A private cloud that tenants can self-serve, built to a failure-domain map
OpenStack private cloud
Where they started
Departments were running their own small virtualisation islands with no central capacity view and no way to charge back usage.
The problem, precisely
A private cloud only works if tenants trust it enough to stop building their own, and trust requires self-service plus a cost number.
What we did
Built service by service with a tenant in the loop from the first sprint, and treated the capacity report as a deliverable rather than a side effect.
Delivered
- Kolla-Ansible deployment with HA API endpoints and documented upgrade path
- Per-tenant project, quota and network isolation with a self-service catalogue
- Volume, image and instance storage consolidated onto Ceph RBD
- Capacity and utilisation report per tenant, suitable for internal chargeback
- Upgrade and patch runbook, one minor release exercised end to end
Technologies used
Measured result
Every number below was measured and reported to the customer. Where we do not have a measured figure, the entry is a fact rather than a percentage — we would rather be precise than impressive.
- 6 tenants migrated off departmental islands
- 0 API outages during three control-plane patches
- — per-tenant utilisation, now chargeable internally
- 1 documented upgrade path instead of a rebuild
Private cloud adoption is a trust exercise. The chargeback number and the upgrade path are what convince tenants, not the API documentation.