fix: fail-fast on the archive-replay error hot-loop (P1.3, match#35)#55
Merged
Conversation
The 2026-07-02 and 2026-07-04 incidents each threw the byte-identical
ArchiveException ("replayPosition ... does not point to a valid frame")
thousands of times from Election.followerLogReplication while the node sat
wedged indefinitely — silent before the truthful status API, DEGRADED-forever
after it, but never terminal.
The Aeron error handler now tracks the consecutive-identical-error streak
(signature = exception class + message, which stayed byte-identical across
5,400+ throws in the incidents). At 200 consecutive identical errors it
signals the shutdown barrier and halts (exit 2) with a loud FAIL-FAST log
naming the signature and pointing at reseed/recovery — a crashed node the
process manager can count and the operator can see, instead of a spinner.
Any different error resets the streak; healthy operation never repeats one
exact error 200 times in a row.
Companion guard (refuse snapshot/housekeeping while a member is down or
lagging — the stranding mechanism itself) lands on the admin-gateway.
Part of match#35 (P1.3).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Part of #35 (P1.3-iv). Both replay-loop incidents (2026-07-02, 2026-07-04) threw the byte-identical
ArchiveExceptionthousands of times while the node sat wedged forever. The Aeron error handler now tracks the consecutive-identical-error streak (class + message signature); at 200 consecutive identical errors it signals the barrier and halts (exit 2) with a loudFAIL-FASTlog naming the signature and pointing at reseed/recovery. Different errors reset the streak — healthy operation never repeats one exact error 200 times in a row.Reuses the existing driver-death exit machinery (barrier + backstop halt +
FATAL_EXIT_STARTEDonce-guard). Companion admin-gateway PR adds the lag guard that prevents the stranding in the first place.🤖 Generated with Claude Code