When working with Kubernetes, one of the fastest ways to roll back a broken release is to use git reset to point your manifests back to a safe commit, then redeploy the Helm chart. This approach cuts through confusion by keeping your code and chart definitions perfectly in sync.
Reset Code and Helm State
Start by finding the commit hash from before the failed chart deployment. Check your Git history:
git log --oneline
Reset your branch:
git reset --hard <commit-hash>
This moves your local tree back to exactly what it was at the known good state.
Redeploy the Chart
With your manifests restored, redeploy:
helm upgrade --install <release-name> <chart-path> --namespace <namespace>
This ensures the Helm release matches the code you just checked out. Every manifest reverts cleanly because you rewound both Git and Helm to the same point in time.
Verify the Rollback
Run:
helm status <release-name>
kubectl get pods --namespace <namespace>
Check logs for anomalies and confirm that containers are running with expected versions.
Why git reset + Helm Works
A broken deployment often comes from mismatched chart templates and application code. git reset aligns the repo exactly, and redeploying with Helm applies those templates without leftover state or drift. It’s faster than editing values files mid-crisis.
Keep your production charts tracked in Git. This guarantees that every rollback can be replicated, tested, and deployed with confidence.
See this workflow live in minutes with hoop.dev — connect your repo, run the commands, and watch it roll back, clean and simple.