The Problem

The NHS had implemented a SAN solution for their Work Area Recovery (WAR) platform using a number of third party products and technologies. The implemented platform for virtualisation was VMware ESX to host virtual machines.

The implemented solution did not cater for recovering virtual machines in the event of a disaster on a data centre level. The IT staff in the NHS needed to manually find the .vmx files of active machines and register all the virtual machines back into work area recovery Virtual Center: a very time-consuming task.

The Challenge

To automate the integration of the current SAN and VMware environment across the two Virtual Centre servers by scanning the datastore and locating VMFS volumes for all virtual machines in the datastore and provide an automatic process to register the virtual machines at the recovery site and start, in priority order, those virtual machines configured to start.

The Solution

Our solution was to use VMware Virtual Service - Orchestrator (VS-O), now re-branded Lifecycle Manager, which is server-based software that provides for the creation and execution of automated, configurable processes for managing virtualised computing environments.

Results

The Tribune process invokes the SAN failover process from the ‘live’ to the ‘WAR’ environments and then scans the datastore restarting any virtual machine discovered on the WAR datastore. Once all virtual machines have been restarted an email notification is sent to the administrator detailing actions taken.

Benefits

By implementing an "invocation workflow" the NHS can now be confident that when they need to invoke Work Area Recovery either in a test scenario or in production, the virtual machines in use at the time of the invocation will be started in the shortest possible time and without impact to their production users.

Copyright 2008 Tribune Business Systems. All rights reserved.