sink(cdc): avoid sinking redundant events in some rare cases with redo enabled

[hicqu]

What problem does this PR solve?

Issue Number: close #10065 .

What is changed and how it works?

Let's focus on file cdc/processor/sinkmanager/table_sink_worker.go. #10065 panics if

• redo is enabled;
• sink memory quota is almost full;
• there are transactions larger than 8M;
• when handling one large transaction, fetchFromCache returns drained=false;
• then the condition for advancer.hasEnoughMem() && !task.isCanceled() is broken immediately;
• changefeed is paused and then resumed, table sink is faster than redo.

After that, the variable lastPos can be less than lowerBound.Prev(), which is not expected.

Check List

Tests

• Unit test

Questions

Will it cause performance regression or break compatibility?

Do you need to update user documentation, design documentation or monitoring documentation?

Release note

None

[ti-chi-bot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: CharlesCheung96, sdojjy

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 [CharlesCheung96,sdojjy]

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:

• 2023-11-14 09:46:37.293797288 +0000 UTC m=+4155994.880907417: ☑️ agreed by sdojjy.
• 2023-11-14 10:05:48.224694386 +0000 UTC m=+4157145.811804532: ☑️ agreed by CharlesCheung96.

[sdojjy]

/test verify