When planning a technical project complexity adds up very fast. Take a virtualization project for example; a lot more things than just the hypervisor installation are coming into play. You’ll need to assess a lot of needs and desires about SANs (snapshots, redundancy, replication, FC, iSCSI, FCoE), network (VLAN, 1/10 Gbps Ethernet, redundancy), disaster recovery/business continuity, hypervisors and there capabilities, management of it all and security. That is a lot of stakes and agendas to take into consideration. And then you haven’t even talked to the business managers, the application owners, and developers. Now, this isn’t limited to virtualization, but this is just a nice example of how so many stakes come together in one project.
One of the major mistakes, that is made again and again even up until this day in the second decade in the 21st century is the fact that entire important or even critical IT systems are being put into place with a plan that can be paraphrased as follows “We’ll just set it up and sort of see how it evolves and just wing it from there”. I have been forced to do this quite often. This creates many problems some of which I will address below.
The single worst problem is that you create a vacuum. That can be storage space, bandwidth, ample resources for a huge amount of virtual machines or a mixture of all this. The results however are always the same and is one of two possibilities. Either they really don’t want and need it so it will never be used. You can also achieve this by keeping it hidden so they can’t use it. The other option is the most natural one. In nature, there is a thing called a “horror vacui”. That means that a vacuum unless protected cannot exist, it has to be filled. Empty LUNSs with data, hypervisor hosts with guests, networks with bandwidth, and backup capacity with even more terabytes. You might think the second option is better than the first one as at least the infrastructure is getting used. Unfortunately, the reality is that this is creating a very expensive mess to run, support & troubleshoot. The legacy this creates is not a valuable inheritance but a bank-breaking, efficiency, and effectiveness ruining debt. Stop doing that right now, you are killing your business. You see technology debt is about more than just old hardware and software. It’s about what you build with it or what grows organically with it. Is that a fertile land that sustains the business or a cancer that is killing it?
The way to prevent this is planning done by competent, involved people with experience and context. No plan is perfect, but a plan gives you a framework to achieve the desired result. Even great people make mistakes but they have the skills and attitude to fix them or work around them.
What are some other problems? Wasting money. Take for example a completely oversized server farm. That thing will consume so much money over a three year period in energy and idle capacity that the amount would be sufficient to replace it with new right sized hardware (more bang for the buck, better energy efficiencies in three years) I don’t know about you but those are very disconcerting numbers.
You can also be wasting money and time. And those who know me I loath wasting time. What if the SAN solution you bought doesn’t perform as planned or isn’t the right fit? There goes 500.000 € or you find yourself in the CEO office explaining why you need an extra 400.000 € to get what is really needed. Oh-oh! Do you have money and time to do it all over again or will you be living with that expensive mistake until the current solution is end of life? Do you have to wait until the CFO and CEO have recovered enough from the shock to allow a new attempt? Or perhaps you bought a SAN solution that is enough to run NASA’s workload and you’ve invested 4.000.000 € in a rather expensive data room heater.
Getting a virtualization project wrong can wreak havoc on a business and create a sizable financial hemorrhage. You can say that that’s not your problem but I beg to differ. If the project goes south that means you’ll have to find another job. The IT world where I live is rather small so you might even have to switch to another field as you’ll be forever known as the guy that sunk company X with his little “plan”.
The reverse, being rewarded for your hard work and success is not a given. In the end, they pay you for getting the job done so results are expected, and to Joe Average manager all ICT is a PC with a software packet to install. So for all you eager beavers who think that with this kind of responsibility and risk management comes big reward when you get it right, I suggest you think again. I have witnessed quite the opposite personally. Even when you’re running multiple enterprise SAN’s, networks, infrastructures like SQL Server, Exchange clusters, Hyper-V clusters, geo clusters, load balancers and providing 2nd and 3rd line support for those and taking 24/7 responsibility for the environment the only thing some managers care about is why the PC they never ordered with the software they never ordered can’t be installed tomorrow. “What kind of a chicken shit outfit are you running here” is what they’ll think when you can’t do that. They’ve read the glossy brochure that IT is a commodity and they expect it cheap and always on, much like electricity. In the end some (incompetent) managers act like ungrateful psychopaths. They’ll just abuse you less when you get it right. Don’t expect anything else. Often it’s the ones that are not capable to integrate things they can’t do or don’t understand into their business. They can not value anything that’s beyond their comprehension so they’ll never recognize it. To them, people are, for all practical purposes, resources that are identical, “Full Time Equivalents”. So don’t buy into the hype that there is a skills shortage from that lot and they can’t fill job openings. The volume in which they often waste talent and flush motivation down the drain is shockingly high and indicates that there is no shortage at all or that they can’t recognize skills when they find it and they’ll hire anyone. Surely they didn’t make a mistake so it must be a skills shortage. So you still want to be some hotshot technical architect? Or does a job that only produces open opinions and optional advice on paper sound more attractive. Per hour worked you’ll earn more, run less risk, and have a lot less stress. My advice? Don’t switch fields if you enjoy what you’re doing, switch jobs. The best career advice I ever got was “don’t work with or for assholes”.
Well if you don’t agree with your bosses and you dare go against them you’re surely playing with your job, you could get fired! So? Does living in fear of being fired make good employees? Does not being strong and confident enough to tell your managers they are doing certain things totally wrong or that they are mistaken make for good advisors? The worst thing a boss can have are a bunch of “yes men” around him or her. That boss should be smarter than that. It doesn’t work. Having trust in the abilities and loyalty of your employees does not mean you need to agree on everything. As a boss you’ll make the final decisions, yes, but you’d better listen very carefully to your advisors and staff or you might as well have hired some monkeys. You can train them to say yes all the time, all it takes are some bananas. As an employee, don’t let yourself be treated like a monkey and if they fire you for throwing the banana back, good for you!
So you’d better love technology and building solutions because that means you are intrinsically motivated to go the extra miles. When you are, select a small group of people with the same attitude. You’ll be able to drag the devil himself out of hell with such a team at a very low cost. Whatever you, do don’t think you can externally motivate or coerce people into achieving this. Charles “Chargin’ Charlie” Beckwith knew that all along when he said “I’d rather go down the river with seven studs than with a hundred shitheads”. And guess what, he wasn’t taught this in some course, by getting a title or by being told this by a manager. He learned it himself by working with the best. These people will keep learning and growing on their own. They don’t need to be told what to do, how to train, what to use, they don’t need nannies & micromanagement. They need an end state and they’ll get it for you. Frankly, that kind of skillset and ability scares the shit out of some bosses as they micromanage actions & items instead of doing their jobs. You can’t use force, treats or authority to make people achievers. In the end, you can cut a diamond, but you cannot create it. Trust me. Putting that amount of pressure on someone that isn’t a diamond only turns them into a heap of crushed remains of what used to be a human being or FTE in your typical HR speak.
“Mate you’re not a conformist” my friend said … you’d better believe I’m not.