So you’ve got your scalable data center. You’ve got your favorite cloud platform at your fingertips. You’ve got a perfect asset inventory of every piece of system hardware and software as well as the dependencies between.
Well, year one you can sit tight and bank on the existing technology to move your company forward.
Well, year 5 you have reached the capacity of the first phase of your local data center construction. At the same time you find a great deal at a COLO facility that perfectly meets your data center needs. You still want to keep about 60% of your servers and SAN at the local DC but you also move about 40% of your infrastructure into the cloud. In order to make this move seamlessly you rapidly create a listing of exactly which servers can be moved. Since you know the dependencies of your server structure you can quickly filter which servers can be migrated versus which cannot.
For example, let’s say x server queries y SQL table. Z server also queries y SQL table. But you want to move z to the cloud and keep x in the local data center. Since you have a good inventory of your dependencies you can make the choice of whether to stay or to go. This extends to every piece of hardware.
Now this is far from easy. In fact, for most companies it’s nearly impossible – at least it’s impossible with legacy equipment. It’s just too difficult to track down dependencies. If your existing infrastructure is highly virtualized it will be easier, but your dev team with hate you for a while.
Alternatively, you can freeze your legacy equipment in place in the local data center. Then institute a comprehensive change management system so that every new piece of hardware and software is accounted for before it is even installed. Moving forward, migration of equipment that has been installed after the change management system has been instituted will migrate with relative ease.
If all of these steps are followed then migration readiness becomes a much simpler and much more basic matter. This will greatly decrease your headaches and anxiety, and make your infrastructure much more reliable when migration actually takes place.