The Slow Pace of Data Center Convergence

Although various approaches to data center convergence have a lot of promise in terms of reducing costs, a combination of economic and internal political challenges is slowing the rate of migration to these new data center architectures.

According to Ashish Nadkarni, a practice lead for Glasshouse Technologies, an IT services provider, the biggest challenge with new system architectures such as Cisco’s Unified Computing System (UCS) or Hewlett-Packard’s BladeSystem Matrix is that most customers lack access to the capital needed to fund what amount to forklift upgrades of their data centers.

Nadkarni says he expects that these systems will eventually make headway in terms of mass adoption, but in the meantime customers are taking an evolutionary approach to data center convergence that favors software-only approaches that make it easier to manage their existing systems. This approach is one for the primary reasons that Dell recently acquired Scalent.

Other issues, says Nadkarni, that are retarding the movement to data center convergence include a lack of industry standards. Customers are concerned that the new system architectures being put forward by the major vendors come with too many proprietary hooks that lock customers into that vendor.

Internal IT politics is playing a role in slowing the general migration to data center convergence as well. Although Nadkarni says he believes this issue will be resolved fairly quickly, as convergence takes hold, the jobs of IT people that formally specialized in server, storage or network management need to be redefined. Until the process is complete, resistance to data center convergence by people that have a vested interest in the existing management paradigm will continue to be a factor.

Of course, Nadkarni adds that the management tools for data center convergence are still fairly rudimentary. And while many IT organizations are curious about data center convergence as part of a general shift toward private cloud computing models, internally, they are not yet prepared to manage IT as a service.

As Nadkarni notes, there’s no point to converging the data center as long as the IT management process that runs it remains fundamentally inefficient. So the challenge facing IT organizations, said Nadkarni, is to reengineer their management processes so they can run IT as a service that embraces data center convergence in order to enable a shift toward private cloud computing.