Data Centre Exit Project – Rehost and Replatform
A global brewing leader were approaching the end of their CoLo Data Centre contracts for both Zurich and Amsterdam. Both Data Centres would need to be vacated.
Each Data Centre housed workloads ranging from SharePoint, SCCM & Airwatch through to more industry specific applications such as DRAMS & Beqom – all provisioned across multiple VMware based hypervisor clusters. The brewer had already migrated their SAP estate and had encountered a large number of project issues along the way. The brewer were looking for a partner to manage the migration of the remaining workloads from each data centre and requested a proposal from bluesource for this project.
The project was challenging for a number of reasons, most notably:
The primary challenge was the project timeline due to the delays incurred as a result of a Cyber Attack. Consequently only 6 months were available to audit, document, plan, migrate & test all in scope workloads from both data centres inclusive of data erasure and hardware removal.
Dispersed Nature of Resources/Teams
Due to the vast and varied array of workloads to migrate, many resources from several organisations and multiple teams were needed. The teams were needed at various and often very specific times to perform actions which would allow for a workload to migrate successfully.
Short Notice Scope Changes
The nature of the migration activities was such that the brewer was also responsible for end to end re-platforming of specific workloads. At times a decision was made to lift and shift a workload if the re-platforming efforts were either failing or no longer deemed to be the best approach – at which point accountability for the migration would be adopted by the bluesource team.
Change Control Restrictions / Cross Organisation Integrations
The brewer has a strict change control process with deadlines on when changes should be logged for them to be discussed within the next weekly CAB meeting. Often this would present a challenge as a change with a moderate risk profile would need to be logged and be ready for review 7 days prior to the next scheduled CAB – this was not always possible as some information to make up the detail of the change may not have been available at the time.
The initial high-level strategy aligned to the Data centre exit project had the following R-Lane strategies:
|Rehost||Lift and Shift as is to Azure|
|Refactor||Re-architect the application to use cloud native PaaS services|
|Repurchase||Replace the workload with a SaaS alternative|
|Replatform||Rebuild the application on newly provisioned servers in Azure and migrate data|
|Retire||Not in scope for migration – virtual machines will simply be switched off and deleted|
|Retain||Keep the workload on-premises – move to an alternative site|
|Revise (Other)||Rebuild the application elsewhere on-premises|
The principal chosen strategies throughout the project lifecycle were Rehost & Replatform.
The project team has worked tirelessly during the lifecycle of the project with very long days both during the week and over weekends – a combination of the effort and capabilities of the team resulted in the successful migration of the final workload within bluesource project scope within a tight deadline.
bluesource has received incredibly positive feedback from the brewer on the project and the success of this project and previous projects has allowed bluesource to have ongoing future projects around their Global Modern Workplace rollout.