Depending on the day of the week, you might hear any given vendor talking about the benefit of integrated servers brought on by data center convergence. In most instances, the argument is anchored around the concept that systems where the server, networking and storage components are all tightly coupled are easier to manage and inherently less costly over the long haul.
The trouble is that a lot of these arguments assume that no other level of convergence can be achieved without ripping and replacing existing servers. And yet, we just saw HP and Brocade standing shoulder to shoulder in support of an approach to convergence that allows virtual machines running on HP servers to access shared storage resources via Brocade switches. That approach does not require HP to make every component in the system to achieve appreciable higher levels of convergence, noted Herman Chao, Brocade senior director for product management for server connectivity.
For Brocade, the announcement with HP reaffirms the value of a long-standing OEM partnership that has been cast into doubt ever since some of the big server vendors started talking about convergence. Of course, the fact that Brocade was reportedly for sale, but found no takers at an appropriate price level, didn’t help matters either.
Before an IT organization is sold on a major fork-lift upgrade to achieve data center convergence that will indeed lower costs, it needs to determine how it wants to get there. Small evolutionary steps can be more tactically sound, and definitely more affordable, than replacing every component in the data center.
The good news is that data center convergence is happening; the better news is that you can pretty much decide to leverage it at any pace you please.
