HyperConverged HyperHype
What is hyperconvergence?
Remember how back in the day every small business tried to find a way to cram the letter ‘x’ into their name to make it look ‘futuristic’?
Or when AOL and IRC screen names had more Xs, dashes and underscores than vowels?
Well, we are BACK! Luckily you don’t make business decisions based on silly product names so let’s keep on going.
I attended a presentation on HyperConvergence a few months ago, hosted internally by one of our Directors here at PEI. Admittedly I was completely unaware of what that word even meant and with hindsight there isn’t much to be missed. The idea of Hyperconverged Infrastructure, just like the word, sounds pretty cool at first but is a really basic concept.
You get 2 things. ONE management console for all your hardware. ONE ‘lego-block’ of compute+memory+storage+network, (get as many blocks as needed to satisfy your capacity needs).
That really is about it.
The whole point is to completely buy into a single provider of ALL your hardware and management plain. I bet you can sense how fond of the idea every risk-analyst is going to be. Oh and by the way, the reason we are mentioning this particular position you probably don’t have in your company, is because Hyperconvergence isn’t small potatoes. We are talking multinational enterprise scale here, everyone else can step aside.
There are more problems. Cost. HUGE cost. Performance – got 50terabytes of storage free but are short one NIC? Doesn’t matter. You have to buy the whole cube again. Wasted capacity: there will be a lot of it, it is part of the design.
Doesn’t sound that great does it? Well…let me tell you how little any of this actually matters.
Say you have 30,000 seats in your org, 2000 servers, 20 branch locations in different countries. How many sys-admins, network engineers, storage architects and virtualization admins do you need? Hundreds. You need hundreds of people with a VERY specific skill set and intimate familiarity with the way YOU run your architecture. Lost a specialized employee or two? Factor in a quarter for search and another quarter to train them before they can push a button on the keyboard and accomplish a task.
What if I told you…
All you need is a small army of identically qualified admins that are readily available and specialize in one thing, and one thing only – Hypercovergence.
Additionally, because the architecture and design is greatly simplified, you should be able to hire employees at a lesser skill level (Outsourcing much?). The whole point is to simplify management and capitalize on massive gains in efficiency achieved by a completely unified hardware infrastructure.
How do I get there?
You don’t. That’s really the biggest barrier in getting Hyperconvergence off the ground. Let me ask you this, what do you think is going to happen when you as a CTO of a large enterprise were to announce ‘we are going to throw away all of our equipment and fire all the high-level staff and replace it with a big black box and a small team of mid-leve sysadmins’ ?
Conversely, any highly-skilled employee with any amount of self-preservation instinct is going to resist this as much as possible.
But hey, that’s okay, they are already doing the same exact thing fighting the cloud.
So how do I really get there?
I swear, there is a great consulting business idea somewhere. Our phone number is below btw.
Jacob R, PEI
