There are multiple reasons why companies decide to move from company-owned data centers to colocation data centers. Regardless of the reason, after deciding to “go colo,” the question, “how do we relocate all of our equipment with minimal downtime for our business?” may come to mind. The answer can seem daunting because relocating IT equipment can be complex and overwhelming.
Successfully relocating equipment is not something that can be rushed into; it is a deliberate and methodical process where the “devil really is in the details.” A lack of attention in one area can cause rippling effects across the project. If you’re planning to move to a colocation data center, follow these five best practices to ensure a smooth transition.
Instinctively, most people would list careful planning as a key best practice when migrating a data center to a colocation site. Careful planning starts by clearly identifying the migration’s objective and defining all key stakeholder roles and responsibilities. The next step is to take inventory of all IT equipment that will either move or will be impacted by the move. The inventory should include names, make/models, serial numbers, cabinets, U locations and associated databases, applications, dependencies, owners, users and information about who supports the equipment. Without a doubt, the inventory effort is arduous, but it’s worth the effort because the process reveals the impact of shutting down equipment, which avoids unexpected outages during the move. It also helps determine optimal dates and times for the move.
The next key best practice following the move planning is move placement, meaning where the equipment should reside at the colo. The colo should participate in this decision to ensure power and cooling loads remain within cabinet capacity limits. Once the equipment’s new location is identified, a move spreadsheet, drawings of the fronts and backs of cabinets and a floor plan should be created. This information will provide clear direction of where equipment currently resides and where it should reside within the colo for the installers. Labeling each piece of equipment is another vital part of move preparation. Labels must clearly identify both the FROM and TO locations in terms of cabinet and unit “U” space.
Another important aspect of move preparation is creating the “seed environment” at the colo. The seed environment means pre-installation of cabinets, power circuits (whips), power strips, power cords, network gear, and data cables. Duplicate assets and supplies need to be purchased, but the extra cost far outweighs the risk of extended outages if unexpected issues arise during the migration such as power plugs that don’t match receptacles, or cables that are the wrong length.
Packing and Shipping
Proper packing and shipping is also a critical part of move preparation. Un-racked equipment should be placed in original boxes and packing material, or placed in specialized crates that are lined with antistatic foam on all sides to prevent damage during transition. Most IT equipment should not be transported in server cabinets unless authorized by a specific equipment and cabinet manufacturer. Equipment moved within cabinets becomes vulnerable to vibration when rolled across uneven floors and when hitting bumps while on the road. Only reputable shippers who specialize in transporting valuable IT equipment should be used. Check to make sure your shipper is available during non-business hours, uses trucks with air-ride suspension, has proper insurance, and can properly secure equipment loaded in the truck.
A final best practice in successfully migrating a datacenter is validating that equipment is truly shut down prior to moving it and validating that it is fully operational once powered up again at the colo. A systems administrator typically shuts down a server remotely rather than being physically at the machine. Once the server is shut down, the administrator gives the okay to unplug and un-rack it. However, green ports may be discovered when technicians attempt to un-rack it. This may be due to equipment being mislabeled and the wrong server being identified. Or, it may be the correct server, but it wasn’t completely shut down yet. Paying close attention to the equipment’s buttons and ports to ensure equipment is truly down is an ironclad way to prevent accidental outages. Once the equipment is racked, connected, and powered up at the colo, testing the equipment for proper operation must be performed. Technicians equipped with cable testers must be prepared to troubleshoot cables if connectivity issues arise. The migration should be considered finished only when all equipment passes validation tests.
Companies decide to move out of their own corporate data centers into colocation data centers for a variety of reasons. Careful planning, proper migration practices and thorough validation are key best practices to follow. Patiently and methodically following these best practices will resolve the dilemma of how to move the equipment with minimal downtime.
Implemented in combination, these best practices can ensure an on-time and error-free move to the colo, reduce disruption and reassure stakeholders that the decision to move to a colo was a wise choice.
Nathaniel Josephs is Senior Manager, Data Center Operations for Parallel Technologies.
Currently rated by 0 people