create manifests then create cluster panic

[e-minguez]

Version

$ openshift-install version
openshift-install unreleased-master-246-g193a3f2cb7b8da375da90126ac4757254cdb17fe

Platform (aws|libvirt|openstack):

libvirt

What happened?

• Installer compiled with libvirt support
• install-config.yaml content:

apiVersion: v1beta1
baseDomain: ocp4.lan
machines:
- name: master
  platform: {}
  replicas: 1
- name: worker
  platform: {}
  replicas: 1
metadata:
  creationTimestamp: null
  name: minwi
networking:
  clusterNetworks:
  - cidr: 10.128.0.0/14
    hostSubnetLength: 9
  machineCIDR: 192.168.126.0/24
  serviceCIDR: 172.30.0.0/16
  type: OpenshiftSDN
platform:
  libvirt:
    URI: qemu+tcp://192.168.122.1/system
    network:
      if: tt0
pullSecret: '{"auths":{"cloud.openshift.com":{"auth":"x","email":"x"},"quay.io":{"auth":"x","email":"x"}}}'
sshKey: |
  ssh-rsa x x@x.com

• mkdir assets
• cp install-config.yaml assets/
• openshift-install create manifests --dir=./assets/ --log-level=debug
• Manifests generated:

assets/
assets/.openshift_install.log
assets/.openshift_install_state.json
assets/manifests
assets/manifests/04-openshift-machine-config-operator.yaml
assets/manifests/09-openshift-service-signer-namespace.yaml
assets/manifests/cluster-config.yaml
assets/manifests/cluster-dns-01-crd.yaml
assets/manifests/cluster-dns-02-config.yml
assets/manifests/cluster-infrastructure-01-crd.yaml
assets/manifests/cluster-infrastructure-02-config.yml
assets/manifests/cluster-ingress-01-crd.yaml
assets/manifests/cluster-ingress-02-config.yml
assets/manifests/cluster-network-01-crd.yml
assets/manifests/cluster-network-02-config.yml
assets/manifests/cvo-overrides.yaml
assets/manifests/etcd-service.yaml
assets/manifests/host-etcd-service-endpoints.yaml
assets/manifests/host-etcd-service.yaml
assets/manifests/kube-cloud-config.yaml
assets/manifests/kube-system-configmap-etcd-serving-ca.yaml
assets/manifests/kube-system-configmap-root-ca.yaml
assets/manifests/kube-system-secret-etcd-client.yaml
assets/manifests/machine-config-server-tls-secret.yaml
assets/manifests/openshift-service-signer-secret.yaml
assets/manifests/pull.json
assets/openshift
assets/openshift/99_binding-discovery.yaml
assets/openshift/99_kubeadmin-password-secret.yaml
assets/openshift/99_openshift-cluster-api_cluster.yaml
assets/openshift/99_openshift-cluster-api_master-machines.yaml
assets/openshift/99_openshift-cluster-api_master-user-data-secret.yaml
assets/openshift/99_openshift-cluster-api_worker-machineset.yaml
assets/openshift/99_openshift-cluster-api_worker-user-data-secret.yaml

• openshift-install create cluster --dir=./assets/ --log-level=debug
  Fails with:

DEBUG   Fetching "Master Ignition Config"...       
DEBUG   Reusing previously-fetched "Master Ignition Config" 
DEBUG   Fetching "Master Machines"...              
DEBUG   Reusing previously-fetched "Master Machines" 
DEBUG Generating "Terraform Variables"...          
panic: interface conversion: runtime.Object is nil, not *v1alpha1.LibvirtMachineProviderConfig

goroutine 1 [running]:
github.com/openshift/installer/pkg/asset/cluster.(*TerraformVariables).Generate(0x8acca90, 0xc420b8b0e0, 0xc420ec79f0, 0x2)
	/home/edu/go/src/github.com/openshift/installer/pkg/asset/cluster/tfvars.go:120 +0x136f
github.com/openshift/installer/pkg/asset/store.(*storeImpl).fetch(0xc420a20e40, 0x53717a0, 0x8acca90, 0x0, 0x0, 0x46633a0, 0x47b48a0)
	/home/edu/go/src/github.com/openshift/installer/pkg/asset/store/store.go:226 +0x7c0
github.com/openshift/installer/pkg/asset/store.(*storeImpl).Fetch(0xc420a20e40, 0x53717a0, 0x8acca90, 0x53717a0, 0x8acca90)
	/home/edu/go/src/github.com/openshift/installer/pkg/asset/store/store.go:76 +0x4b
main.runTargetCmd.func1(0x7fff94670566, 0x9, 0xc420a81f00, 0xc42049fbf8)
	/home/edu/go/src/github.com/openshift/installer/cmd/openshift-install/create.go:157 +0x13f
main.runTargetCmd.func2(0x8986700, 0xc420a81da0, 0x0, 0x2)
	/home/edu/go/src/github.com/openshift/installer/cmd/openshift-install/create.go:182 +0x81
github.com/openshift/installer/vendor/github.com/spf13/cobra.(*Command).execute(0x8986700, 0xc420a81d60, 0x2, 0x2, 0x8986700, 0xc420a81d60)
	/home/edu/go/src/github.com/openshift/installer/vendor/github.com/spf13/cobra/command.go:766 +0x2c1
github.com/openshift/installer/vendor/github.com/spf13/cobra.(*Command).ExecuteC(0xc420a34780, 0x0, 0xc420a35900, 0xc420a348d0)
	/home/edu/go/src/github.com/openshift/installer/vendor/github.com/spf13/cobra/command.go:852 +0x30a
github.com/openshift/installer/vendor/github.com/spf13/cobra.(*Command).Execute(0xc420a34780, 0xc42049feb8, 0x1)
	/home/edu/go/src/github.com/openshift/installer/vendor/github.com/spf13/cobra/command.go:800 +0x2b
main.installerMain()
	/home/edu/go/src/github.com/openshift/installer/cmd/openshift-install/main.go:56 +0x132
main.main()
	/home/edu/go/src/github.com/openshift/installer/cmd/openshift-install/main.go:40 +0x9b

What you expected to happen?

Create a proper OCP4 cluster on libvirt.

How to reproduce it (as minimally and precisely as possible)?

See before

Anything else we need to know?

Running only create cluster works:

• Installer compiled with libvirt support
• install-config.yaml content:

apiVersion: v1beta1
baseDomain: ocp4.lan
machines:
- name: master
  platform: {}
  replicas: 1
- name: worker
  platform: {}
  replicas: 1
metadata:
  creationTimestamp: null
  name: minwi
networking:
  clusterNetworks:
  - cidr: 10.128.0.0/14
    hostSubnetLength: 9
  machineCIDR: 192.168.126.0/24
  serviceCIDR: 172.30.0.0/16
  type: OpenshiftSDN
platform:
  libvirt:
    URI: qemu+tcp://192.168.122.1/system
    network:
      if: tt0
pullSecret: '{"auths":{"cloud.openshift.com":{"auth":"x","email":"x"},"quay.io":{"auth":"x","email":"x"}}}'
sshKey: |
  ssh-rsa x x@x.com

• mkdir assets
• cp install-config.yaml assets/
• openshift-install create cluster --dir=./assets/ --log-level=debug

...
DEBUG                                              
DEBUG State path: /tmp/openshift-install-168351161/terraform.tfstate 
INFO Waiting up to 30m0s for the Kubernetes API... 
...

References

[wking]

This or similar is probably the issue (and it would be cross-platform). I'll work on this today.

[dgoodwin]

Affecting hive as well, we're getting this on every execution and we're mounting in install-config for the installation pods, so this is consistent with what was reported above.