Repository navigation
refactor(budget): share pacing state - #8464
Draft
mediocregopher wants to merge 2 commits into
Draft
mediocregopher wants to merge 2 commits into
mediocregopher wants to merge 2 commits into
Conversation
mediocregopher
added
codex
A-block-building
Contributor
👁️ Cyclops Security ReviewLast reviewed: no commit yet |
This was referenced Oct 20, 2026
This branch has not been deployed
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Shares build and validation feedback between consensus and the payload builder. Keeps the existing multiplier algorithm and fixed consensus budget, and passes the same handle through startup, overrides, and the e2e harness.
Part 2/7 of the stack splitting #7838. Added unit tests from that PR are omitted; existing tests are preserved.
Stack: #8463 → #8464 → #8465 → #8466 → #8467 → #8468 → #8469. Merge in order using merge commits to preserve the stack ancestry.
Four busy-load pairs measured mean TPS 15,249 → 15,021 (−1.5%), with individual changes −7.0%, −0.6%, −2.1%, and +4.2%. The last pair reversed the revision order; the shared-state revision was faster there. Results vary with skipped views and long-tail stalls, and do not show a consistent throughput regression. First run, repeat, reversed order.
At 100 TPS offered load, measured chain TPS was 88 → 90, with pooled block-time p50/p75 637/667 ms → 646/691 ms. Low-load comparison.