Flux HelmRelease "install retries exhausted" in CI

helm-controller attempted the release, it failed, and after the configured spec.install.remediation.retries the controller gave up. The real Helm error is in the release history and the HelmRelease status, above the "retries exhausted" summary.

What this error means

A HelmRelease reports "False" with "install retries exhausted" or "upgrade retries exhausted", and flux get helmreleases shows it not Ready.

Helm install failed: install retries exhausted

Common causes

The chart install itself keeps failing

A templating error, a failing hook, or an invalid resource makes each attempt fail, so retries are exhausted with the same underlying cause.

A transient failure exceeded the retry budget

A slow dependency or image pull failed on every attempt within the retry window, so the controller stopped retrying.

How to fix it

Read the underlying Helm error

1. Describe the HelmRelease to see the real Helm error under the summary.
2. Check helm-controller logs for the failing template, hook, or resource.
3. Fix the cause, then reconcile the HelmRelease.

kubectl -n flux-system describe helmrelease web
flux logs --kind HelmRelease --name web
flux reconcile helmrelease web --with-source

Reset a failed release before retrying

If the release is stuck in a failed state, suspend and resume, or increase retries, once the root cause is fixed.

flux suspend helmrelease web
flux resume helmrelease web

How to prevent it

• Fix the underlying Helm failure; retries alone will not recover a broken chart.
• Set remediation.retries to a value that covers transient failures, not permanent ones.
• Reconcile with the source so the release uses the intended chart version.

Related guides