Bladewatch.com

Many of the big banks/enterprises are acquiring more data center space, some are closing down their older facilities and having to move equipment. What things would I think about in terms of moving a volume of servers to a new physical data center?  Well as I write this on the train home, here they are:

• Consider naming conventions – for example that server called LONSapplication1, is it still called that if it’s in Milton Keynes? In Wales?
• Network latency – are the applications your validated to cope with any latency/network performance issues – real or perceived
• Network resolutions/addressing – does my ip address move? Do we need to scope out a new IP arrange for the new site?
• Storage – do we need any new storage/replication to the site?
• Infrastructure – what new infrastructure do we need? Another domain controller/SMS server, a new build server for the linux servers
• Hardware failures – what level of spares are to be kept during and after the move – what’s the agreement/SLA for fixing a faulty system on installation at the new site
• Hardware support contracts – it might be a remote site hosting blades for a grid application – but what’s the service level agreement, if the enclosure fails when will the vendor be on site to diagnose and repair the issues?
• Efficiency – do we want to move the legacy servers to a new data center? Is the objective virtualization/consolidation? Or is it to get everything more efficiently racked and managed in the new facility.
• Procedures – what are the procedures for granting access to this new data center?

No related posts.

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