DM may stuck at syncer initialization

[lance6716]

What did you do?

https://asktug.com/t/topic/903490/8?u=lance6716

goroutine 358 [select]:
github.com/pingcap/tidb/ddl.(*ddl).DoDDLJob(0xc0111c3780, {0x3974878?, 0xc011ec1200?}, 0xc0007d6d80)
	github.com/pingcap/tidb@v1.1.0-beta.0.20220528045048-5495dc6c4360/ddl/ddl.go:699 +0x752
github.com/pingcap/tidb/ddl.(*ddl).BatchCreateTableWithInfo(0x2bdda00?, {0x3974878, 0xc011ec1200}, {{0xc00057f980?, 0x1008ba698?}, {0xc00057f980?, 0x4?}}, {0xc012181800, 0x95, 0x100}, ...)
	github.com/pingcap/tidb@v1.1.0-beta.0.20220528045048-5495dc6c4360/ddl/ddl_api.go:2451 +0x706
github.com/pingcap/tiflow/dm/pkg/schema.(*Tracker).BatchCreateTableIfNotExist(0xc000f1c7e0, 0x3253c7d?)
	github.com/pingcap/tiflow/dm/pkg/schema/tracker.go:458 +0x21e
github.com/pingcap/tiflow/dm/syncer.(*RemoteCheckPoint).LoadIntoSchemaTracker(0xc0010da4e0, {0xc000dffe40?, 0xc00127f538?}, 0x5?)
	github.com/pingcap/tiflow/dm/syncer/checkpoint.go:1104 +0x1de
github.com/pingcap/tiflow/dm/syncer.(*Syncer).Run(0xc000475680, {0x391a0b8?, 0xc000dffe40?})
	github.com/pingcap/tiflow/dm/syncer/syncer.go:1703 +0xeb1
github.com/pingcap/tiflow/dm/syncer.(*Syncer).Process(0xc000475680, {0x391a0b8, 0xc000dffd80}, 0xc000175320?)
	github.com/pingcap/tiflow/dm/syncer/syncer.go:702 +0x392
created by github.com/pingcap/tiflow/dm/dm/worker.(*SubTask).run
	github.com/pingcap/tiflow/dm/dm/worker/subtask.go:260 +0x56c

What did you expect to see?

the internal call of BatchCreateTableWithInfo should not stuck

What did you see instead?

see above

Versions of the cluster

v6.1.0

current status of DM cluster (execute query-status <task-name> in dmctl)

(paste current status of DM cluster here)

[lance6716]

this should be solved by #5334 in v6.3.0

[lance6716]

the user says it has 800 tables in one task, and the goroutine stack shows it is waiting at

https://github.com/pingcap/tidb/blob/b99aebee3110e539bec53bbd8838fa554ad4e653/ddl/ddl.go#L490-L495

800*0.1s = 80s

will check if we can step aside this check later