Helm release status and CI/CD log correlation for post‑install hook failures

Question

Determine whether a failing post‑install hook changes the Helm release status reported by helm status and whether that change is observable in the CI/CD pipeline logs without enabling additional debugging.

In Helm v3 each release is persisted as a Secret, and hook execution logs are written to the hook’s pod rather than to the release resource. The CI system typically captures the helm command’s stdout and stderr, but hook output appears only if the pod logs are collected separately. It is unclear whether Helm automatically marks the release as FAILED when a hook errors, leaving the status as DEPLOYED while the failure is visible only in hook logs.

Does Helm automatically set the release status to FAILED when a post‑install hook returns a non‑zero exit code? If the status stays DEPLOYED, what specific log lines in the CI output indicate the hook failure? Can the CI pipeline rely on helm status alone to judge deployment success, or must it also inspect hook logs?

Tasadduq Burney · Answer

Confirmed facts
In Helm v3 a failing post‑install hook causes the release to be marked FAILED. The helm install or helm upgrade command exits with a non‑zero status and its stderr contains an error line such as:
Error: post‑install hook failed: …
or
hook  failed: …
The release Secret that helm status reads reflects this state, so helm status  shows STATUS: FAILED.
Likely explanation
Hook execution is performed synchronously during the install/upgrade sequence; it is not a background activity that Helm ignores. Therefore a hook error aborts the release transaction and updates the release record to FAILED. The --wait flag only influences waiting for non‑hook resources to become ready; it does not change how hook failures are reported.
Steps for this case

Run the Helm command exactly as used in your CI (e.g., helm install --wait … or helm upgrade …).
Capture the command’s exit code and its combined stdout/stderr.
If the exit code is non‑zero and the output contains the hook‑failure error line above, treat the deployment as failed.
Optionally run helm status  to confirm that the STATUS field reads FAILED; this matches the CI observation.
No additional debugging or hook‑pod log collection is required to determine success/failure.

Helm release status and CI/CD log correlation for post‑install hook failures

1 answer

Confirmed facts

Likely explanation

Steps for this case

Your perspective belongs here.

0 question comments