I’ve had some thoughts rattling around in my head for a while and I wanted to take a minute to write them down to share with others as well as to help myself add some order to these thoughts. My career these days is pretty perfect, I get to create architecture and designs, implement them, then move on. I don’t have to deal with the internal organizational politics, sell ideas to my management, etc. I just get to participate when it’s time for you to do the “fun” projects! It wasn’t long ago though that those less savory items were my life working as an IT Infrastructure manager in a fairly large organization…4000+ users and a few hundred servers. I’m saying this to try to convey some level of empathy for those of you working with organizational politics that might think I’m just an ignorant consultant.
There is a ton of waste in IT environments. I can’t tell you the last time I found an organization that by design was fully utilizing their IT infrastructure. Most VMware environments are significantly underutilized. I see org after org adding servers to their VMware farm when they don’t need to. That’s great for the VAR I work for but it’s partially the reason people are looking at IAAS instead of internal infrastructure. I have 2 theories on why people add infrastructure when they don’t need to and this applies to more than just VMware farms, you could apply this to almost all IT infrastructure.
First theory, IT overbuys infrastructure because it’s hard to “go back to the well” for more money later, they’ve been conditioned to buy more than what they need because they can’t easily buy infrastructure later so when a big project comes along they buy as much as they can whether they need it or not knowing there will be future needs that won’t get approved that this one big purchase will make up for. I know this happens, I hear it constantly from IT groups we work with, I did it myself…heck I even went so far as to keep a few dozen servers under maintenance when they were no longer needed just so when the CIO wanted to cut costs I could cut a few servers from their maintenance contract and “save money”. Wouldn’t it have been better for the company if I had cut those servers immediately from their maintenance contract? Sure! But my reality of cutting costs wasn’t based in reality or what was best for the business, it was just a number I had to cut regardless of the impact to the business. So to make up for it I made sure there was always lots of waste. Sound a bit like government?
Second theory, your IT department is technically clueless. So back to that example I used of organizations buying more servers for their VMware farm than they need? Pure lack of understanding. Maybe they had a performance issue and didn’t know how to troubleshoot it and decided in the future they’ll make sure they’re not exposed like that again and just buy more hardware than they need. Maybe they really don’t know how any of this virtualization works and are just nervous that it actually works…and since they lack any tools to measure or simulate workloads they don’t have any way to argue against an end user that says things are slower now that they are virtualized. It’s like taking your car to a repair shop that doesn’t know how your car works, they just start replacing everything and when in doubt bigger is better. Every management book I’ve ever read has said to hire people smarter than you and that people are the difference in an organization. Yet every day I meet organizations that knowingly hire people dumber than them, keep people in key positions they know in their gut aren’t right for the business. Why? Do you think these people don’t affect your organization negatively? I doubt it…I’d bet on it that you’re wrong. Even in the organization I work at I think the cost a bad employee is vastly underestimated and understood. Smart people are the difference…always. I’ll hire smart over tenure (or any other metric) every day…it is the only thing that matters.
So why do I bring all this up? Because your CEO is about to try to replace you with “Cloud” and you need to understand why. I’m not saying IAAS isn’t the right solution for you I’m just saying that in order to have internal infrastructure in the future you need to start acting a little more like IAAS providers internally because those providers hire talent and don’t overbuy. More on that in a future post.