ddl: clone table meta for recover table

[joechenrh]

What problem does this PR solve?

Issue Number: close #55442

Problem Summary:

As described in #55442, the infoschema checker has found inconsistent with v1 and v2 in TestRecoverTableByJobID, thus throw a panic.

This is because the tableinfo used by recovery is shallow copied from infocache and modified by subsequent process.

tidb/pkg/ddl/executor.go

Lines 1429 to 1444 in c74a233

	tbInfo.State = model.StateNone
	job := &model.Job{
	SchemaID: schemaID,
	TableID: tbInfo.ID,
	SchemaName: schema.Name.L,
	TableName: tbInfo.Name.L,
	Type: model.ActionRecoverTable,
	BinlogInfo: &model.HistoryInfo{},
	Args: []any{recoverInfo, recoverCheckFlagNone},
	InvolvingSchemaInfo: involvedSchemas,
	CDCWriteSource: ctx.GetSessionVars().CDCWriteSource,
	SQLMode: ctx.GetSessionVars().SQLMode,
	}
	err = e.DoDDLJob(ctx, job)
	return errors.Trace(err)

What changed and how does it work?

Make a clone for table meta. This can fix TestRecoverTableByJobID failure.

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

[tiprow]

Hi @joechenrh. Thanks for your PR.

PRs from untrusted users cannot be marked as trusted with /ok-to-test in this repo meaning untrusted PR authors can never trigger tests themselves. Collaborators can still trigger tests on the PR using /test all.

I understand the commands that are listed here.

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]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 75.2201%. Comparing base (c74a233) to head (309014c).
Report is 12 commits behind head on master.

Additional details and impacted files

@@               Coverage Diff                @@
##             master     #55443        +/-   ##
================================================
+ Coverage   74.9176%   75.2201%   +0.3025%     
================================================
  Files          1576       1576                
  Lines        365707     443461     +77754     
================================================
+ Hits         273979     333572     +59593     
- Misses        71965      89769     +17804     
- Partials      19763      20120       +357     

Flag	Coverage Δ
integration	49.6556% <0.0000%> (?)
unit	72.3540% <100.0000%> (-1.5541%)	⬇️

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

Components	Coverage Δ
dumpling	52.9567% <ø> (-2.2327%)	⬇️
parser	∅ <ø> (∅)
br	52.5429% <ø> (+4.7246%)	⬆️

[ti-chi-bot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: tangenta, tiancaiamao

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

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [tangenta,tiancaiamao]

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

[ti-chi-bot]

[LGTM Timeline notifier]

Timeline:

• 2024-08-16 01:13:58.486782426 +0000 UTC m=+489723.190252069: ☑️ agreed by tiancaiamao.
• 2024-08-16 01:47:46.198352098 +0000 UTC m=+491750.901821742: ☑️ agreed by tangenta.

[D3Hunter]

why it causes "[schema:1050]Table 't_recover' already exists" if not clone?

[joechenrh]

why it causes "[schema:1050]Table 't_recover' already exists" if not clone?

Because recover table will check if it exists table with same name.

tidb/pkg/ddl/executor.go

Lines 1455 to 1458 in 2045418

	// Check not exist table with same name.
	if ok := is.TableExists(schema.Name, tbInfo.Name); ok {
	return infoschema.ErrTableExists.GenWithStackByArgs(tbInfo.Name)
	}

I have to mention that, on master branch, this test can always pass no matter whether this patch is applied or not. The key problem is that meta data retrieved from infoschema cache will be modified by recover table process, but it seems like recover table doesn't check table state. So this problem is not exposed without infoschema checker.

Besides, do you think we should check table state before executing RecoverTable?

[hawkingrei]

/retest