It certainly does if you're in Enterprise IT.
I seem to be spending quite a lot of my time at the moment talking to customers about why it is so important to ensure that their IT infrastructures are “right sized”. We should call this the Goldilocks approach - not too much, not too little, but just right!As IT professionals we are taught to over provision from the day we leave the womb. It is always better to have too much than to have too little - right? The industry is designed to leverage this mindset and infrastructures are often just too big as a consequence. After all the “tin” is just a commodity and it’s the applications that are important. Although I understand this way of thinking I personally think the truth is a little more complicated than this.
The size of an infrastructure does matter, and it matters a lot! See, I knew it! They were just trying to spare my feelings! Building big does safeguard application performance, but burdens the business with so much unnecessary cost. Why? Because the “tin” is the basis on which everything else is built.
A typical enterprise data centre with virtualisation operates at an efficiency of around 25% - that’s 75% of the physical resources deployed not being used. If that were the end of it then I would agree that we don’t have a massive problem. We still have a problem, it’s just not a massive one. In today’s data centre that isn’t where it ends - we still have virtualisation, operating systems, and increasingly both Software Defined Networking (SDN) and Software Defined Storage (SDS) to consider.
Key point: All of these are priced in direct relationship to the size of the “tin” and they all (typically) cost significantly more than the “tin”.
So by definition, in the same data centre, we have 75% waste in an increasing number of higher cost items. Wow, in fact double wow!! It’s like compound interest on steroids - and not in a good way.
This is such a massive waste and it is the reason why I get all “preachy” when talking about the “right sizing” of enterprise infrastructure. I understand the reasons why things happen the way they do but I just happen to believe that there is a better way. In fact I happen to know that there is a better way - and it isn’t that complicated. Automating efficiency levels within the virtualised data centre and moving resource utilisation more towards 70% has the potential to cut IT infrastructure costs in half and still leave headroom for busy periods. If you want to know more and you want the recipe to the secret sauce then drop me a line - I’d be happy to share the wealth.
Watch out for those 3 bears!