Repository navigation
test(two-node-sim): model F25 (takeover waits for the append lock on a frozen peer's JM) - #1507
Merged
Merged
Conversation
…ck on a frozen peer's JM Bed c12/02c (image with ultra #219 incl. F24): no survivor abort any more, but the new leader's first write waited for its append-lock request on the frozen peer's JM until the peer aborted on wake (6-8 s client stall, 0 IO errors). - simnode: CAP_APPENDER_LOCK_SKIPS_PEER_JM (ultra #219, F25 commit) - actors: with the F23+F24 capabilities but not F25's, a solo takeover of a frozen peer's LVS lasts until the peer's frozen_until, violation "F25" - scenarios: F25-frozen-primary-takeover-append-lock, judged on the client stall (the F24 scenario has no stall limit, which is why it missed this); F25's capability added to every pending-fix set that includes F24's - test_f25_needs_its_own_fix Co-Authored-By: Claude Opus 5.5 <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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Bed run c12 (2026-10-19, node image = ultra main + #217 + #219 incl. F24, CP main-ceff9f53): 02c (the vol1 primary sn2 frozen 8 s) no longer aborts the survivor (F24 fixed), but stalls 6/7/8 s with 0 IO errors.
Mechanism (proven on the bed logs): at the takeover the new leader re-locks the appender on every JM and
helper_ensure_append_locks_readywaits for all lock requests. The one on the frozen peer's JM (a READ onremote_jm_…) neither completes nor fails until the peer aborts on wake or the 8 s NVMe IO timeout fires ("locke failed _res_lock -3"), and the lvstore update's used-blobids write, with all client IO, waits until then.This PR models it:
CAP_APPENDER_LOCK_SKIPS_PEER_JM(the fix: ultra Adding lvol id to lvol events #219, commit b4d92c71)frozen_until→ violationF25F25-frozen-primary-takeover-append-lock, judged on the client stall. The F24 scenario hadstall_limit_s=None, which is why the simulator did not catch this before the bed did.test_f25_needs_its_own_fixLocal run against FDB: 64 passed, 2 skipped, 8 xfailed.
🤖 Generated with Claude Code