Deployment with 1controller and 1 compute in CI is failing with "error": "No IPs found for Controller role on ctlplane network"

[Sandeep Yadav]

Context:

Validation squad is trying to create a new multinode job using 1controller and 1 compute nodeset.

Patch with new job definition: https://review.rdoproject.org/r/c/rdo-jobs/+/36416

Failure:-

https://logserver.rdoproject.org/17/36417/3/check/periodic-tripleo-ci-centos-8-containers-multinode-network-master-validation/ba8a0dc/logs/undercloud/home/zuul/overcloud_deploy.log.txt.gz

~~~
2021-11-08 10:09:26 | 2021-11-08 10:09:26.694757 | fa163eea-27a6-466c-f979-00000000000d | FATAL | Generate ansible inventory | localhost | error={"changed": false, "error": "No IPs found for Controller role on ctlplane network", "msg": "Error generating inventory for overcloud: No IPs found for Controller role on ctlplane network", "success": false}
~~~

As per KCS: https://access.redhat.com/solutions/6310901

This type of issue can occur when HostnameMap doesn't match the DeployedServerPortMap.

This seems the case as per templates:-

https://logserver.rdoproject.org/17/36417/3/check/periodic-tripleo-ci-centos-8-containers-multinode-network-master-validation/ba8a0dc/logs/undercloud/home/zuul/config-download.yaml.txt.gz

~~~
  DeployedServerPortMap:
    node-0001753171-ctlplane: >>>>>>>>>>>>>>>>>>>>> node-0001753171-ctlplane
      fixed_ips:
        - ip_address: 192.168.24.3
      subnets:
        - cidr: 192.168.24.0/24
      network:
        tags:
          - 192.168.24.0/24
    node-0001753172-ctlplane:
      fixed_ips:
        - ip_address: 192.168.24.4
      subnets:
        - cidr: 192.168.24.0/24
      network:
        tags:
          - 192.168.24.0/24
~~~

https://logserver.rdoproject.org/17/36417/3/check/periodic-tripleo-ci-centos-8-containers-multinode-network-master-validation/ba8a0dc/logs/undercloud/home/zuul/hostnamemap.yaml.txt.gz

~~~
parameter_defaults:
  HostnameMap:
    overcloud-controller-0: overcloud-controller-foo-0 >>>>>>>>>>>>>>>>> overcloud-controller-foo-0
    overcloud-novacompute-0: overcloud-novacompute-bar-0
~~~

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.opendev.org/c/openstack/tripleo-quickstart/+/817032

[Sandeep Yadav (sandeepyadav93)]

containers-multinode job have correct HostnameMap in sync with DeployedServerPortMap:-

https://logserver.rdoproject.org/openstack-periodic-integration-main/opendev.org/openstack/tripleo-ci/master/periodic-tripleo-ci-centos-8-containers-multinode-master/c2aaf68/logs/undercloud/home/zuul/hostnamemap.yaml.txt.gz

~~~
parameter_defaults:
  HostnameMap:
    overcloud-controller-0: node-0001753128
~~~

https://logserver.rdoproject.org/openstack-periodic-integration-main/opendev.org/openstack/tripleo-ci/master/periodic-tripleo-ci-centos-8-containers-multinode-master/c2aaf68/logs/undercloud/home/zuul/config-download.yaml.txt.gz

~~~
  DeployedServerPortMap:
    node-0001753128-ctlplane:
      fixed_ips:
        - ip_address: 192.168.24.3
      subnets:
        - cidr: 192.168.24.0/24
      network:
        tags:
          - 192.168.24.0/24
~~~

In 1ctlr.yml[1] we are generating HostnameMap in an automated way:-

https://opendev.org/openstack/tripleo-quickstart/src/branch/master/config/nodes/1ctlr.yml#L6

~~~
    hostnamemap_override: "{{ hostvars[groups['overcloud'][0]].ansible_hostname }}"
~~~

1ctlr_1comp.yml have harcoded value, which seems to be the issue-

https://opendev.org/openstack/tripleo-quickstart/src/branch/master/config/nodes/1ctlr_1comp.yml#L7

~~~
    hostnamemap_override: overcloud-controller-foo-0
~~~

[OpenStack Infra (hudson-openstack)]

Change abandoned by "Sandeep Yadav <email address hidden>" on branch: master
Review: https://review.opendev.org/c/openstack/tripleo-quickstart/+/817032
Reason: Didn't work