Repository navigation
Fix interrupted transaction recovery catchup - #450
Merged
Merged
Conversation
heifner
requested changes
Jun 27, 2026
heifner
left a comment
Contributor
There was a problem hiding this comment.
Targeting the wrong branch or needs to be rebased.
huangminghuang
force-pushed
the
fix/interrupt-trx-recovery
branch
from
June 27, 2026 17:54
df8a50d to
5a4d34e
Compare
heifner
approved these changes
Jun 27, 2026
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 the interrupted transaction recovery path that could leave a restarted producer stuck in LIB catchup after it had already reached the peer's known fork DB root.
Failure
Failure CI run: https://github.com/Wire-Network/wire-sysio/actions/runs/28244239953/job/83678838341
The failing job timed out in
interrupt_trx_test.pyafter the producer was relaunched. During recovery, the restarted node could receive duplicate or already-present blocks at/above the peer's known fork DB root. Because those blocks were not newly applied, the old sync path did not emit the accepted-block transition that normally advances LIB catchup to head catchup. The node stayed inlib_catchup, kept handshaking/waiting, and the test never reached the recovered producer round.Fix
sync_managerhelper that completes LIB catchup when the local fork DB root already satisfies the known peer root.interrupt_trx_test.pyso the recovery failure is reported directly instead of silently hanging until the outer CI timeout.Branch Base
This branch was rewritten to include only the delta from PR #356 to this fix, then replayed onto current
origin/master. The PR now contains one commit,5a4d34ecf1, and only these files:plugins/net_plugin/src/net_plugin.cpptests/interrupt_trx_test.pyReview
A subagent review found the missing timer cancellation in the first version of the patch. That was fixed, then re-reviewed with no remaining findings.
Validation
git diff --check origin/master...HEADcmake --build build/codex-interrupt-trx --target net_plugin nodeop clio sys-util kiod sysio_testing_contracts_copy -- -j6timeout 600 python3 tests/interrupt_trx_test.py -vNote: local validation used
-DBUILD_OPP_BUNDLES=OFFbecausewire-protobuf-bundleris not installed on this machine.