A few key tips on successful data center consolidation
Agilex's Bob Otto offers advise on how to clear the potential hurdles of time, budget and culture
Data center consolidation is a huge undertaking that will take years for federal agencies to successfully complete. Time, budget and culture are road blocks that must be overcome if the government is going to reduce the 2,094 data centers spread across agencies. The government hopes that stemming the growth of data centers will reduce energy use, spur IT cost decreases and improve security. As CIO and chief technology officer of the U.S. Postal Service, Bob Otto helped reduce 20 data centers down to two. Now the executive vice president for IT Advisory Services at Agilex, an IT professional services company, Otto is advising agencies on consolidation efforts. He recently gave his thoughts on what it takes for agencies to be successful through an e-mail exchange.
- By Rutrell Yasin
- Nov 19, 2010
GCN: Why is data center consolidation so challenging?
Otto: It is important to recognize that most data center consolidation projects encompass two phases. During the initial period, agencies can achieve "quick wins" by quickly closing and consolidating remote sites, smaller server farms and secondary data centers, which are often little more than multiple servers in a small facility. The second phase is often more challenging as you consolidate, automate and modernize major facilities.
By the middle of next year, I would expect that all [federal] agencies will have a very detailed plan in place for execution. And that several agencies would have made significant progress in consolidating the remote smaller sites and secondary data centers. In reality, the biggest driver here should be information security and continuity of operations, as these facilities are often a huge liability. And while the number of data centers may quickly drop, keep in mind that the majority of the work for most agencies will still be ahead of them over the next few years.
How can agencies achieve success?
Data center consolidation is a multiyear project encompassing significant change management. To do this correctly and achieve the full benefits, agencies will need to put in place leadership, planning, asset management, project tracking, measurements, assessments and evaluations, governance and training. Coupled with the fact that this change impacts active systems, you need to proceed with caution to ensure that you don’t jeopardize current operations while you are building the future.
Can you talk about the importance of asset management as a starting point?
As a starting point, effective asset management is the foundation for successful data center consolidation. By documenting your existing systems and processes, you can identify areas of duplication. And having established a baseline and a detailed plan, we should expect that [agencies’] confidence in reaching their goal will grow.
IT leaders recognize three potential hurdles in implementing the Federal Data Center Consolidation Initiative – time, budget and culture – that are driving their hesitation. First, this is both a major engineering and change management project so time will be needed to get it right. And while the long-term goal is cost-savings, little upfront money has been set aside. To reduce sustainment costs, we are going to need to invest in more agile and automated infrastructure. Finally, potential inertia from the rank and file and lack of prioritization and focus from management are additional risks. Overcoming all of these obstacles is critical to achieving success with this initiative.
Myth buster: Is data center consolidation a fast path to cost savings?
Are there creative ways that agencies can get the money to move toward consolidation?
While the cost issue might seem insurmountable, there are ways that it can be addressed. Obviously, this is an area where cloud computing can play a huge role. However, I think it’s even more important to commit to a phased approach in which incremental cost savings are used to fund subsequent investments.
There are a number of areas where I typically look to realize these savings. For example, use baseline budgets to reevaluate requirements annually versus simply bumping costs to keep pace with inflation. Going a step further, I used to challenge my managers to actually lower their costs each year.
Due to the risk of complacency, you should also regularly review your existing contracts to ensure that you’re getting optimal value. In some cases, you might seek to extend current relationships in exchange for additional savings. Likewise, the refresh cycles on many technologies can be extended as it has become more mature.
We should also look to consolidate our other cost centers to achieve additional efficiencies. Customer support and desktop maintenance is one area often ripe for savings. Another is the tools and applications used throughout the organization. Incentivizing and/or penalizing users can often drive them to shift from non-standard systems to more cost-effective solutions.
What are some of the other obstacles/roadblocks to data center consolidation as you see it?
In terms of change management, you need to secure the full buy-in of your staff and contractors while ensuring that this initiative remains a top management objective for the agency.
Obviously, the implications for your existing workforce are huge. It’s important to take the time to get their perspective on how the program can be implemented most effectively. You also need to provide a clear roadmap for how their roles will change and how they will be supported in this transition.
From a management perspective, it is important to recognize that this is a multiyear program spanning multiple administrations and CIOs. As such, it must be embraced as a strategic agency program and not simply as one person’s initiative. It’s also important that your best people and best thinking be devoted to this from day one. Nothing else will have as broad and lasting an impact.
Ultimately, you are going to need to redefine what technology means within your organization, shifting it from an IT-maintained product to a customer-driven service. We need to embrace a portfolio mindset in which resources are continually realigned with the areas in which they are having the biggest impact.
At the same time, we need to ensure that "server huggers" within IT, as well as arguments regarding "unique missions" aren’t used to retain legacy architecture and systems. Part of the cultural change that we need to implement is about ensuring shared ownership of the initiative and its success in order to overcome these obstacles.
Rutrell Yasin is is a freelance technology writer for GCN.