planner: add HasTiflash to avoid generating unnecessary mpp task | tidb-test=pr/2660

[hawkingrei]

What problem does this PR solve?

Issue Number: close #65335

Problem Summary:

What changed and how does it work?

checkCanPushDownToMPP will be called twice in the following code. The other problem is that if tables without tiflash replications are used in the query, we don't need to check the mpp task.

tidb/pkg/planner/core/operator/physicalop/physical_hash_agg.go

Lines 52 to 108 in 79ec3b6

	func getHashAggs(lp base.LogicalPlan, prop *property.PhysicalProperty) []base.PhysicalPlan {
	la := lp.(*logicalop.LogicalAggregation)
	if !prop.IsSortItemEmpty() {
	return nil
	}
	if prop.TaskTp == property.MppTaskType && !checkCanPushDownToMPP(la) {
	return nil
	}
	hashAggs := make([]base.PhysicalPlan, 0, len(prop.GetAllPossibleChildTaskTypes()))
	taskTypes := []property.TaskType{property.CopSingleReadTaskType, property.CopMultiReadTaskType, property.RootTaskType}
	// aggregation has a special case that it can be pushed down to TiKV which is indicated by the prop.NoCopPushDown
	if prop.NoCopPushDown {
	taskTypes = []property.TaskType{property.RootTaskType}
	}
	// lift the recursive check of canPushToCop(tiFlash)
	canPushDownToMPP := la.SCtx().GetSessionVars().IsMPPAllowed() && checkCanPushDownToMPP(la)
	if canPushDownToMPP {
	taskTypes = append(taskTypes, property.MppTaskType)
	} else {
	hasMppHints := false
	var errMsg string
	if la.PreferAggType&h.PreferMPP1PhaseAgg > 0 {
	errMsg = "The agg can not push down to the MPP side, the MPP_1PHASE_AGG() hint is invalid"
	hasMppHints = true
	}
	if la.PreferAggType&h.PreferMPP2PhaseAgg > 0 {
	errMsg = "The agg can not push down to the MPP side, the MPP_2PHASE_AGG() hint is invalid"
	hasMppHints = true
	}
	if hasMppHints {
	la.SCtx().GetSessionVars().StmtCtx.SetHintWarning(errMsg)
	}
	}
	if prop.IsFlashProp() {
	taskTypes = []property.TaskType{prop.TaskTp}
	}
	taskTypes = admitIndexJoinTypes(taskTypes, prop)
	for _, taskTp := range taskTypes {
	if taskTp == property.MppTaskType {
	mppAggs := tryToGetMppHashAggs(la, prop)
	if len(mppAggs) > 0 {
	hashAggs = append(hashAggs, mppAggs...)
	}
	} else {
	childProp := &property.PhysicalProperty{ExpectedCnt: math.MaxFloat64, TaskTp: taskTp, CTEProducerStatus: prop.CTEProducerStatus, NoCopPushDown: prop.NoCopPushDown}
	// mainly to fill indexJoinProp to childProp.
	childProp = admitIndexJoinProp(childProp, prop)
	if childProp == nil {
	continue
	}
	agg := NewPhysicalHashAgg(la, la.StatsInfo().ScaleByExpectCnt(la.SCtx().GetSessionVars(), prop.ExpectedCnt), childProp)
	agg.SetSchema(la.Schema().Clone())
	hashAggs = append(hashAggs, agg)
	}
	}
	return hashAggs
	}

Check List

Tests

• Unit test
• Integration test
• Manual test (add detailed scripts or steps below)
• No need to test

  • I checked and no code files have been changed.

Side effects

• Performance regression: Consumes more CPU
• Performance regression: Consumes more Memory
• Breaking backward compatibility

Documentation

• Affects user behaviors
• Contains syntax changes
• Contains variable changes
• Contains experimental features
• Changes MySQL compatibility

Release note

Please refer to Release Notes Language Style Guide to write a quality release note.

None

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 78.8974%. Comparing base (e8a60bd) to head (74c7d31).
⚠️ Report is 1 commits behind head on master.

Additional details and impacted files

@@               Coverage Diff                @@
##             master     #65250        +/-   ##
================================================
+ Coverage   77.7750%   78.8974%   +1.1223%     
================================================
  Files          2001       1922        -79     
  Lines        545527     532954     -12573     
================================================
- Hits         424284     420487      -3797     
+ Misses       119581     111014      -8567     
+ Partials       1662       1453       -209     

Flag	Coverage Δ
integration	48.9488% <ø> (+0.7840%)	⬆️

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

Components	Coverage Δ
dumpling	56.7974% <ø> (ø)
parser	∅ <ø> (∅)
br	66.0791% <ø> (+5.0887%)	⬆️

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[hawkingrei]

/retest

[hawkingrei]

/retest

[hawkingrei]

/retest

[hawkingrei]

/retest

[hawkingrei]

/retest

[hawkingrei]

/retest

[winoros]

I think we don't need to add this together in this pr.
Current physical optimization needs to generate the mpp task type first because we don't know whether its children can be converted to mpp plan.
If we want to do the pruning completely, we'll need to change the physical opt framework itself.

[hawkingrei]

But under the current circumstances, most users actually won't have TiFlash replicas. So when building the data source, it can quickly determine whether it contains TiFlash based on the table metadata.

[hawkingrei]

/retest

[Copilot]

Pull request overview

Copilot reviewed 99 out of 99 changed files in this pull request and generated 4 comments.

[Copilot]

TiKVSingleGather.PreparePossibleProperties drops HasTiflash from its child (it only forwards Order). This will make upstream GetHasTiFlash() calculations incorrectly become false after a gather, potentially preventing MPP/TiFlash plan generation. Consider propagating childrenProperties[0].HasTiflash as well.

[Copilot]

Typo in comment: "wether" should be "whether".

[Copilot]

Doc comment has a duplicated word ("SetHasTiFlashSetHasTiFlash"). Consider renaming it to "SetHasTiFlash" for clarity.

[Copilot]

PreparePossibleProperties can panic when any childInfo is nil (childInfo.HasTiflash dereference). Also, returning nil here causes preparePossibleProperties() to propagate nil upwards, which can lead to nil dereferences in callers that access .Order/.HasTiflash. Consider treating nil child infos as HasTiflash=false and returning a non-nil PossiblePropertiesInfo (even when no order properties), so HasTiFlash can propagate safely through operators that rely on the base implementation.

[tiprow]

@hawkingrei: The following test failed, say /retest to rerun all failed tests or /retest-required to rerun all mandatory failed tests:

Test name	Commit	Details	Required	Rerun command
fast_test_tiprow	74c7d31	link	true	/test fast_test_tiprow

Full PR test history. Your PR dashboard.

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. I understand the commands that are listed here.

[ti-chi-bot]

@hawkingrei: The following tests failed, say /retest to rerun all failed tests or /retest-required to rerun all mandatory failed tests:

Test name	Commit	Details	Required	Rerun command
idc-jenkins-ci-tidb/mysql-test	74c7d31	link	true	/test mysql-test
pull-mysql-client-test-next-gen	74c7d31	link	true	/test pull-mysql-client-test-next-gen
idc-jenkins-ci-tidb/unit-test	74c7d31	link	true	/test unit-test
idc-jenkins-ci-tidb/check_dev	74c7d31	link	true	/test check-dev
pull-unit-test-next-gen	74c7d31	link	true	/test pull-unit-test-next-gen
idc-jenkins-ci-tidb/check_dev_2	74c7d31	link	true	/test check-dev2

Full PR test history. Your PR dashboard.

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. I understand the commands that are listed here.