Skip to content

fix: degrade condition overwritten by the next reconcile (backport #220) - #230

Merged
alicefr merged 1 commit into
release-0.1from
mergify/bp/release-0.1/pr-220
Oct 12, 2026
Merged

alicefr merged 1 commit into
release-0.1from
mergify/bp/release-0.1/pr-220

Conversation

@mergify

@mergify mergify Bot commented Oct 12, 2026

Copy link
Copy Markdown

When tag resolution fails for a BootcNodePool, the Degraded condition is correctly set but lost on the next reconcile: the blanket healthy-reset clears it and the deferred resolution path returns early without re-applying it.

Re-set the TagResolutionError condition when resolution is not yet due and TargetDigest is still empty, meaning no prior resolution has succeeded.

Fixes: #217


This is an automatic backport of pull request #220 done by Mergify.

Replace the blanket Degraded=False reset at the top of each reconcile
with per-reason clearing. Each check clears its own degraded reason on
success via clearPoolDegradedReason, and rollout-path reasons are
cleared before syncMembership/driveRollout re-evaluate them.

This avoids the False->True->False ping-pong on LastTransitionTime that
caused a spurious status write every reconcile when tag resolution was
deferred.

Assisted-by: AI
Signed-off-by: Alice Frosi <afrosi@redhat.com>
(cherry picked from commit 1c1997c)
@alicefr
alicefr merged commit 926b407 into release-0.1 Oct 12, 2026
15 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant