Databases are becoming so huge, particularly for analytics and especially with big data, that they are essentially immobile, says EMC COO Pat Gelsinger, one of the many high profile CEOs to stop by theCube during VMworld 2011 (full video below). “You can’t even back up a 10-or 15-Pbyte environment. It will take you weeks just to back it up once, and by the time you’re done you start all over again.”

This is the big problem with data, he told Wikibon Co-Founder David Vellante and SiliconAngle Founder John Furrier in an interview with  theCube at this year’s event. This means that at least for deep analysis, as opposed to transactional systems, “data mobility is one of the key challenges of hybrid cloud environments. Distance is expensive; latency is expensive. The storage environments are getting so big, so heavy, that you can’t move them around. So you’re either going to find intelligent ways to make them transparent, cache them over distance, or find some other solution.”

One answer, he said, is to take advantage of the mobility of VMs in a virtualized environment to basically bring the software to the data, rather than the other way around. That already is the strategy for Hadoop big data analysis, but he suggests that vMotion in VMware will make that easier to do. And, of course, it also implies that the storage arrays of the future will need built-in processing, both to run tools such as data dedup, compression, and archiving and to run the VM analysis engines that need to access the data.

Some industry observers have raised the question of whether storage technology will basically stall as it disappears under the new virtualization layer. But Gelsinger argues that many tasks are best done at the storage layer, and even though that will be invisible to most observers under the virtualization blanket, a lot of innovation will be happening over the next few years. If security in a virtual environment is a do-over, data mobility is a “get started” – today nothing is out there to do over.

“These types of problems demand new technologies,” he said. “Because now an IT guy doesn’t look at his data centers the same way. He looks at the data center as a collection of these database running transparently, active-active, over distance, plus the services he’s getting from his third-party cloud providers, which is just an extension of his data center.”

Meeting that need is a big challenge, he said. “And frankly, big problems are good business.”


Watch live video from SiliconANGLE.com on Justin.tv