Orchestrating a PingAccess Cluster Deployment

Use kustomize for the PingAccess cluster deployment from your local pingidentity-devops-getting-started/20-kubernetes/04-clustered-pingaccess directory (the location of the YAML files), and call into your local pingidentity-devops-getting-started/20-kubernetes/01-standalone directory for the base product configurations.

Use the server profile in our pingidentity-server-profiles/pa-clustering repository.

We use separate deployments for the PingAccess admin node (env_vars.pingaccess) and the PingAccess engine node (env_vars.pingaccess-engine and pingaccess-engine.yaml). To scale out replicas, use the PingAccess engine node.

The env_vars.pingaccess and env_vars.pingaccess-engine files:

• Contain the environment variables to use for pingidentity-server-profiles/pa-clustering
• Set the clustering (operational) mode for each deployment: CLUSTERED_CONSOLE for pingaccess and CLUSTERED_ENGINE for pingaccess-engine

kustomization.yaml does the following:

• References your local pingidentity-devops-getting-started/20-kubernetes/01-standalone/pingaccess directory for the base product configurations
• Uses patches to remove the pingaccess engine port (3000)
• Replaces the environment variables in the parent configMap with those in the specified env_vars.pingaccess and env_vars.pingaccess-engine files

Deploying the Cluster

1. To orchestrate the replicated PingAccess deployment, from your local pingidentity-devops-getting-started/20-kubernetes/04-clustered-pingaccess directory, enter:

   kustomize build . | kubectl apply -f -
   

2. Scale up the engines:

   kubectl scale deployment pingaccess-engine --replicas=2
   

Cleaning Up

To clean up when you're finished, enter:

kustomize build . | kubectl delete -f -