Skip to content

Getting started

In a nutshell

  1. Deploy ORC to your Kubernetes cluster
  2. Create a Kubernetes secret containing a clouds.yaml
  3. Create an OpenStackCloud object pointing to that clouds.yaml
  4. Deploy your OpenStack infrastructure as Kubernetes custom resources

Deploy ORC to your Kubernetes cluster

From the git repository, run:

make deploy IMG=quay.io/orc/openstack-resource-controller

Create a Kubernetes secret containing a clouds.yaml

kubectl create secret generic openstack-clouds \
    --from-file=${XDG_CONFIG_HOME}/openstack/clouds.yaml

Note:

The command above will upload your entire clouds.yaml to your Kubernetes cluster! If that is not appropriate, you may want to upload a slimmed version of it.

Create an OpenStackCloud object pointing to that clouds.yaml

Your clouds.yaml contains a YAML dictionary of one or more clouds. If the name of the cloud you want to target is, for example, openstack-one, your OpenStackCloud object should look like this:

kubectl apply -f- <<EOF
apiVersion: openstack.k-orc.cloud/v1alpha1
kind: OpenStackCloud
metadata:
  labels:
    app.kubernetes.io/name: openstackcloud
    app.kubernetes.io/instance: openstackcloud-gettingstarted
    app.kubernetes.io/part-of: gettingstarted
  name: osp1
spec:
  cloud: openstack-one # <-- replace with your cloud name in clouds.yaml
  credentials:
    source: secret
    secretRef:
      name: openstack-clouds
      key: clouds.yaml
EOF

Check that the resource is ready in Kubernetes:

kubectl get OpenStackCloud osp1

If the credentials are valid, you get:

$ kubectl get OpenStackCloud osp1
NAME   READY   ERROR   STATUS
osp1   True    False   Ready

Deploy your OpenStack infrastructure as Kubernetes custom resources

This is the definition of a subnet. You can apply it to your cloud:

kubectl apply -f- <<EOF
apiVersion: openstack.k-orc.cloud/v1alpha1
kind: OpenStackSubnet
metadata:
  labels:
    app.kubernetes.io/name: openstacksubnet
    app.kubernetes.io/instance: openstacksubnet-gettingstarted
    app.kubernetes.io/part-of: gettingstarted
  name: subnet-1
spec:
  cloud: osp1
  resource:
    name: subnet-1
    network: network-1
    allocationPools:
    - start: 192.168.1.5
      end: 192.168.1.60
    cidr: 192.168.1.0/24
    ipVersion: IPv4
EOF

The controller will only attempt creating the subnet when the corresponding OpenStackNetwork exists:

$ kubectl get openstacksubnet
NAME       READY   ERROR   STATUS
subnet-1   False   False   Waiting for the following dependencies to be ready: network:default/network-1

Let's create it:

kubectl apply -f- <<EOF
apiVersion: openstack.k-orc.cloud/v1alpha1
kind: OpenStackNetwork
metadata:
  labels:
    app.kubernetes.io/name: openstacknetwork
    app.kubernetes.io/instance: openstacknetwork-gettingstarted
    app.kubernetes.io/part-of: gettingstarted
  name: network-1
spec:
  cloud: osp1
  resource:
    name: network-1
EOF

After a few seconds, both resources become ready:

$ kubectl get openstacknetwork
NAME        READY   ERROR   STATUS
network-1   True    False   Ready

$ kubectl get openstacksubnet
NAME       READY   ERROR   STATUS
subnet-1   True    False   Ready

The subnet can be inspected through its Kubernetes representation, under .status.resource:

kubectl get openstacksubnet subnet-1 -o yaml

apiVersion: openstack.k-orc.cloud/v1alpha1
kind: OpenStackSubnet
metadata:
  # [...]
  finalizers:
  - openstacksubnet.k-orc.cloud
  labels:
    app.kubernetes.io/instance: openstacksubnet-gettingstarted
    app.kubernetes.io/name: openstacksubnet
    app.kubernetes.io/part-of: gettingstarted
    cloud.openstack.k-orc.cloud/osp1: ""
    network.openstack.k-orc.cloud/network-1: ""
  name: subnet-1
  namespace: default
  uid: 9f256fd8-91fb-4c6d-a1fa-8a4bda1a92f5
spec:
  cloud: osp1
  resource:
    allocationPools:
    - end: 192.168.1.60
      start: 192.168.1.5
    cidr: 192.168.1.0/24
    ipVersion: IPv4
    name: subnet-1
    network: network-1
status:
  conditions:
  - message: Ready
    reason: Ready
    status: "True"
    type: Ready
  - message: ""
    reason: NoError
    status: "False"
    type: Error
  resource:
    allocationPools:
    - end: 192.168.1.60
      start: 192.168.1.5
    cidr: 192.168.1.0/24
    enableDHCP: true
    gatewayIP: 192.168.1.1
    id: 0e4a1c78-d5c3-4670-88ba-880cb56c4188
    ipVersion: 4
    name: subnet-1
    networkID: 9d4ecc67-ea5b-4344-b919-2051f0255c06
    projectID: 90dce24f8e6748bfbc319a9223d0a7a6
    tenantID: 90dce24f8e6748bfbc319a9223d0a7a6

Reset

To reset both Kubernetes and Openstack to their original state, delete the resources and undeploy ORC:

kubectl delete OpenStackSubnet subnet-1
kubectl delete OpenStackNetwork network-1
kubectl delete OpenStackCloud osp1
kubectl delete secret openstack-clouds
make undeploy