Skip to content

fix: reconcile native service recovery before fleet updates - #23

Merged
jjziets merged 1 commit into
devfrom
fix/fleet-service-recovery-20261102
Nov 2, 2026
Merged

jjziets merged 1 commit into
devfrom
fix/fleet-service-recovery-20261102

Conversation

@jjziets

@jjziets jjziets commented Nov 2, 2026

Copy link
Copy Markdown
Contributor

Block new native updater work while a failed service job still needs recovery, and add a strict private receipt for verified restoration of the original IPMI container. The companion host CLI can create that receipt before adopting the newer controller; historical job and active-pointer bytes remain intact.

Submit, selection and worker checks revalidate the restored runtime and predecessor. A verified IPMI success retires its predecessor even if a later unrelated service rolls back. Failure diagnostics retain bounded phases without raw Engine errors or private configuration.

Validation: focused proxy and selector suites pass, including actual submit/worker partial-success, receipt durability and wrong-identity denial cases. Independent correctness and adversarial reviews are clear after the regression repair. Packaged Linux legacy bootstrap and live activation remain required; this PR does not update either site.

@jjziets
jjziets merged commit 608ee48 into dev Nov 2, 2026
6 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