Tweak reconstruction batch size #6668
Merged
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
michaelsproul
added
ready-for-review
michaelsproul
added a commit
that referenced
this pull request
Dec 9, 2024
Squashed commit of the following: commit e5f2be5 Author: Michael Sproul <michael@sigmaprime.io> Date: Mon Dec 9 11:00:55 2024 +1100 Tweak reconstruction batch size
Merged
jimmygchen
approved these changes
Dec 10, 2024
jimmygchen
added
ready-for-merge
There's no metric, although it might be good to add one. There is a log:
I didn't notice if it did, by the time I went to check it had already run out of space and been re-synced |
|
@mergify queue |
✅ The pull request has been merged automaticallyThe pull request has been merged automatically at b2590bc |
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Outdated suggestions cannot be applied.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.
Issue Addressed
Address a performance issue reported by @mcdee which boils down to reconstruction taking too long and starving the finalization migration.
Proposed Changes
Jim's node with per-slot diffs spent 110 epochs doing reconstruction before picking up another finalization migration. This is very slow, and causes severe issues for the caches (similar to 110 epochs of non-finality).
To address this I've reduced the maximum number of blocks processed in a single reconstruction batch by a factor of 32, down to 1024. This should mean that a node like Jim's spends around 110/32 = 3.5 epochs doing reconstruction before it does another finalization migration. This is much more manageable for the hot state cache (4 epochs * 32 slots = 128 states, which fits inside the cache with default sizing).