Skip to content

Provisioning Solutions#

Provisioning a service in SCO is the same as applying any Kubernetes resource — write a YAML claim and apply it. The platform handles the rest.

Applying a Claim#

Every service has its own resource kind and API group. Apply a claim with kubectl:

kubectl apply -f my-claim.yaml

For example, to provision a virtual machine:

apiVersion: compute.cloud.stakater.com/v1
kind: VirtualMachine
metadata:
  name: my-dev-vm
spec:
  parameters:
    flavour: rhel9
    instanceType: o1.medium
    connection: private
    sshPublicKey: "ssh-rsa AAAAB3NzaC1yc2EAAAA..."

Monitoring Status#

# Check status
kubectl get virtualmachine my-dev-vm

# Watch for changes
kubectl get virtualmachine my-dev-vm -w

# Full details and events
kubectl describe virtualmachine my-dev-vm

Typical provisioning times:

Service Typical time
Virtual Machine 2–5 minutes
OpenShift Cluster 10–20 minutes
Database 2–5 minutes

Updating a Service#

Edit the claim and reapply:

kubectl apply -f my-updated-claim.yaml

Crossplane reconciles the change. Not all fields are mutable after creation — check the API reference for details.

Deleting a Service#

kubectl delete virtualmachine my-dev-vm

This triggers a full teardown of all resources backing the service. Deletion may take a few minutes.

Warning

Deletion is irreversible. Any data stored in the service (VM disks, database contents) will be permanently removed. Back up any data you need before deleting.

Troubleshooting Stuck Provisioning#

If a service stays in Provisioning state longer than expected, check for errors:

kubectl describe virtualmachine my-dev-vm

Look at the Conditions and Events sections. Common causes:

  • Quota exceeded — request a quota increase from your administrator
  • Invalid parameter — a value was rejected; the error message will indicate which field
  • Infrastructure issue — contact your platform administrator

See Common Issues for more.

What's Next?#