fix(statewrite): retry an empty or errored manifest re-fetch instead of parsing it - #542
Merged
Merged
Conversation
…of parsing it On a rejected optimistic write, CommitWithRetry re-fetches the current manifest and re-applies. The Contents API client swallows a transient GET failure and returns empty content, and that empty content was passed straight to the mutation, which parsed it and hard-failed the whole write with 'manifest file missing required ci key at top level' even though the committed manifest was intact. Treat an errored or empty re-fetch as a retryable condition, like a 409: back off and retry within the existing bounded loop rather than parsing it, and on exhaustion return a distinct non-parse error rather than a spurious missing-key error. The happy path, the conflict-merge behavior, and single-component behavior are unchanged. The git-based re-apply path already fails closed on a bad fetch and reads the manifest only after a successful reset, so it needs no change. Signed-off-by: Joshua Temple <joshua.temple@stablekernel.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Problem
On a rejected optimistic state write,
CommitWithRetryre-fetches the current manifest and re-applies. The Contents API client swallows a transient GET failure and returns empty content, and that empty content was passed straight to the mutation, which parsed it and hard-failed the entire write withmanifest file missing required 'ci' key at top level- even though the committed manifest was intact. Observed live on a fleet promote finalize; the committed manifest was verified clean, so the bad parse was on a transient empty re-fetch inside the write.Fix
Treat an errored or empty/whitespace-only re-fetch as a retryable condition, like a 409: back off and retry within the existing bounded loop rather than parsing it. On exhaustion by a persistent empty/errored re-fetch, return a distinct non-parse error rather than a spurious missing-key error. Happy path, conflict-merge behavior, and single-component behavior unchanged. The git-based
WithReapplyre-apply path already fails closed on a bad fetch (reads the manifest only after a successful reset), so it needs no change.Verification
Red-before/green-after tests: a write whose first re-fetch returns empty (and separately, errors) retries and succeeds; a persistently-empty re-fetch exhausts with a non-parse error and zero PUTs (no spurious missing-key); and a guard test that a genuinely invalid non-empty manifest still errors immediately (real invalidity not masked).
go build,go test ./...(2690),-race,golangci-lintall clean.