Phase 7.7 — First node on the new platform

The first server comes back, no longer on VMware.

Network firmware updated, drives refitted to the new layout, storage arrays rebuilt, the new platform installed and configured from code, and the workloads imported back — thin-provisioned this time, which reclaims a substantial amount of space that was being held and not used.

Then it simply runs for a week before anything else is touched. That week is the rollback position: the other two servers are still on the old platform and still hold everything else, so if the new one disappoints, nothing has been lost but time.

The management cluster comes back here.