-
Notifications
You must be signed in to change notification settings - Fork 3.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
release-23.1: changefeedccl: fix initial scan checkpointing #123970
Merged
wenyihu6
merged 1 commit into
cockroachdb:release-23.1
from
wenyihu6:backport23.1-123625
May 10, 2024
Merged
release-23.1: changefeedccl: fix initial scan checkpointing #123970
wenyihu6
merged 1 commit into
cockroachdb:release-23.1
from
wenyihu6:backport23.1-123625
May 10, 2024
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Initially, all span initial resolved timestamps are kept as zero upon resuming a job since initial resolved timestamps are set as initial high water which remains zero until initial scan is completed. However, since cockroachdb@0eda540, we began reloading checkpoint timestamps instead of setting them all to zero at the start. In PR cockroachdb#102717, we introduced a mechanism to reduce message duplicates by re-loading job progress upon resuming which largely increased the likelihood of this bug. These errors could lead to incorrect frontier and missing events during initial scans. This patches changes how we initialize initial high water and frontier by initializing it as zero if there are any zero initial high water in initial resolved timestamps. Fixes: cockroachdb#123371 Release note (enterprise change): Fixed a bug in v22.2+ where long running initial scans may incorrectly restore checkpoint job progress and drop events during node / changefeed restart. This issue was most likely to occur in clusters with: 1) changefeed.shutdown_checkpoint.enabled (v23.2) is set 2) Multiple table targets in a changefeed, or 3) Low changefeed.frontier_checkpoint_frequency or low changefeed.frontier_highwater_lag_checkpoint_threshold.
Thanks for opening a backport. Please check the backport criteria before merging:
If your backport adds new functionality, please ensure that the following additional criteria are satisfied:
Also, please add a brief release justification to the body of your PR to justify this |
rharding6373
approved these changes
May 10, 2024
stevendanna
approved these changes
May 10, 2024
blathers backport release-23.1.22-rc |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Backport 1/1 commits from #123625.
/cc @cockroachdb/release
Initially, all span initial resolved timestamps are kept as zero upon resuming a
job since initial resolved timestamps are set as initial high water which
remains zero until initial scan is completed. However, since
0eda540,
we began reloading checkpoint timestamps instead of setting them all to zero at
the start. In PR #102717, we introduced a mechanism to reduce message duplicates
by re-loading job progress upon resuming which largely increased the likelihood
of this bug. These errors could lead to incorrect frontier and missing events
during initial scans. This patches changes how we initialize initial high water
and frontier by initializing it as zero if there are any zero initial high water
in initial resolved timestamps.
Fixes: #123371
Release note (enterprise change): Fixed a bug in v22.2+ where long running
initial scans may incorrectly restore checkpoint job progress and drop events
during node / changefeed restart. This issue was most likely to occur in
clusters with: 1) changefeed.shutdown_checkpoint.enabled (v23.2) is set 2)
Multiple table targets in a changefeed, or 3) Low
changefeed.frontier_checkpoint_frequency or low
changefeed.frontier_highwater_lag_checkpoint_threshold.
Release justification: Fixed a severe bug in v22.2+ where long running
initial scans may incorrectly restore checkpoint job progress and drop events
during node / changefeed restart.