Uploaded image for project: 'OpenShift Bugs'
  1. OpenShift Bugs
  2. OCPBUGS-29391

AWS HyperShift clusters' nodes cannot join cluster with custom domain name in DHCP Option Set

XMLWordPrintable

    • Low
    • No
    • Hypershift Sprint 250, Hypershift Sprint 251, Hypershift Sprint 252
    • 3
    • False
    • Hide

      None

      Show
      None
    • Hide
      * Previously, an AWS policy issue prevented the Cluster API Provider for AWS from retrieving the necessary domain information. As a consequence, installing an AWS hosted cluster with a custom domain failed. With this update, the policy issue is resolved. (link:https://issues.redhat.com/browse/OCPBUGS-29391[*OCPBUGS-29391*])
      Show
      * Previously, an AWS policy issue prevented the Cluster API Provider for AWS from retrieving the necessary domain information. As a consequence, installing an AWS hosted cluster with a custom domain failed. With this update, the policy issue is resolved. (link: https://issues.redhat.com/browse/OCPBUGS-29391 [* OCPBUGS-29391 *])
    • Bug Fix
    • Done

      Description of problem:

      AWS HyperShift clusters' nodes cannot join cluster with custom domain name in DHCP Option Set

      Version-Release number of selected component (if applicable):

      Any

      How reproducible:

      100%

      Steps to Reproduce:

      1. Create a VPC for a HyperShift/ROSA HCP cluster in AWS
      2. Replace the VPC's DHCP Option Set with another with a custom domain name (example.com or really any domain of your choice)
      3. Attempt to install a HyperShift/ROSA HCP cluster with a nodepool

      Actual results:

      All EC2 instances will fail to become nodes. They will generate CSR's based on the default domain name - ec2.internal for us-east-1 or ${region}.compute.internal for other regions (e.g. us-east-2.compute.internal)

      Expected results:

      Either that they become nodes or that we document that custom domain names in DHCP Option Sets are not allowed with HyperShift at this time. There is currently no pressing need for this feature, though customers do use this in ROSA Classic/OCP successfully.

      Additional info:

      This is a known gap currently in cluster-api-provider-aws (CAPA) https://github.com/kubernetes-sigs/cluster-api-provider-aws/issues/1691

              jparrill@redhat.com Juan Manuel Parrilla Madrid
              mshen.openshift Michael Shen (Inactive)
              Jie Zhao Jie Zhao
              Laura Hinson Laura Hinson
              Votes:
              0 Vote for this issue
              Watchers:
              16 Start watching this issue

                Created:
                Updated:
                Resolved: