Disruption or failure - The painful choices of migrating to the cloud

I managed to get along to the Cloud World Forum in London this week. It was a busy, fast and noisy event (sometimes very noisy) which provided a way to sample a lot of views in a short space of time. The fundamentals of cloud computing are well established so there were no paradigm shaking revelations (and this wasn’t the kind of premier global event where such things might get announced anyway). I did pick up some subtle signals which might be of interest to organisations wanting to move to the cloud, suppliers competing for cloud business and engineers working with cloud technologies. This first post focuses on the customer perspective.

The painful truth about cloud migration

For customers wanting to make the move to cloud computing a number of distinct strategies seem to be appearing. A very common strategy - allow me to call this Filling the Attic - is the same strategy organisations have used for earlier waves of technology. This involves refreshing the long term target architecture to include cloud computing (often with a “cloud-first” or “cloud by default” policy) and then letting individual business units and change projects implement the target piece by piece. Over a number of years the changes accumulate and the organisation moves towards the new cloud computing architecture. But they will never quite reach the destination - stubborn pockets of legacy architectures will remain and can’t be thrown away. If executed well this is actually a nearly optimal way to run IT but most organisations lack the discipline and governance necessary to keep this technical sprawl under control.

A more proactive approach is Self Disruption where an organisation creates a new business operation and, like a startup, builds a clean cloud computing architecture from scratch. Gradually, business is diverted to the new operation and the legacy business is strangled along with its legacy IT. This is a bold approach which should probably be viewed as a digital business transformation rather than an IT initiative. The UK Government is doing something like this with GDS.

Technical Bankruptcy

I suspect many organisations will arrive at their cloud destination by Failing Forward. Effectively an IT or business shock creates the opportunity to leap onto a new cloud computing architecture. The usual burden of trying to maintain business or service continuity can be ignored as the business or service has already failed. This is effectively an IT version of bankruptcy. No organisation would be wise to deliberately chose this option as many will not survive the shock but some organisations (e.g. financial services organisations which might need to spend billions to escape from their mainframes) may have run out of alternatives.

Unfortunately, that is the end of the list of options. At the conference there was some talk of legacy migration but, in practice, applications are either built for the cloud or are incompatible with the cloud. At the moment application migration tools are, at best, short cuts to application replacement. Keep looking if you want but don’t say I didn’t warn you.

In my next post I will look at the situation from the perspective of the suppliers. Believe me, their options aren’t any better! What patterns are you seeing in the cloud computing market? Add a comment or use the Twitter button to let me know what you think.

** Related post**: DON’T BOTHER migrating legacy apps to the cloud, says CTO