Skip to content
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

admission: increase file count based overload threshold #138699

Open
wants to merge 1 commit into
base: master
Choose a base branch
from

Conversation

sumeerbhola
Copy link
Collaborator

The current value of 1000 was too inconsistent with Pebble's compaction scoring, in that compaction scoring had 1 sub-level == 500 L0 files, and admission control had 1 sub-level == 50 L0 files. The value is increased to 4000, so 1 sub-level == 200 L0 files. There is a long code comment elaborating on this.

Informs #138655

Epic: none

Release note (ops change): The default value of cluster setting admission.l0_file_count_overload_threshold is changed to 4000.

The current value of 1000 was too inconsistent with Pebble's compaction
scoring, in that compaction scoring had 1 sub-level == 500 L0 files, and
admission control had 1 sub-level == 50 L0 files. The value is increased
to 4000, so 1 sub-level == 200 L0 files. There is a long code comment
elaborating on this.

Informs cockroachdb#138655

Epic: none

Release note (ops change): The default value of cluster setting
admission.l0_file_count_overload_threshold is changed to 4000.
@sumeerbhola sumeerbhola requested a review from a team as a code owner January 8, 2025 23:42
Copy link

blathers-crl bot commented Jan 8, 2025

It looks like your PR touches production code but doesn't add or edit any test code. Did you consider adding tests to your PR?

🦉 Hoot! I am a Blathers, a bot for CockroachDB. My owner is dev-inf.

@cockroach-teamcity
Copy link
Member

This change is Reviewable

@andrewbaptist
Copy link
Collaborator

It might be better to drop the value for the pebble compaction scoring. Many of the current tests only go a little over 1000 (in the 2000-3000 range) and would need to be updated to run longer to hit the higher target. That isn't necessarily a problem but it means that we might miss things that would happen on a real system.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants