Anonymised case studies

Two infrastructure journeys, from constraint to operations

These case studies describe delivery contexts and methods without identifying the organisations involved. Exact volumes, schedules, costs and contractual indicators are not published. No outcome figures are extrapolated.

CASE 01 · OPENSTACK

Modernising legacy OpenStack without an uncontrolled cutover

Multi-tenant enterprise environment with old releases, Ceph storage and strong continuity constraints.

ChallengeReduce release debt while preserving workloads and rollback capability.

Key decisionObjectively compare an in-place upgrade with a parallel platform before selecting the migration path.

Assessment

Map Nova, Neutron, Cinder and Keystone releases and their dependencies.
Assess Ceph pools, volume types, images, networks and external integrations.
Classify workloads by criticality, rebuild capability and acceptable outage window.

Delivery path

Inventory

Parallel target

Pilot

Waves

The target platform is prepared separately with Kolla-Ansible. A pilot group validates the migration process, acceptance criteria and rollback before workloads move in controlled waves.

Evidence and outcomes

Migration path validated before critical workloads were touched.
Cutover and rollback runbooks tested on the pilot group.
Target architecture, automation, acceptance criteria and ownership documented.
Knowledge transferred to the operating team to reduce external dependency.

CASE 02 · VMWARE → OPENSTACK

Proving VMware exit before industrialising it

Mixed Linux and Windows portfolio with varied network dependencies and application requirements.

ChallengeValidate technical and operational eligibility before building a migration programme.

Key decisionBuild a representative pilot instead of limiting the demonstration to one simple VM.

Qualification

Inventory operating systems, disks, snapshots, interfaces, flows and backup mechanisms.
Identify VirtIO drivers, guest agents and application licence constraints.
Select a panel covering several risk and dependency profiles.

Pilot path

Export

Conversion

Integration

Acceptance

Each pilot VM follows a repeatable path: controlled export, conversion, driver preparation, Glance integration, network configuration and application acceptance.

Evidence and outcomes

Eligibility matrix separating direct migration, prior remediation and rebuild.
Repeatable procedure for validated workload profiles, with pre- and post-cutover checks.
Wave plan based on dependencies rather than VM count alone.
Decision criteria that reduce uncertainty before programme costing.

Is your environment different?

A short audit identifies the right first proof before committing to a complete programme.

Frame an audit