VReplication: improve reliability of log management

[mattlord]

Description

In helping someone debug a VReplication failure, we had seen the following:

1. We failed to create a _vt.vreplication_log record and the apparent cause was that the message was too long
2. The truncated error message seen in the _vt.vreplication record showed the beginning of a vcopier bulk INSERT query that failed — truncated at 1000 bytes — but not the reason why / error it generated
3. This _vt.vreplication_log record generation failure may have caused problems in the workflow (e.g. if this happens after the point of no return in a SwitchTraffic then we'll be left in an inconsistent state that needs manual cleanup)

We address all of these issues with the following changes in this PR:

1. We truncate the message as needed before inserting it into the _vt.vreplication_log table
2. That truncation happens in the middle of the message so that we don't lose the error shown e.g. after a failed query
3. We no longer consider failing to write a _vt.vreplication_log record as "fatal" — meaning that the workflow operation does not stop and return the log generation error — and instead only logs an error in the vttablet log

I also moved the existing truncation for the _vt.vreplication.message field over to this new truncation method so that we have the error message there as well.

Related Issue(s)

I didn't create an issue since this felt more like a minor internal cleanup, but I can create one quickly if others disagree.

Checklist

• "Backport to:" labels have been added if this change should be back-ported to release branches
• If this change is to be back-ported to previous releases, a justification is included in the PR description
• Tests were added or are not required
• Did the new or modified tests pass consistently locally and on CI?
• Documentation was added or is not required

[vitess-bot]

Review Checklist

Hello reviewers! 👋 Please follow this checklist when reviewing this Pull Request.

General

• Ensure that the Pull Request has a descriptive title.
• Ensure there is a link to an issue (except for internal cleanup and flaky test fixes), new features should have an RFC that documents use cases and test cases.

Tests

• Bug fixes should have at least one unit or end-to-end test, enhancement and new features should have a sufficient number of tests.

Documentation

• Apply the release notes (needs details) label if users need to know about this change.
• New features should be documented.
• There should be some code comments as to why things are implemented the way they are.
• There should be a comment at the top of each new or modified test to explain what the test does.

New flags

• Is this flag really necessary?
• Flag names must be clear and intuitive, use dashes (-), and have a clear help text.

If a workflow is added or modified:

• Each item in Jobs should be named in order to mark it as required.
• If the workflow needs to be marked as required, the maintainer team must be notified.

Backward compatibility

• Protobuf changes should be wire-compatible.
• Changes to _vt tables and RPCs need to be backward compatible.
• RPC changes should be compatible with vitess-operator
• If a flag is removed, then it should also be removed from vitess-operator and arewefastyet, if used there.
• vtctl command output order should be stable and awk-able.

[codecov]

Codecov Report

Attention: Patch coverage is 73.21429% with 15 lines in your changes are missing coverage. Please review.

Project coverage is 65.45%. Comparing base (fb4abd5) to head (9a0e780).
Report is 5 commits behind head on main.

Files	Patch %	Lines
go/vt/vttablet/tabletmanager/vreplication/utils.go	58.33%	5 Missing ⚠️
go/vt/binlog/binlogplayer/mock_dbclient.go	50.00%	4 Missing ⚠️
.../vt/vttablet/tabletmanager/vreplication/vcopier.go	25.00%	3 Missing ⚠️
go/vt/binlog/binlogplayer/dbclient.go	60.00%	2 Missing ⚠️
go/vt/vtctl/workflow/server.go	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #15374      +/-   ##
==========================================
+ Coverage   65.43%   65.45%   +0.02%     
==========================================
  Files        1561     1562       +1     
  Lines      193638   193877     +239     
==========================================
+ Hits       126704   126902     +198     
- Misses      66934    66975      +41     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[rohit-nayak-ps]

Looks good!

Had a deja vu about this: I thought we had added this truncation a long time ago, because we had seen this error before for long queries.

Turns out we had done it for _vt.vreplication while setting the ubiquitous message column:

func (ct *controller) setMessage(dbClient binlogplayer.DBClient, message string) error {
	ct.blpStats.History.Add(&binlogplayer.StatsHistoryRecord{
		Time:    time.Now(),
		Message: message,
	})
	query := fmt.Sprintf("update _vt.vreplication set message=%v where id=%v", encodeString(binlogplayer.MessageTruncate(message)), ct.id)
	if _, err := dbClient.ExecuteFetch(query, 1); err != nil {
		return fmt.Errorf("could not set message: %v: %v", query, err)
	}
	return nil
}

Maybe we can just update the MessageTruncate function to use your idea of logging enough of the prefix and suffix and use it in this PR as well.

[rohit-nayak-ps]

Nice work.