Your Data Center’s Moving Day: Planning for Minimal Business Disruption

Is your business moving its headquarters or branch offices – and its data center with it? As IT leadership, you should have been given sufficient notice – six months or more – to carefully inventory your current server racks, networking gear, applications and power and connectivity needs. You also have an understanding about your apps’ dependencies and service-level requirements, meaning you know how long each of your systems can tolerate downtime. Using this knowledge, you can carefully plan a staged move, in order to minimize business disruption or avoid it altogether.

Moving Basics for IT

Any good moving plan starts with a thorough inventory of current IT assets, their connections, power sources, and their uses. DCIM (Data Center Infrastructure Management) software can be a wise investment. Such a tool tracks assets, their network connections and capacities, their power, space, and cooling requirements.  You’ll also need a camera; before the move, shoot every device on every rack, front and rear, and label every photo.

A thorough inventory may turn up many servers with very low or even zero usage rates. This presents a good opportunity for safely scrapping dead iron and freeing up rack space. Several servers with low usage are ripe for virtualizing and consolidating into one physical machine.

In addition to upper management, IT should have some say at the very beginning of the moving process, when picking the new location. Especially if you’re an online business where little to no downtime is acceptable, it’s critical to ensure the new building has backup power and similarly, multiple carriers to support failover.

When it comes to physically moving your equipment, you might want to consider hiring extra manpower, experienced in moving technology. Your IT staff will be too busy overseeing reconnections and reboots to be dealing with boxes and trucks.

Lay In and Test the New Network First

Hire a contractor to verify the labeling for every wall socket and patch panel, to test the new cabling end to end, and provide you with a report on the results. The most common IT relocation horror story seems to prominently feature delayed data line delivery or misconfigured networking equipment.

“The process of getting connectivity into the new location should start as soon as the ink is dry on the lease or purchase agreement,” says Dave Hopson, managing partner of Triumphus, a Houston-area IT consultancy. “Often there are long lead times.”  Coordinating the move of Internet service providers, static IP addresses and voice services generally takes 90 days to ensure a smooth transition, notes Travis Grundke, operations manager at managed IT services provider Ashton Technology Solutions.

Sanjiv Bawa, CEO and founder of Chicago-based cloud provider Chi Networks, echoes many when he says that new core routing, switching and network lines all have to be ready, tested and waiting in the new location before moving any services. And for applications that cannot be interrupted – like previously noted e-commerce servers and manufacturing systems – he replicates and tests the relevant data center components completely. “I set up an entire disaster recovery system in the new facility,” he says, “and then fail it over.” He also recommends leasing a dedicated, point-to-point circuit between old and new locations to make this move, system by system and subnet by subnet.

Applications that can be down for a Saturday or a whole weekend are easier to cut over without business disruption, assuming the physical move is a short distance. But the same rules of planning apply.

Move Up, Not Over

An IT move is also a prime opportunity to make the leap into cloud migration. This could simply mean virtualizing and updating your servers for the new in-house data center: “If you have 20 old boxes you can consolidate down to four,” says Bawa – that’s a great savings in space and moving costs. “Maybe some apps that you want to replicate are using old hardware,” he continues, recommending a refresh to faster, more elastic computing capacity and newer operating systems.

Or you may want to virtualize and replicate your servers to a third-party, off-site cloud. “Many customers are saying, ‘Why do I even need a data center?’” says Bawa. An off-site cloud migration trades capital expenses for operating expenses. It also gives you the time to deploy and test with all your end users from your old location, as well as backup in the old data center that you’ve left in place.

An IT move presents a great opportunity to assess your project load and processes in a holistic way. Can you keep up with day-to-day maintenance and management while also developing new solutions to support strategic business goals? IT is being asked to create more and manage less. Third-party expertise combined with virtualized (and possibly hybrid) cloud environments can provide some of the balance you need to accomplish both.

Finally, recognize that you have another project to manage in decommissioning your old data center. This must be done not only in a way that permanently and safely destroys the data that’s been replicated in the new site, but recycles, recovers, and disposes of electronic devices in accordance with current environmental standards. As with relocation, there are companies that perform these services for enterprises of all sizes.

Through proper preparation and clear understanding of current IT assets, moving your data center can be a seamless process without causing your business detrimental downtime.