Reducing Downtime During Data Center Relocation

One key metric central to any data center relocation event is downtime. Obviously, there will be some degree of operational downtime, but how much downtime is considered acceptable? Are there data center relocation scorecards that can be used to judge a move event? And better yet, how can you reduce the amount of downtime during your next move?

Transitional Data Services employs a variety of strategies, best practices and tools to ensure minimal downtime during its data center relocation projects. Based on our extensive experience relocating over 1,000,000 square feet of data centers, here are some tips that are sure to help you prepare to move your own data center. .

Move Readiness Assessments - These are generally low cost engagements, sometimes even offered in the form of a half-day consultation. The goal of a move readiness assessment is to review your operational plans, resource plans, communication plans, logistics, calendars and issue resolution plans and then identify gaps and/or areas for improvement. If your organization is considering managing a relocation event on its own, these low cost services can help reduce downtime and highlight obvious “gotcha’s” that may otherwise cause headaches.

No New Equipment or S/W Upgrades - Contrary to what some OEM’s may tell you, a data center relocation event is not the right time to upgrade your most critical applications and server hardware. There’s already enough risk involved in data center relocation without adding the complexities of an upgrade or replacement to the mix. Virtualization and server consolidation is different, and should be carefully considered as a strategy worth pursuing.

Application Dependency and Move Bundling - Although not usually an issue with smaller move events, any large relocation event absolutely requires a detailed analysis of application dependency and move bundling. TDS uses a variety of sophisticated tools to analyze and plan complex move events, tools that are usually available for purchase if your organization wishes to perform this analysis internally.

Application Performance and Network Testing - Unlike application dependency and move bundling, application performance testing and network testing are increasingly becoming a standard practice on moves of any size. For example, moving an application from the LAN to the WAN, only to discover the application was never developed for the latency and speed provided by the WAN is all too common. TDS uses state-of-the-art simulation tools and a host of other services to minimize risk, avoid move day disasters and to properly size needed network bandwidth.

Inventory Confirmation -In 10 years of data center consulting we rarely find an organization whose physical and application inventory accuracy was greater than 50 to 75%. Adding an update and verification of your asset inventory is vital to project success. Move day is not the time to add 20 overlooked pieces of equipment to the relocation puzzle. Find out how TDS collects physical inventory using barcodes and its own purpose-built software called Transition Manager.

Site Inspection / Walkthrough -Loading dock accessibility, truck restrictions, elevator & door heights and weight restrictions are just some of the items to be verified before your move event begins. And of course – this step must be performed at both the source and target data centers.

Test Moves - Every Broadway Musical has at least one dress rehearsal before opening night, and your data center relocation is at least as complex. A “dress rehearsal” for your move event is not difficult to implement, and you’d be surprised how many relocation projects fail because of the simplest of oversights. For example, do you know if the UPS guy sorts all of his pick-ups on your loading dock at 6:00PM on Friday nights because no one’s ever there? Are you SURE that oversize rack you’ve had for years will REALLY fit through all those doors? Testing move events with a small selection of non-production or isolated, low-priority equipment, well in advance of your actual move event, is a great opportunity to work out the kinks and test your end-to-end processes and procedures from end-to-end.

Cabling - One of the most common “gotcha’s” is the one in which customers believe cabling should be run on the actual move day. The answer is no. Under no circumstances. While additional planning may be required to determine proper cable length from a device in Rack 3, U position 22, back to the core, move day is not the correct time and place to do it. Don’t wait until move weekend to run your cabling. It almost certainly will increase your downtime and may even cost you more, while potentially causing significant outages.

Un-racking and re-racking equipment - Most racks are not manufactured so they can be moved around while filled with equipment. In fact, most rack manufacturers recommend against it. And, believe it or not – un-racking equipment doesn’t add much time to the process, allows more flexibility in placing servers in their new location, and results in a better long-term solution. Learn how TDS un-racks and re-racks using barcode labels.

New Racks and Pre-Staged Rails - Especially on larger move events, ordering new racks is a “must” and saves hours of move day downtime. It allows network equipment to be pre-staged and pre-cabled. If you have to prioritize, choose new racks over new rails, but updating both is ideal.

Choosing the right logistics partner - It may sound obvious, but choosing the right logistics partner, one with extensive experience relocating high value equipment is critical. Move day is not the right time to find out that your equipment will be delivered right after the trucker unloads a couple of tons of used auto parts.

Staffing - Like all successful projects, the quality of your move event comes down to the skill and expertise of the people who develop and execute the plan. But unlike IT equipment, people do not come with 99.999% uptime! How will you cope when one of your employees has to make an unexpected hospital visit? On how many other projects are your employees working in addition to the data center relocation? What happens if one of your Sys Admins gets stuck troubleshooting a critical application problem for 4 hours on the morning of your move? Don’t forget to include back-up resources in your risk mitigation plans