Checking out the various views on the storage world's future is alwaysinteresting. Most see brighter times coming, eventually -- for storagespecifically and for the economy in general. Of course, many predictedthat would happen in the second half of this year, but I think that'sreally optimistic. I believe that the recent roller-coaster stock marketmeans that any significant upswing won't happen until next year at theearliest. I'm no financial analyst, but everything I've been readingpoints to a very slow recovery.
Plus, pitching the boss for more storage when volume utilization is onlyaround 50%, as is the case in many Windows and Unix shops, is mightydifficult. (Mainframe volume utilization runs higher, at around 80% ormore, I'm told.) And yes, I realize that many shops keep extra storagecapacity on hand so they can quickly respond to users' needs and tospikes in application performance. That said, though, you're stillbetter off right now trying to wring more out of what you've already gotinstead of continually adding more storage to the mix. Plus, you'll payfor all that extra hardware down the line when you're consolidatingservers and trying to make everything work together. Hardwareconsolidation is happening in many open shops, and it's promising to bea bear.
So instead of throwing another array onto the barbie, you might want tolook into things like storage resource management (SRM) and otherrelated disciplines. Do you know how much storage you've got, whatyou're using, and how much is available at any given time? Ifapplication A goes down, will it impact anything on another application?
And what will that down time cost your company?
If you're like most people, you don't know. Not really. (Be honest,now.) So now's a good time to find out. You'll look good to yourmanagement if you can make the case that a $50,000 software investmentwill save a few million in hardware over the next three years. And itjust might.
No matter how efficient you are, though, eventually you're going to needmore hardware. A CIO magazine survey published earlier this year, basedon responses from 100 CIOs and other IT execs, said that their averagegrowth for storage this year will be around 39%, as compared to 36% lastyear, with around 22% of their overall IT budgets allotted to storagethis year. Most of the demand will be from business continuity, ERP,CRM, and Internet applications.
As you're sorting out the strategy that makes most sense for you, hereare a couple of different looks ahead for storage.
On the technical side, predictions were not in short supply at therecent Gartner PlanetStorage conference. SAN management software is out,storage-area management is in. Second, ATA is going to be very big, ascompanies move to that technology from direct-attached storage. There isa down side: although ATA can save you money, there are some potentialcompatibility problems with legacy storage hardware. Third, shy awayfrom any storage vendor that's not willing to play nice and share itsAPIs in the name of standards and openness. Fourth, plan on theconvergence of SAN and NAS into FAS (fabric-attached storage). SANs willlikely run over Fibre Channel and IP, with NAS available for FibreChannel.
This leads to the second round of up-and-coming technology predictions,this from Brad Stamas, chairman of the Storage Networking IndustryAssociation. The group is working on three proposed TCP/IP-basedtechnologies for SANs: Internet SCSI, Fibre Channel IP and InternetFibre Channel Protocol. When products become available to support thesedifferent initiatives -- sometime next year, probably -- it will meanthat customers get to decide whether they want IP or FC-based storagenetworks based on their business needs, not on economics that make sensefor the vendor community. Yee-ha.