jobs: improve the status reporting of index backfill jobs #39773
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
|
This change is |
Before this change index backfill workers would write their updated spans concurrently to the job's row which can potentially lead to blow up in the jobs table. This PR is a first step in improving this. Instead of updating the finished spans in each worker, we propagate this info to the gateway node and do bulk updates there. Next step is to let the workers run longer by not canceling them at 2 mins which is currently done for historical reasons. Fixes cockroachdb#36601. Release note: None
spaskob
force-pushed
the
index-backfill-progress
branch
from
8f4be12
to
9bce82e
Compare
dt
approved these changes
Aug 23, 2019
dt
reviewed
Aug 23, 2019
| @@ -843,10 +841,16 @@ func (sc *SchemaChanger) distBackfill( | |||
| otherTableDescs = append(otherTableDescs, *table.TableDesc()) | |||
| } | |||
| } | |||
| rw := &errOnlyResultWriter{} | |||
| metaFn := func(_ context.Context, meta *distsqlpb.ProducerMetadata) { | |||
| if meta.BulkProcessorProgress != nil { | |||
There was a problem hiding this comment.
I think we avoid wrapping todoSpans access in a mutex because all the access from the callback is from the same goroutine (the distsql inbound goroutine) and all happens while this goroutine is blocked on Run.
If that is correct, that's a slightly brittle assumption but it works for me for now.
|
bors r+ |
craig bot
pushed a commit
that referenced
this pull request
Aug 23, 2019
39773: jobs: improve the status reporting of index backfill jobs r=spaskob a=spaskob Before this change index backfill workers would write their updated spans concurrently to the job's row which can potentially lead to blow up in the jobs table. This PR is a first step in improving this. Instead of updating the finished spans in each worker, we propagate this info to the gateway node and do bulk updates there. Next step is to let the workers run longer by not canceling them at 2 mins which is currently done for historical reasons. Fixes #36601. Release note: None Co-authored-by: Spas Bojanov <spas@cockroachlabs.com>
Build succeeded |
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.
Before this change index backfill workers would write their updated
spans concurrently to the job's row which can potentially lead to blow
up in the jobs table. This PR is a first step in improving this. Instead
of updating the finished spans in each worker, we propagate this info to
the gateway node and do bulk updates there. Next step is to let the
workers run longer by not canceling them at 2 mins which is currently
done for historical reasons.
Fixes #36601.
Release note: None