I am a huge proponent of designing your own servers. I firmly believe there is no better way to save your organization time and money than to design, build, and maintain your own hardware. Every server and desktop here is custom designed and built by me for a specific purpose. My users and colleagues will tell you there are very few times where an application or device has been offline and created a problem. There’s no way you can avoid downtime, but you can avoid how badly it cripples your business and its ability to serve its customers (remember, your users are your customers too). You first need to identify how much downtime is acceptable (zero is not an option here) and work around it. More on that later.

I can hear you saying now “Well what about technical support and part availability? How can I ever build something better than Dell or HP?” You can, easily. Armed with an understanding of what function certain components perform and what the demands of your application are, I firmly believe you can cut the cost of your infrastructure very significantly, and I have the data to back it up.

I was recently tasked with a project that required me to create a very large amount of storage space for our Marketing team. Now, when I say large, we’re talking 30TB of useable space. Part of the project also required me to design a way to access that data more efficiently, because let’s be honest, even at gigabit speeds, copying terabytes of data is going to take a while. Storage providers were coming in with prices starting at $25,000 and topping out around $40,000, and that didn’t fully address my requirement of faster network access. It turns out that many manufacturers aren’t supporting 10GB/s yet and the ones that did support it were installed in systems that were much larger than we required. I could get 10GB/s if I wanted to go with flash-based storage (and I would have loved to do that, but I didn’t dare take that request to the CFO).

So, what to do now? The quotes from the storage vendors were way over my budget and I had a project to complete. Fortunately, the answer was easy; I’d build my own. I ended up with an array that created an extra 51TB of storage, was accessible at 10GB/s speeds, and came in under my $15,000 budget. Installation and configuration was a breeze, and my users still don’t know that anything happened, nor do they care. All they care about is that it works and that they have enough space to store what they need to store. Building my own equipment let me pick and choose what components I wanted to use while at the same time not being forced to use parts I didn’t want to use.

Mission accomplished.

A future blog post will delve into more detail as to what you need to look for when building your own machines, so keep checking back.

In the meantime, if you have no idea on how to start taking control of your data center, contact us and we’ll point you in the right direction.

Leave a Reply

Your email address will not be published. Required fields are marked *

 characters available