Repository navigation
test(e2e): quarantine pre-existing "untriggered" HA flake (re-trigger), track root-cause in #89 - #90
Merged
Merged
Conversation
…a pre-existing HA flake On the multi-node e2e stack a freshly triggered workflow intermittently stays in "untriggered" (trigger accepted, actor not spawned/claimed in time) until the test times out. Same commit observed both passing and failing, and the failure predates the ADR-0031 Phase 3 work across unrelated workflows — i.e. a pre-existing HA trigger/claim flake, not a workflow-logic regression. TriggerAndWaitTerminal re-triggers only when a workflow is stuck specifically in "untriggered" (each trigger yields a fresh id); any terminal state (incl. error) is returned as-is so real failures still surface. Wired into the execution suite + sum-rand-branch. Root cause tracked in #89. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This branch was previously 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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Why
On the multi-node e2e stack (
docker compose --profile e2e), a freshly triggered workflowintermittently stays in
untriggereduntil the test times out — the trigger is accepted but theactor isn't spawned/claimed in time. This is a pre-existing HA flake, not a regression:
df827dcboth passed and failed E2E within ~2h;still in status "untriggered"failure occurred on16867aa(before the ADR-0031Phase 3 work), across multiple unrelated workflows.
Root cause is tracked in #89.
Change
TriggerAndWaitTerminalre-triggers a workflow only when it is stuck specifically inuntriggered(each trigger yields a fresh id; bounded to 3 attempts). Any terminal state —including
error— is returned as-is, so genuine failures still surface. Wired into the workflowexecution suite and
TestSumRandBranch_Finishes.No production code changes; e2e test harness only.
🤖 Generated with Claude Code