april 20

Could this finally break vendor lock-in?
XAM creates the handshake for CAS
Computerworld Opinion by GlassHouse CTO, Jim Damoulakis

http://www.computerworld.com/action/article.do?command=viewArticleBasic&articleId=9017320

Spending this week at SNW, I've been impressed on several fronts, but, for me, there have been two standouts. At the risk of cheerleading, one simply can't ignore the renewed level of interest and enthusiasm in the many different virtualization offerings targeting all levels of the market. It's very easy to forget that just a few years ago virtualization had become a dirty word. I attribute the change primarily to two factors: first, the enormous success of VMware in reopening people's minds to the notion of virtualization, and second, the normal product maturation process resulting in the current generation of storage virtualization products that provide real value. Rather than stressing virtualization for its own sake, they focus on enhanced functionality that just happens to be made possible through virtualization. Specifically, I'm referring to products from companies like 3Par, BlueArc, FalconStor, Attune Systems, and others where manageability is being dramatically simplified and functionality improved through the virtualization technologies.

Long-term data retention also got some much-needed attention this week. The not-so-secret shortcoming of almost every data archiving offering available today is the (mostly) unintended consequence of vendor lock-in that exists at both the software and hardware layers. Consider, for example, the fact that each e-mail archiving application stores data in its own proprietary format. Additionally, if you happen to be storing the archived data in a content addressed storage (CAS) system, it is indexed and stored in yet another vendor-specific format. This proprietary double-whammy makes any attempt to migrate to another hardware platform or switch e-mail archiving applications a virtual (no pun intended) impossibility.

To address at least the lack of CAS portability, the Storage Networking Industry Association announced an initiative to develop a XAM (eXtensible Access Method) software development kit (see "XAM working group to standardize metadata"). XAM is a proposed standard API for fixed content storage, and members of the initiative include EMC, IBM, Hitachi Data Systems, Hewlett-Packard, Permabit and others. If and when the standard is adopted, it will enable practical migration of data from one supported platform to another.

To address software lock-in, it will also be necessary for application vendors to also support a common data standard as well. At this point, it is not clear whether or not there is support for an initiative like XAM among the e-mail archiving, ERP or database vendor communities. Some argue that a storage-level API is not the best place to standardize and suggest that it be done at the protocol level. One such initiative, JSR (Java Specification Request) 170, is a content repository API and has the support of application providers like Oracle and SAP. While XAM supporters are hoping to enlist application support, it is unclear how this will eventually play out.

At any rate, it is good to see positive movement on this front, and I wish the SNIA Fixed Content Aware Storage initiative well. One way or another, if we ever want to effectively archive data, we need to address this issue, and I urge companies to demand that their vendors -- both hardware and software -- support these standardization efforts.

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.