Skip to content

fix: block rollouts on node disk pressure#595

Closed
lthibault wants to merge 2 commits into
masterfrom
fix/disk-pressure-rollout-guard
Closed

fix: block rollouts on node disk pressure#595
lthibault wants to merge 2 commits into
masterfrom
fix/disk-pressure-rollout-guard

Conversation

@lthibault

Copy link
Copy Markdown
Contributor

Summary

  • refuse master image rollouts before the Recreate deployment can disrupt service when an eligible node has DiskPressure or insufficient node/image filesystem capacity
  • check only nodes that can currently receive ww-master, retry transient preflight failures, and require an explicit completion sentinel
  • cover pressure, capacity, malformed-data, imagefs, and multi-node failure paths

Verification

  • bash tests/test_deploy_capacity_check.sh
  • bash tests/test_deploy_context.sh
  • ShellCheck

Related: wetware/infra#23

@lthibault
lthibault marked this pull request as ready for review July 23, 2026 13:37
@lthibault

Copy link
Copy Markdown
Contributor Author

Superseded by #596’s manual promotion boundary and infra#26’s guarded manual rollout path. Do not merge; this PR restores the direct CI deployment path that #596 intentionally removed.

@lthibault lthibault closed this Jul 23, 2026
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