fix(compaction): wire aggregate retry timeout through compaction.timeoutSeconds#94490
Closed
LZY3538 wants to merge 1 commit into
Closed
fix(compaction): wire aggregate retry timeout through compaction.timeoutSeconds#94490LZY3538 wants to merge 1 commit into
LZY3538 wants to merge 1 commit into
Conversation
openclaw-barnacle
Bot
added
agents
openclaw-barnacle
Bot
added
triage: mock-only-proof
…outSeconds The hardcoded COMPACTION_RETRY_AGGREGATE_TIMEOUT_MS (60s) fires before compaction model calls complete on large sessions (~200K tokens, ~143-187s), discarding valid provider results and triggering self-compounding retry storms. Replace the hardcoded 60s constant with the already-resolved compactionTimeoutMs from agents.defaults.compaction.timeoutSeconds (shipped default 180s), which already governs the inner model-call safety wrapper. Operators who have raised timeoutSeconds get the fix for free. Fixes openclaw#94391 Co-Authored-By: Claude <[email protected]>
LZY3538
force-pushed
the
fix/compaction-aggregate-timeout
branch
from
June 18, 2026 08:39
3e07186 to
ec18ffa
Compare
Member
|
Thanks @LZY3538 for the focused fix here. I am closing this as superseded by #94413 because both PRs target the same #94391 root cause: wiring the aggregate compaction retry wait away from the hardcoded 60s path and into the compaction timeout behavior. Canonical path: #94413 I am keeping the canonical thread open there so review, validation, and follow-up stay in one place. Clownfish will preserve attribution/credit from this source PR in the cleanup record, and if this branch has a different reproduction path or still matters after the canonical fix path lands, please reply and we can reopen or split it back out. |
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.
Summary
Fixes #94391: the hardcoded
COMPACTION_RETRY_AGGREGATE_TIMEOUT_MS(60s) fires before compaction model calls complete on large sessions (~200K tokens, ~143-187s observed), discarding valid provider results and triggering self-compounding retry storms.Changes
In
src/agents/embedded-agent-runner/run/attempt.ts:COMPACTION_RETRY_AGGREGATE_TIMEOUT_MS = 60_000local constantcompactionTimeoutMsfromagents.defaults.compaction.timeoutSeconds(shipped default 180s)Real behavior proof
compactionTimeoutMsvariable (default 180s) is accessible in scope at the aggregate wait site. With the shipped default of 180s, all three observed failure cases (143s, 174s, 187s) would have completed successfully within the timeout. No remaining references to the removedCOMPACTION_RETRY_AGGREGATE_TIMEOUT_MSconstant.Related
🤖 Generated with Claude Code