Moving Data Centres and Servers: 3 methods
Posted by Angelo Bramani, on May 14 2014
So you’ve made the decision to move your data centre or servers to a new location. Although you might have thought making the final decision was tricky, the process has only just begun. There are a few different methods you can undertake when moving data centres and servers which I’ll aim to guide you through in this weeks post.
The expensive solution
For some organisations, budget and resources aren’t really an issue. If this is you, you have the opportunity to undertake the easiest method which would be to build a brand-new data centre in the new location that your organisation is moving to and then migrate everything across through virtual tools. In layman’s terms, this basically means you’re replacing the old data centre rather than moving it.
This method requires quite a large budget but you’ll enjoy very minimal downtime and a brand-new computing environment with new servers, storage, and core networking at the new site. You will also save a bunch of time since planning and scheduling considerations are much more flexible. It would be unfair if I didn’t acknowledge that building a new data centre is a lot more complicated than I just explained so consider that this method is not for the faint-hearted.
‘All in’ approach
This is the most cost efficient solution but it is also the hardest and longest process. The new data centre held in the new location is only provided with power and cooling whilst you move everything from point A to point B: racks, servers, storage, everything. For small relocations this might sound feasible but for large organisations, institutions and government relocations, it just isn’t possible.
The hybrid approach
This kind of approach calls for a ‘mix and match’ solution from the above two methods. Here, some part of the new data centre is built before the relocation, like racks and core networking. Once you’re ready to do the relocation, the old data centre is shut down and the servers and storage are physically moved to the new location. Finally, you can re-rack, re-cable, and bring them back online.
This solution comes at a much lower cost than the ‘replicate and replace’ approach but it also involves at least a day of downtime which opens the threat of data and service loss. As well, it’s usually a high-pressure process with critical services and applications down for hours at a time. If a problem arises, your downtime will grow as the new problem is sought and dealt with.
Planning is key
Most companies involved in moving data centres and servers will opt for some blend of the last two methods. The first expensive method is essentially guaranteed to succeed and offers plenty of time to get everything just right. But the way to ensure that the other approaches go smoothly is with copious amounts of planning.
No matter your approach, The First Choice Removalz team are well acquainted with IT and server relocations and we’re ready to start planning with you today. Get in touch!
For more of a discussion on this topic check out this post on PC World.