Bladewatch.com

Chris asked me what things I would consider when implementing and completing a virtualization project for a large organization. Key to both issues would be what the project were to achieve, I know many companies have tried to do both, to virtualize some systems and use it as a server provisioning tool.

That’s fine vmware can accommodate that, but the project needs to be done in two phases:

1. Server provisioning
2. Server consolidation or virtualization

Two phases unless you’ve got the budget, business buy in, the CIO and CFO agreeing that this is the way forward, the way to continue to deliver the business need without buying/renting more datacenter space.

We must prove that we can provide a development team/application team with: Stable/scalable platform for business which works and can perform to their needs Flexible infrastructure bringing the benefits of virtualization within our organizational constraints

Handle the challenges virtualization brings – rapid deploy/rollback/expectations.

How you do this, whether it’s “From Monday at 10am physical servers become an exception.”, or you phase it in by production/non-production split is up to you. But the key is offering production from the start, you’d be surprised how many ‘applications’ there are that aren’t business critical, aren’t real time and where performance isn’t important, these are the key virtual targets, the key ones to prove IT can deliver with the new technology.

Once we’ve proved the provisioning model that we can build, deploy and support a virtual infrastructure we can then retrospectively start virtualizing platforms, looking at those systems we can be made virtual in the quickest most efficient way, all those legacy windows boxes. The key questions are in the next blog entry.

No related posts.

Related posts brought to you by Yet Another Related Posts Plugin.