Settle wait only after sustained, restart-free checks - #130
Merged
Conversation
Settle only after stableChecks consecutive checks with full replicas, no active update, and no newly failed tasks. A new task failure keeps the service unsettled for a full stableChecks window, so a container that crash-loops faster than that window can never report success.
firecow
force-pushed
the
fix/wait-stability-window
branch
from
June 22, 2026 12:58
2967de2 to
d14451d
Compare
firecow
changed the title
Member
Author
|
claude --resume a603d5df-1c35-4772-a907-7d3ce96ff079 |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The
waitcommand declared success on the first poll where every service had its desired replicas running and no activeUpdateStatus. Swarm marks a taskrunningthe moment it starts, so a crash-looping container isrunningbetween restarts; a single lucky poll yielded a false green (wheatley-server 2.4.0 crash-looping in prod while the deploy job went green on retry).A service is now settled only when, for
stableChecks(default 3) consecutive checks: replicas are fully running, no active update, and no newly failed task has appeared. A new task failure puts the service on arestartingcooldown for a fullstableCheckswindow, so a container that crash-loops faster than that window can never reach success.Verified on a local single-node swarm (alpine
sleep 2; exit 1, restart loop ~6s period):Reconciliation succeededLimitation: detection needs the observation window
(stableChecks-1) × interval(10s at defaults) to exceed the crash-restart period. wheatley restarts ~7s, so defaults catch it; a slower crash-loop (period >10s) would need a higherstableChecksorinterval. Jest is unrelated/pre-broken onmain(ESM config); tsc and eslint pass.🤖 Generated with Claude Code