support (tikv#10041)

5830565

ref tikv#9764 Signed-off-by: lhy1024 <admin@liudos.us>

lhy1024 force-pushed the release-8.5-affinity branch from b3248da to ef62a65 Compare December 16, 2025 12:52

lhy1024 marked this pull request as ready for review December 16, 2025 12:52

ti-chi-bot Bot removed the do-not-merge/work-in-progress Indicates that a PR should not merge because it is a work in progress. label Dec 16, 2025

lhy1024 added 2 commits December 16, 2025 21:00

schedule: add affinity checker (tikv#10040)

b9da740

ref tikv#9764 Signed-off-by: lhy1024 <admin@liudos.us>

pd-ctl: support affinity group commands (tikv#10043)

cecd033

ref tikv#9764 Signed-off-by: lhy1024 <admin@liudos.us>

lhy1024 force-pushed the release-8.5-affinity branch 3 times, most recently from 0467ae4 to cecd033 Compare December 16, 2025 14:33

HunDunDM reviewed Dec 17, 2025

View reviewed changes

bufferflies reviewed Dec 19, 2025

View reviewed changes

bufferflies approved these changes Dec 19, 2025

View reviewed changes

ti-chi-bot Bot added the needs-1-more-lgtm Indicates a PR needs 1 more LGTM. label Dec 19, 2025

okJiang reviewed Dec 19, 2025

View reviewed changes

affinity, operator: fix affinity operator count (tikv#10091)

cad0ed5

ref tikv#9764 Signed-off-by: lhy1024 <admin@liudos.us>

okJiang reviewed Dec 19, 2025

View reviewed changes

lhy1024 requested review from HunDunDM and okJiang December 19, 2025 05:40

rleungx reviewed Dec 19, 2025

View reviewed changes

lhy1024 mentioned this pull request Dec 19, 2025

Support data affinity schedule for performance optimization #9764

Open

28 tasks

rleungx approved these changes Dec 19, 2025

View reviewed changes

ti-chi-bot Bot added lgtm and removed needs-1-more-lgtm Indicates a PR needs 1 more LGTM. labels Dec 19, 2025

HunDunDM approved these changes Dec 19, 2025

View reviewed changes

okJiang approved these changes Dec 19, 2025

View reviewed changes

niubell approved these changes Dec 19, 2025

View reviewed changes

ti-chi-bot Bot added approved cherry-pick-approved Cherry pick PR approved by release team. and removed do-not-merge/cherry-pick-not-approved labels Dec 19, 2025

ti-chi-bot Bot merged commit 029eb6e into tikv:release-8.5 Dec 19, 2025
28 of 34 checks passed

ti-chi-bot Bot added do-not-merge/cherry-pick-not-approved cherry-pick-approved Cherry pick PR approved by release team. and removed cherry-pick-approved Cherry pick PR approved by release team. do-not-merge/cherry-pick-not-approved labels Mar 5, 2026

lidezhu mentioned this pull request Mar 8, 2026

schemastore: fix panic for rename tables job from TiDB <= v8.1 pingcap/ticdc#4391

Merged

Conversation

lhy1024 commented Dec 14, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Conflict Resolution Details

Commit1: PR #9993 - Affinity Storage & Config

Commit2-4: PR #9997, #9998, #9999

Commit5: PR #10038 - Affinity Filter & kvproto

Commit6: PR #10050 - Refactor

Commit7: PR #10041 - API & Client Support

Commit8: PR #10040 - Affinity Checker

Commit9: PR #10043 - pd-ctl Support

Commit9: PR #10081 - merge fix

Commit10: PR #10080 - other fix

What problem does this PR solve?

What is changed and how does it work?

Check List

Release note

Uh oh!

ti-chi-bot Bot commented Dec 14, 2025

Uh oh!

ti-chi-bot Bot commented Dec 14, 2025

Uh oh!

codecov Bot commented Dec 14, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ti-chi-bot Bot commented Dec 19, 2025

[LGTM Timeline notifier]

Uh oh!

lhy1024 commented Dec 19, 2025

lhy1024 commented Dec 14, 2025 •

edited

Loading

codecov Bot commented Dec 14, 2025 •

edited

Loading