Two smart guys
Most of the emails and phone calls I get from vendors and PR people feature overblown phrases like "industry-leading," and are peppered with the typical tech litany of something being the first, fastest, biggest (or smallest) and so on. These exaggerated claims can sometimes overshadow the fact that there are plenty of smart people in the storage game.
I was reminded of that last week when I met with two insightful industry experts. The meetings were eye-opening--maybe not epiphany-like experiences, but they made me think about some things a little differently.
Cheng Wu is the founder of Acopia Networks, purveyors of file virtualization appliances. He sees virtualization more as a first step toward truly manageable storage networks rather than a solution unto itself. Virtualization can serve as the framework for disaster recovery, information lifecycle management and data protection, and still address the things commonly associated with virtualization like better capacity control, management of mixed systems and data migration. Virtualization becomes part of the infrastructure because, according to Wu, you won't be able to do any of those things effectively across disparate systems without it.
He also contends that data migration and its associated tiering has to be a two-way street because data that seems to be of a low value now can become valuable again later. This means all company data is treated as an asset that doesn't necessarily retire to some offline media, never to return. Think of the repercussions of not making this accommodation: What if your business can't get to data when it's needed?
Wu also spoke about "Tier 0" storage, a capability recently introduced in his company's product line. Tier 0, reserved for the most demanding apps, is actually virtualized server memory that can act as a high-performance virtual disk. Tier 0 can tap existing server memory without disrupting operations, or it can use one or more servers loaded with memory specifically designated for this use. I don't know how well it really works, but it sounds like a cost-effective alternative to some other solid-state storage products, and it's certainly a cool way to use installed resources.
A second conversation, with Cisco Systems' George Kurian, was equally enlightening. Kurian is VP and general manager of Cisco's application delivery business unit, and oversees WAFS and WAN optimization products. He suggests that the benefits of wide-area storage management are far from limited to protecting remote site data. In fact, he says consolidation may be the key factor for implementations today, and that the cost of wide-area gear is often easily outweighed by the savings realized by eliminating remote storage and servers.
But the consolidation process is far more than just putting all of one's eggs in a single basket; it's about discovery of intellectual assets. Using WAN technology to centralize data and storage operations will likely reveal these hidden assets that often lurk undetected in DAS. It also provides the opportunity to reclassify data and eliminate--or relegate to cheap media--data that's no longer useful, but still clogging company storage.
One remote application often overlooked during WAFS evaluations is VoIP. Kurian says VoIP is growing rapidly and could easily suck up bandwidth and hinder other wide-area consolidation efforts. All of this boils down to broadening one's perception of WAFS and WAN optimization. Kurian says not to think of it as "just a widget," but rather a system solution.
File virtualization and WAFS/WAN optimization may offer more than first meets the eye. I don't think anyone who's serious about storage views these technologies as point solutions. But, after talking with Wu and Kurian, I'm convinced their respective technologies may offer real opportunities to restructure entire storage environments.