Repository navigation
feat(storage): add the dormant Longhorn worker preparation step - #5686
Conversation
Slice 4a of the split of the V2 preparation change. Adds the guarded prepare-worker operation, its journal and the deploy gate's allowance for a cordon that journal holds. Talos layout changes stay in the next slice, so this has no live effect: a dispatch holds before any write. Part of #5655 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@coderabbitai full review |
Action not completed
|
|
Warning Review limit reachedYou've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository. Next included review available in 54 minutes. View limit details
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. |
@codex review for safety of the destructive worker reset path: journal recovery, what may run before the reset, and whether any dispatch can write on today's configuration |
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
|
Codex Review: Didn't find any major issues. Already looking forward to the next diff. Reviewed commit: ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
If Codex has suggestions, it will comment; otherwise it will react with 👍. Codex can also answer questions or update the PR. Try commenting "@codex address that feedback". |
Evaluation at
Not proven: nothing here has run on the cluster, and no real Talos volume read has been made through these checks. The one shared-path change is that deploy gate, which only widens when a matching journal exists, and none does. |
Why
Moving storage to Longhorn V2 needs each permanent worker's local disk rebuilt, and that is the most dangerous step of the migration. The large change that first held this step never converged in review, because a finding anywhere restarted review of everything, so this step now gets its own review.
What
Adds the guarded step that empties a worker, wipes only its temporary partition and brings it back, with a saved record so an interrupted run recovers instead of repeating the wipe. It also lets a deploy pass a worker that this step is deliberately holding. Nothing runs on the cluster from this change: the step refuses before writing anything until the new disk layout, the V2 capacity and the host settings land in later changes.
Part of #5655