Skip to content

OPCT - Cluster Installation Checklist

This checklist should be provided for every new support case, or when any items have been changed (for Example, Compute flavor). This will help the Red Hat engineers while working in your partner support case during the validation process.

If you have any questions you can:

Provider Information

  • Provider Name: # Example: MyCloud
  • Validation Type: # Options: (Standard | Upgrade | Disconnected)
  • OpenShift Version: # Example: 4.12.11
  • OpenShift Topology: # Example: Standard HA

Review Checklist

Deployment Documentation

  • [ ] The Cluster Diagram has been added
  • [ ] The Cluster Installation steps have been added, including any Cloud-based resources (ex.: external Load Balancers)

Compute

  • [ ] The Control Plane nodes meet the minimum requirements
  • [ ] The Compute nodes meet the minimum requirements

  • Control Plane pool flavor:

  • Compute pool flavor:
  • Public documentation with the details of the flavor offering:

See more at User Installation Review > Compute

Load Balancers

Load Balancer deployment:

  • [ ] I am using one Load Balancer dedicated to API
  • [ ] I am using one Load Balancer dedicated to Ingress
  • [ ] I am using the same Load Balancer for API and Ingress

  • Load Balancer flavor/type used by API:

  • Load Balancer flavor/type used by Ingress:

API Load Balancer details:

  • [ ] I have reviewed all the Health Check requirements
  • [ ] The Health Checks for kube-apiserver listeners are HTTP or HTTPS
  • [ ] The DNS api-int.<cluster>.<domain> is properly configured with private Load Balancer address
  • [ ] I have reviewed the Hairpin connection problem, and the Load Balancer used to kubernetes-api is not impacted by it
  • [ ] The API load balancer is configured to take a maximum of 30 seconds from the time the API server turns off the /readyz endpoint to the removal of the API server instance from the pool.

See more at User Installation Review > Load Balancers

Component-specific Review

etcd

  • [ ] I have checked the etcd logs while running OPCT
  • [ ] I ran the etcd performance tool to measure the performance of the disk used by the mount point used by etcd(/var/lib/etcd) and it reported below than 10 ms (milliseconds).

See more at User Installation Review > Components > etcd

image-registry

  • [ ] The persistent storage has been set to the Image Registry
  • Describe the name of the persistent storage used on the internal image registry:

  • [ ] I can push the image to the registry

  • [ ] I can pull images from the registry
  • [ ] I can create resources (deployment) with custom images

See more at User Installation Review > Components > image-registry