redo(ticdc): fix resolved moves too fast when part of tables are not maintained redo writer #5587
Conversation
amyangfei
added
type/bugfix
|
[REVIEW NOTIFICATION] This pull request has been approved by:
To complete the pull request process, please ask the reviewers in the list to review by filling The full list of commands accepted by this bot can be found here. Reviewer can indicate their review by submitting an approval review. |
ti-chi-bot
added
do-not-merge/needs-triage-completed
release-note
|
/run-check-issue-triage-complete |
…tained redo writer
amyangfei
force-pushed
the
fix-redo-log-advance-too-fast
branch
from
463fbf8
to
a6ecc1f
Compare
|
/run-all-tests |
|
/run-verify |
Codecov Report
Flags with carried forward coverage won't be shown. Click here to find out more. @@ Coverage Diff @@
## master #5587 +/- ##
================================================
+ Coverage 56.0768% 56.2697% +0.1928%
================================================
Files 535 528 -7
Lines 70143 70027 -116
================================================
+ Hits 39334 39404 +70
+ Misses 27078 26898 -180
+ Partials 3731 3725 -6 |
| for tableID, rts := range rtsMap { | ||
| m.rtsMap[tableID] = rts | ||
| for tableID := range m.rtsMap { | ||
| if rts, ok := rtsMap[tableID]; ok { |
There was a problem hiding this comment.
why part of tables are not maintained in redo log writer?
There was a problem hiding this comment.
If there is no data of this table, and flush resolved ts of this table has not been executed, the redo log writer doesn't know this table.
The redo log writer only records the information of the table that it has met.
|
@nongfushanquan: Thanks for your review. The bot only counts approvals from reviewers and higher roles in list, but you're still welcome to leave your comments. In response to this: 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 ti-community-infra/tichi repository. |
ti-chi-bot
added
status/LGT2
|
/merge |
|
This pull request has been accepted and is ready to merge. Commit hash: a6ecc1f
|
ti-chi-bot
added
the
status/can-merge
Rustin170506
changed the title
|
/merge |
|
/run-all-tests |
|
/run-integration-test |
|
/run-verify |
|
/run-verify |
2 similar comments
|
/run-verify |
|
/run-verify |
|
@amyangfei: Your PR was out of date, I have automatically updated it for you. At the same time I will also trigger all tests for you: /run-all-tests If the CI test fails, you just re-trigger the test that failed and the bot will merge the PR for you after the CI passes. 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 ti-community-infra/tichi repository. |
|
In response to a cherrypick label: new pull request created: #5617. |
|
In response to a cherrypick label: new pull request created: #5618. |
|
In response to a cherrypick label: new pull request created: #5619. |
What problem does this PR solve?
Issue Number: close #5486
What is changed and how it works?
redo manager queries redo log writer to update table resolved ts, but when part of tables are not maintained in redo log writer, the redo manager doesn't respect the resolved ts of these tables and could move resolved ts forward too fast, which leads to data loss.
Check List
Tests
Questions
Will it cause performance regression or break compatibility?
Do you need to update user documentation, design documentation or monitoring documentation?
Release note