Rewrite production lab guide for OC-managed provisioning
Rewrites sections 0-6 from scratch with OC-managed clustering (tokens,
OC ISOs, auto-registration) replacing manual cluster formation. Updates
sections 7-15 with correct naming (oc-node-* instead of lab-node-*),
fixes device override bug, and adds sequential launch guidance.
Key changes:
- Architecture: new VMID scheme (920/400-402), IP scheme (.120/.140-142)
- Node-01 disk increased to 250 GiB for Aether golden image
- OC provisioning flow: token → older-channel ISO → auto-update
- Fixed root device config: device override instead of device set
- Added image download timing notes (sequential, not concurrent)
- Updated OC Dashboard section for OC-managed deployment
- Updated node replacement procedure with correct VMIDs/IPs
- Updated cleanup to use --cleanup-all --deep --yes
Full lab validated from scratch: deploy, cluster, OVN, workloads,
isolation, peering, load balancing, live migration, evacuation/restore.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Maarten
6d193158d5