George Crump over at Storage Switzerland has a pretty good introductory primer to File/NAS Virtualization. George and I haven’t always seen eye-to-eye, but that’s no surprise, I’m one of those people that wants analysts to prove they’re working on solid ground, and he’s an analyst. Both being type A personalities just guarantees that once in a while we’ll get a little sparky. This time though, he’s got a good intro written up, and though he doesn’t come out and mention it, the standalone, heterogeneous solution he talks about as the second type is pretty much what our ARX series products are all about.

The only thing he missed that I think it is imperative to clearly delineate in any storage virtualization solution primer is that NAS/File Virtualization does not suffer from the horrid “my virtualization box went down, how to find my data” problem that SAN virtualization largely failed because of. With File Virtualization you are still saving the file with the name that the user typed in, it just might be on a different NAS or share than the user thinks – you’re masking where, not what. I know that some vendor’s NAS solutions used an indexing scheme that masked what and where, I presume those have all improved or gone the way of the dodo (though I haven’t looked in a while… If I find some when I look over the next few months, I’ll let you know. If I do find any still  extant, I’ll leave the ridiculing to others though, since those in the File/NAS virtualization space are our competitors ;-)).

The other thing that made me a throw up a little in the back of my throat was his use of the dread phrase “ILM” (Information Lifecycle Management). I shudder when our marketing organization uses it too. ILM had such a huge hype curve that it was doomed to fail when reality showed that its most useful bits were easier merged into other applications and appliances than implemented as stand-alone solutions. ILM products like dedupe and tiering have survived, but this functionality is also merged into existing products. The reason stand-alone products still live is because they are heterogeneous and offer a more data-center wide strategy than the one NAS box you have doing these functions for itself and ignoring all the other NAS your organization owns.

So I prefer some other terminology for tiering and rule-based movement, both of which the ARX does smashingly well (how smashingly well you ask? Watch this space, I’m working up a series). But in essence the tiering bit being associated with ILM is fair, ILM or HSM is where its roots lie anyway. I just think that as early as 2004 some smart people were pointing to the rebound problem and trying to slow the hype curve before it reached the clouds. Sound familiar? Yeah, it happens a lot – XML, Java, SOA, Cloud… All got their share of over-hype, but most found a home. ILM was a group of related ideas and while large chunks found a home in products like ARX, it did not really survive as a single field of technology. So I don’t use it unless I’m talking about other people’s writing.

The biggest problem I see with File/NAS virtualization is one of education – most people don’t fully understand what they can hope to gain from it. Since I’m doing my bit to educate people on the issues (find a full list of my file virtualization articles on my DevCentral “About the Team” page or by searching for “Don MacVittie Virtualization” on Ulitzer), I’m happy to point to Mr. Crump’s article as a good starting point for those of you who are still trying to figure out why you’d bother implementing File virtualization in your organization.

Meanwhile, I’m cooking up that series, expect to hear more from me in the near future – maybe even an “unboxing” video of an ARX, depending upon how mine is packaged when it comes.

Until then, read on, and enjoy our increasingly virtual world!


Fun with full disclosure. I’m an employee of F5 Networks, producer of the ARX family of products, file virtualization solutions. If you think that is enough to make me say good things about the product line, I suspect that you don’t know me very well…

(Picture is By Ian Wilson, released via Wikipedia under Creative Commons license.

Click the image to be taken to the picture in its original context)