Created attachment 1903905 [details] journalctl log from node0 Description: I was testing the DHCP scenario where only rendezvousIP is specified in the agent-config.yaml and no NMStateConfig is embedded. create-cluster-and-infraenv.service fails on node0 when networkConfig is missing from agent-config.yaml. /etc/assisted/manifests/nmstateconfig.yaml is an empty file. agent-config.yaml used: metadata: name: ostest namespace: cluster0 spec: rendezvousIP: 192.168.122.2 Steps to reproduce: 1. Create agent.iso using install-config.yaml and agent-config.yaml 2. Deploy cluster using agent.iso 3. Log into node0 and create-cluster-and-infraenv.service will be displayed as a failed unit. Expected: create-cluster-and-infraenv.service in success state Actual: create-cluster-and-infraenv.service in failed state Aug 05 08:27:59 control1 podman[2681]: time="2022-08-05T08:27:59Z" level=info msg="releaseImage version 4.11.0-0.okd-2022-08-04-074610 cpuarch x86_64" Aug 05 08:27:59 control1 create-cluster-and-infraenv[2693]: time="2022-08-05T08:27:59Z" level=info msg="Registered cluster with id: 1cc3ea1a-5bbc-4c4d-ad66-6e052800fb0c" Aug 05 08:27:59 control1 create-cluster-and-infraenv[2693]: time="2022-08-05T08:27:59Z" level=info msg="Registering infraenv" Aug 05 08:27:59 control1 podman[2681]: time="2022-08-05T08:27:59Z" level=info msg="Registered cluster with id: 1cc3ea1a-5bbc-4c4d-ad66-6e052800fb0c" Aug 05 08:27:59 control1 podman[2681]: time="2022-08-05T08:27:59Z" level=info msg="Registering infraenv" Aug 05 08:27:59 control1 create-cluster-and-infraenv[2693]: time="2022-08-05T08:27:59Z" level=fatal msg="Failed to register infraenv with assisted-service: nmstateconfig should have at least one label set matching the infra-env label selector" Aug 05 08:27:59 control1 podman[2681]: time="2022-08-05T08:27:59Z" level=fatal msg="Failed to register infraenv with assisted-service: nmstateconfig should have at least one label set matching the infra-env label selector" Aug 05 08:27:59 control1 systemd[1]: create-cluster-and-infraenv.service: Main process exited, code=exited, status=1/FAILURE Aug 05 08:27:59 control1 systemd[1]: create-cluster-and-infraenv.service: Failed with result 'exit-code'. Aug 05 08:27:59 control1 systemd[1]: Failed to start Service that creates initial cluster and infraenv. /etc/assisted/manifests/nmstateconfig.yaml is an empty file. [core@control1 ~]$ sudo cat /etc/assisted/manifests/nmstateconfig.yaml [core@control1 ~]$
Since the create-cluster-and-infraenv code is in the release payload, it's effectively not fixable for installing the 4.11 payload. We'll need to either create a dummy NMStateConfig to keep it happy, or maintain the restriction that at least 1 network config must be supplied (even if the rendezvousIP is also set).
PRs for review: https://github.com/openshift/installer/pull/6299 https://github.com/openshift/assisted-service/pull/4349
Bug has been verified with 4.12.0-0.nightly-2022-09-26-111919 release. create-cluster-and-infraenv service started without any issue when only provided rendezvousIP in the agent-config.
OCP is no longer using Bugzilla and this bug appears to have been left in an orphaned state. If the bug is still relevant, please open a new issue in the OCPBUGS Jira project: https://issues.redhat.com/projects/OCPBUGS/summary