When you’re moving a business, it’s important not to overlook your digital infrastructure. If something goes wrong when moving servers to a new location, delays or damage could sideline your company.
Earlier this week we sat down with IT guru Mark Henderson Leary, vice president of Aldridge, a Houston-based managed IT services company, to discuss the unique challenges with moving a server room.
Companies looking to relocate are often faced with specific concerns when it comes to moving servers to a new location. What’s the first step?
Mark: Moving servers to a new location is a very a complex procedure because moving a company and its digital infrastructure poses a unique set of obstacles. The way I see it, there’s really two facets to the move: the physical move and the information move.
I really can’t underscore enough that you need to have a plan. With so many moving pieces, something will invariably not work as it should and when that happens, you need to know what you’re going to do in order to minimize down time.
With that in mind, the first question that needs to be asked is “How much down time can you afford?” If your business is able to shut down the servers on Friday afternoon and be offline until Monday morning, that’s ideal. With a full weekend, you should have sufficient time to move anything that doesn’t require a large amount of time to set up or install. That way you have the luxury of being able to reuse all major pieces of equipment.
However, for those companies that can’t afford to be offline for an extended length of time, the question shifts to “How much infrastructure am I willing to duplicate in the name of a smooth transition?”
Some systems can’t be moved because that means they’re off and if they’re off, the company isn’t doing business and that might not be an option.
Would you recommend companies use this transition to move their infrastructure to the cloud?
Mark: When you talk about the cloud, things can get tricky, because what does that even mean? There are several different levels of cloud dependency. Does that mean you have your email in the cloud? Does it mean you have your entire infrastructure in the cloud? Does it mean that you have your phone system in the cloud, in addition to your files, applications, emails, and everything else?
Answering these questions is critical because even if you’re 100 percent in the cloud, there’s still physical stuff in your office. There’s a computer and phone on your desk and some hardware in a wiring closet. That stuff has to be addressed, but it’s really simple. If you’re down to the bare necessities and the rest of your infrastructure is in the cloud, then you’re probably pretty close to being ready to go. You’d be able to set up the equipment in the new office and have it waiting for you when you get there. Everybody could literally walk in and plug in their laptop and phone and it’s off to the races.
If you’re a company that’s planning a move and considering upgrading your infrastructure to be more cloud based, that’s a great idea, but that conversation needs to start anywhere from three to six months out. Three months is about as tight as I would want that.
However, you can get cloud backup running in 30 days. Cloud email is a similar deal, but if they’re talking about the core of their infrastructure to the cloud then we need to have that conversation three to six months in advance and talk about what that plan looks like. Ultimately, the cloud dramatically simplifies a move.
Turning back to the physical move now, do you have any advice for businesses that are looking to move physical servers?
Mark: The simplest advice is often the best: only move stuff that has to be moved. It’s a cost-benefit analysis; if it didn’t cost anything you could just set up everything brand new, test it in advance, and then just flip a switch during the move.
Unfortunately, most of us don’t have unlimited funds at our disposal, so we have to move really valuable equipment. This is why it’s important to pick the right movers. As you’re planning for the move, you want to make sure things are on schedule and on budget, with minimal downtime.
The main thing is experience in moving a server room or moving high-value infrastructure. You can’t simply unplug servers, throw them in the back of a truck and reconnect them in a new location. You need to make sure you work with somebody who understands the ramification of moving such sensitive equipment, someone who is bonded and insured properly for that type of stuff.
These sound like obvious points, but I’ve seen what happened when an off-brand moving company moved a company’s servers and over $350,000 worth of equipment literally fell off the back of the truck. It was a disaster.
What do you look for in a company for this sort of move?
Mark: Proper transportation, knowledge, and skill are key. Certain precautions should be taken to transport electronics like special packing and a controlled transit to keep the equipment safe. But it’s a sliding scale. I would say that if you’re a small business you do what you can afford. If you’re a big business you better have somebody with some credentials who can demonstrate they’re good at moving servers to a new location.
Do you have any experience with a move that’s gone wrong?
Mark: Yes. In fact, here’s a nightmare story. We had a client who didn’t have any organization on their side of the move. They had an office administrator in charge of moving their equipment and they were moving their entire office across town. For some reason they thought it made sense to move over the course of a week. Everything was done piecemeal, which is pretty abnormal.
They hired some moving company that misunderstood every direction possible. We had labeled the equipment and told them not to take anything out of these racks because everything that needs to be removed has been removed. When we got to the new site everything had been taken out of the racks and stacked all around the room. Nothing was in the spot where it was supposed to be. We had to spend four or five days trying to get it back up. They were down for almost a week, not doing business, and paying us the full billable rate while we pieced it back together. They ended up with more than $14,000 in overage bills and this was a very small company.
It just goes to show that if someone on the client side doesn’t know exactly what they’re looking at and can’t manage the relocation of a box from one site to another, or tell the movers exactly where things need to go, there will be trouble. There needs to be someone that can step in and do that. It could be a vendor of some kind, as long as it’s someone who understands the ramifications and can really project manage that move down to the last cable.
How do companies that are moving a server room avoid such a disastrous experience?
Mark: It’s a common scenario to shut a server down with no issues, put it in a car, gently drive across town, put it back in the rack, and it just won’t come on. That happens.
I would say the majority of the moves I’ve been a part of have gone fine, but in that same majority there’s always one little thing that is unexpected; one hard drive didn’t want to spin up and one network card died, whatever the case may be. What’s important is being ready for that inevitability and making sure there’s a contingency plan built into the move.
If you sit down with the moving company and they can’t explain to you what their process is, or if they don’t have a technical team that understands, that’s a big red flag. Your IT folks need to feel very comfortable with what they’re being told, and what the timing is going to be like. There should be some specific requirements. If you’re a moving company and you tell me that you’re going to savvy enough to move 20 different servers, I expect you to ask me some questions that I might not be able to answer at first. That’s important.
So when something goes wrong, if you lose a server, now what? How do I get it back? Say the server stays offline and it’s dead and it’s not coming back. I’ve got that backup, but what do I do with it? Do we need to test it? Do we need to be sure I can restore that data? How many hours will it take to restore that data? If the server dies Sunday night when we’re putting it online and Monday morning we’re supposed to do X, but it’s going to take two to three days to restore that server, what’s that going to mean?
To have a clean move, you have to know the answer to these questions before anything gets moved.
About Mark Henderson Leary
Mark is vice president of Aldridge, which specializes in providing managed IT and cloud computing solutions for small to midsize businesses. Founded in Houston in 1984, Aldridge provides a broad range of outsourced custom IT services for clients across the nation.