october 25

De-dup technology -- a killer app for cutting cost
"A killer VTL feature soon-to-be-available on VTLs is data de-duplication".
Computerworld opinion by Jim Damoulakis, GlassHouse CTO

http://www.computerworld.com/action/article.do?command=viewArticleBasic&articleId=9004425

In last week's column (see: "Speeding up backup: Can VTLs keep up with tape? "), I discussed a scenario where the failure to apply proper architecture and design considerations to a virtual tape library (VTL) implementation led to a reduction, rather than an increase, in performance over a tape library being replaced. This week, I'd like to extend the discussion to highlight some additional factors that you might want to consider when making a VTL selection.

Considering that a VTL fundamentally consists of a special-purpose software application running on one or more servers and connected to a pool of storage, factors like performance, scalability, and availability are dependent to a large degree on the specific hardware selected. However, software design also plays a significant role: some VTLs are designed as standalone entities while other are architected to support multiple data movement nodes supervised by a master. Functionality like remote vaulting, non-disruptive updates, transferring data to physical tape devices are also subject to design goals and tradeoffs, and one approach may be more appropriate than another for a given environment.

A killer VTL feature in the "current or soon-to-be-available" category is data de-duplication or redundancy elimination technology. Today most VTLs offer some level of software-based data compression (at the expense of dramatically reduced performance), and a few have incorporated hardware-based compression similar to that found in tape drives -- providing roughly a reduction ratio of 2:1. Data de-duplication technology extends this much further, ranging anywhere from 15:1 to 100:1! It isn't hard to figure out that this dramatically reduces the disk quantities required for a VTL and will likely conclusively tip acquisition cost calculations in favor of disk over tape. Such a reduction in data also dramatically improves the feasibility of remotely replicating backup data thereby further reducing or eliminating the need for backup tapes.

Data de-duplication technology isn't trivial - there are significant implications relating to performance and scalability, and if you're considering VTL technology, it is important to understand your vendor's position and approach. Not all VTLs are created equal, and as the technology matures, vendors continually strive to differentiate their offerings. Making the right choice still comes back to how well you understand your requirements and priorities.

Jim Damoulakis is chief technology officer of GlassHouse Technologies Inc., a leading provider of independent storage services. He can be reached at jimd@glasshouse.com