Virtualization Meets Its Organizational Limit
We've been staring a simple truth in the face for a few years but generally trying to ignore it because the implications are big. That truth: If we are to get the most out of highly virtualized, cloud-ready environments we are probably going to have to rethink the way parts of IT are organized.
Mon, November 22, 2010
Network World — We've been staring a simple truth in the face for a few years but generally trying to ignore it because the implications are big. That truth: If we are to get the most out of highly virtualized, cloud-ready environments we are probably going to have to rethink the way parts of IT are organized.
Virtualized environments, after all, blend compute, storage, network and even application resources. While you can manage that using traditional infrastructure teams organized by technology stack -- and most seem to be doing just that -- it isn't optimum.
Worse, it can limit what you can achieve. Some observers say companies tend to hit a wall when they get 20% to 40% of their environment virtualized because they don't have the management structure and expertise to deal with key questions such as backup/restore and compliance requirements. That stuff is hard enough on a good day, but is different altogether when unified compute/storage/network virtual assets move around freely.
There isn't, however, a one-size-fits-all way to deal with this new reality. Add to that the fact that organizational overhauls are hard and time consuming, and it isn't surprising we've been slow to face up to the matter.
One company that has grabbed the bull by the horns is Cisco. John Manville, vice president of IT Network and Data Services, had a team of 450 organized in the traditional silos; he had a platform team, a storage team, a network team, etc. (See here for a full interview with Manville.)
The result was "sub-optimal system level designs," he says. "We had networking optimally designed for just the network, and the platform group designed optimally for platforms. But as an end-to-end system, it wasn't optimally designed and we decided to change that."
So Manville tore that apart and installed "an architecture team, a design team, an implementation team and, effectively, an operations team." Then he built what amounts to horizontal virtual service teams. "So there's a network service and we have a service owner for that, and that person has virtual members in all the other groups, and their role is to make sure the overall network service meets the requirements," he says.
It took a year to see it through, Manville says. "And I can tell you, it wasn't easy because various teams were worried their role was going to get taken over by other people."
But he thinks it's working. "I think most people have realized their role becomes richer and they can have more direct impact on running this organization," he says. And "while it's not perfect, this is a much better organizational structure, especially as we get closer to virtualization."
Read more about data center in Network World's Data Center section.