perf(v2): optimize compaction planning #4149
Closed
Conversation
This file contains hidden or 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
kolesnikovae
commented
Apr 28, 2025
Comment on lines
-388
to
-391
| // Assuming that block IDs (16b ULID) are globally unique. | ||
| // We could achieve the same with more efficiency by marking visited | ||
| // batches. However, marking visited blocks seems to be more robust, | ||
| // and the size of the map is expected to be small. |
Collaborator
Author
There was a problem hiding this comment.
It was sort of expected, but was pessimized prematurelly.
Collaborator
Author
|
After #4216 things look much better |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.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.
It's been observed that the compaction process might be very wasteful in terms of CPU resources. The reason is that the compaction planner inspects the block queues, cancelling the planned compaction job if the conditions are not met. The problem is that we do not skip the queues that we could have already checked.