Helm3 Charts for Sonatype Products

These charts are designed to work out of the box with minikube using both ingess and ingress dns addons.

The current releases have been tested on minikube v1.12.3 running k8s v1.18.3

User Documentation

See docs/index.md which is also https://sonatype.github.io/helm3-charts/

Contributing

See the contributing document for details.

For Sonatypers, note that external contributors must sign the CLA and the Dev-Ex team must verify this prior to accepting any PR.

Updating Charts

Charts for Nexus IQ and for NXRM can be updated in charts/ directories. The most common updates will be to use new application images and to bump chart versions for release.

There should likely be no reason to update anything in docs/ by hand.

Test a chart in a local k8s cluster (like minikube) by installing the local copy from within each charts directory:

helm install --generate-name ./

Packaging and Indexing

Sonatype CI build will package, commit, and publish to the official helm repository.

Upon update of the charts/, run build.sh from here in the project root to create tgz packages of the latest chart changes and regenerate the index.yaml file to the docs/ directory which is the root of the repo site.

The build process requires Helm 3.

Testing the Helm Charts

To test Helm Charts locally you will need to follow the next steps:

1. Install docker, helm, kubectl, and minikube, if you don't already have it on your local workstation.
2. Start up minikube: minikube start
3. Confirm minikube is up and running: minikube status
4. List the existing pods in the cluster: kubectl get pods (There should not be anything listed at this point.)
5. Install the helm chart in any of these ways:
   • From a copy of the source: helm install iq {path/to/your/helm3-charts}/charts/nexus-iq
   • From our production online repo: Add our helm repo locally as instructed at https://sonatype.github.io/helm3-charts/
6. List installed servers with helm: helm list
7. Watch the server start in kubernetes by repeatedly running: kubectl get pods
8. Use the pod name you get from last command to follow the console logs: kubectl logs -f iq-nexus-iq-server-xxx
9. Confirm expected version numbers in those logs.
10. Forward a localhost port to a port on the running pod: kubectl port-forward iq-nexus-iq-server-xxx 8070
11. Connect and check that your fresh new server is successfully running: http://localhost:8070/
12. Uninstall the server with helm: helm delete iq
13. Confirm it's gone: helm list && kubectl get pods
14. Shutdown minikube: minikube stop

To run unit tests:

1. Install the unittest plugin for Helm: https://github.com/quintush/helm-unittest
2. Run the tests for each individual chart:
   • cd charts/nexus-iq; helm unittest -3 -t junit -o test-output.xml .
   • cd charts/nexus-repository-manager; helm unittest -3 -t junit -o test-output.xml .

Further Notes on Usage

Resolver File and Ingress-DNS

Use the sample values files provided here.

• helm install nexus-iq sonatype/nexus-iq-server -f iq-values.yaml
• helm install nexus-repo sonatype/nexus-repository-manager -f repo-values.yaml

If you want to use the custom values file for the demo environment that expose the apps on a local domain of *.demo which is done by creating a resolver file. On a Mac it's /etc/resolver/minikube-minikube-demo with the following entries:

domain demo
nameserver 192.168.64.8
search_order 1
timeout 5

You'll need to update the IP address to match the running instance's IP address. Use minikube ip to get the address

Docs for Ingress-dns are here https://github.com/kubernetes/minikube/tree/master/deploy/addons/ingress-dns

413 Errors with Nginx

The default setting for Nginx allows for very small upload sizes. Add this annotation to the ingress for each product to remove the limit:

nginx.ingress.kubernetes.io/proxy-body-size: "0"