cacheport

Ecosystem · Containers

Helm chart repository cache for CI

Deploy pipelines and GitOps controllers fetch charts on every run. Cacheport mirrors classic chart repositories and OCI chart registries, so deploys do not fail because a chart host is slow.

CP-HELMManifest
Upstream
Any HTTP chart repository, OCI registries
Endpoint
https://acme.cacheport.cc/helm/<repo>/
Config
helm repo add with --username/--password
index.yaml
Refreshed every 15 minutes
Chart packages
Cached forever per version
OCI charts
Through Container Mirror

How Helm caching works

A classic Helm repository is an index.yaml plus a set of .tgz chart packages. You register the upstream repository in your workspace under a short name, and Cacheport serves it at /helm/<name>/.

Charts stored in OCI registries are container artifacts. They go through Container Mirror with the same token.

What gets cached
index.yaml
Refreshed every 15 minutes, stale copy on failure.
Chart packages
Cached forever per version.
Provenance files
.prov files cached alongside charts for helm verify.

Set it up

Add a mirrored repository

shell
helm repo add prometheus-community https://acme.cacheport.cc/helm/prometheus-community/ \
  --username ci --password "$CACHEPORT_TOKEN"
helm repo update
helm install monitoring prometheus-community/kube-prometheus-stack

Keep the repository name the same as the upstream one, so existing chart references like prometheus-community/kube-prometheus-stack keep working.

Use it in CI

Store the token as a secret named CACHEPORT_TOKEN, then pick your provider.

GitHub Actions

.github/workflows/deploy.yml
jobs:
  deploy:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: azure/setup-helm@v4
      - run: |
          helm repo add prometheus-community https://acme.cacheport.cc/helm/prometheus-community/ \
            --username ci --password "${{ secrets.CACHEPORT_TOKEN }}"
          helm dependency build charts/app

More for GitHub Actions: GitHub Actions setup guide.

GitLab CI

.gitlab-ci.yml
deploy:
  image: alpine/helm:3
  script:
    - helm repo add prometheus-community https://acme.cacheport.cc/helm/prometheus-community/ --username ci --password "$CACHEPORT_TOKEN"
    - helm dependency build charts/app

More for GitLab CI: GitLab CI setup guide.

GitOps controllers

Argo CD and Flux re-fetch chart indexes on their reconcile interval. Pointing them at Cacheport takes that recurring load off the upstream chart host and keeps deploys working if it goes away.

Questions

Can I mirror any chart repository?

Any repository reachable over HTTPS. Register it once in the workspace and give it a name.

Do chart digests change?

No. Chart packages are served unmodified, so digests in Chart.lock match.

Does helm push work?

No. Publish charts to your own repository or registry.

Other ecosystems people set up next

Start on the free plan

20 GB of cache, every registry, unlimited users. Upgrade when your pipelines outgrow it.