SaaS Data Centre Migration

SaaS UK Data Centre Migration

This global corporation, with Headquarters in Mountain View USA, was running a programme of transformational change to optimise their use of co-located data-centre space. The programme was focussed on ensuring that their global network of data-centres could handle the capacity plans that had been projected for the next 3 years. One of the key deliverables was the migration of services from the largest site in the company’s European cluster to a new facility.

Accentium_Datacentre

The Challenge

One of the data-centres due for expansion in the UK was effectively full and could not offer the space needed to satisfy the new capacity requirement. The programme needed to source a new facility and perform a complex migration of all the SaaS infrastructure whilst maintaining the 100% service availability target that was in place with the company’s global client base.

The UK site in question was one of the largest in the company’s operation, with over 60 racks of equipment, including enterprise-class servers and networking equipment. It also hosted mission critical systems such as the web based administration portal used by the global client base.

About the Project

The project required a team to be established across many functions including service delivery, client support, platform support, network, security and product management. Two parallel streams were kicked off initially. The first was focused on sourcing the new co-located space within a dedicated suite that would comply with the company’s ISO27001 and SSAE 16 accreditation, whilst the second performed an audit of all the systems and services which would be impacted. The audit revealed several gaps in current service procedures which needed to be filled before the migration could take place. The gaps identified by the audit were resolved and detailed analysis was undertaken to establish an accurate set of dependencies which would inform the day by day migration plan. It was imperative that the plans took into account the 100% service availability target.

The migration took place over 6 weeks, and was split into 5 phases. This approach enabled both sites to be run in parallel whilst the migration took place, with services effectively split across the two locales until the final network migration had been completed.

Phase 1 – Build of co-location space, core infrastucture (power, HVAC etc), and core network

Phase 2 –  Pilot move to prove the logistics, refine the outage timings and identify improvements to the process

Phase 3 – Servers running email and web based scanning services that could be load balanced across other data-centres to maintain cluster capacity

Phase 4 – Servers and SAN controlling the web administration portal

Phase 5 – Network migration

A ticketing system was used to help track the migration of each system. This covered the lifecycle of every movement, from system shutdown through to re-commissioning.

Our Role

Our role was to lead the Programme from start to finish; building and managing the large cross functional team to ensure that the objectives were met. We also had to set up and manage 3rd party relationships with the new data-centre management team and logistics partner to ensure that their elements of delivery aligned with the plan.

The programme’s collaborative approach and ultimately successful delivery led to a global award from the company for ‘Cross Functional Team Work’.