Will EMC's rising tide float all storage boats?

Given that the phrase "our current economy" has such a negative connotation, EMC's third-quarter earnings report last week was downright upbeat.

EMC surpassed its own optimistic guidance for the fiscal quarter just ended by 4 percent. Wow. Other phrases like "increasing confidence among customer in spending their IT budgets" and "very weak first quarter progressing to more normal third and fourth quarters (of 2009)" were heard on the conference call with analysts. Cool. Happy days are here again for the storage industry, right? Well, it depends.

EMC is still regarded as a bellwether for the rest of the storage industry. Indeed, Information Storage as a line item in EMC's third-quarter 2009 financial report accounted for 65 percent of EMC's Information Infrastructure revenues. Upbeat forecasts are often extrapolated to other publicly and privately held storage companies. But EMC is changing. While it's still "where information lives," strategic initiatives include cloud computing, security, broader IT services, and of course, virtualization. It's a very different storage company than the rest of the pack in that all of these initiatives now contribute directly and indirectly to the growth of its Information Storage division. No other storage company operates under EMC's model.

Additionally, there are forces currently at work in the marketplace that tend to mitigate against either raw storage growth or profit margins. Data deduplication technologies are now in the mainstream of IT operational processes. Reducing physical storage requirements by a factor of 20:1 for a growing number of data types is increasingly commonplace, reducing the demand for raw physical storage. And while cloud storage is the latest craze, it's a tough place to generate profit dollars. Cloud storage buyers like commodity hardware, open-source software, and a DIY attitude.

I agree with EMC's executives, who are predicting that the storage industry can expect to see purchasing activity returning to somewhat normal levels in the coming quarter and extending into 2010. But I don't think we'll see a rising tide that automatically floats all boats. Rather, positioning and an ability to integrate with other things--platforms, applications, processes--will be key predictors of success.

Again, let's look at EMC. No doubt, VMware is now an indirect growth contributor to EMC's Information Storage division. The same is true for 3Par, Dell/EqualLogic, HP/LeftHand, and NetApp. It should by now be clear that integration with server virtualization is goodness if you're a storage vendor.

Positioning in the storage cloud is chic but risky at the moment. Nevertheless, if you as a storage vendor can survive the harsh profit environment, your reward will come in due course. Cloud service providers are getting away with the data center equivalent of murder right now. Security is all but nonexistent and incidents of data loss are increasing. Yet providers are still able to sign-up customers who ignore the fine print in the service agreement that absolves them of all liability. That will change. And when it does, the demand for infrastructure that supports a higher quality of service will rise.

Then there are the "stack" vendors--those who are now pursuing an end-to-end, application-to-disk and everything-in-between product integration strategy. These include IBM, HP, Oracle/Sun (yes the deal gets done), and Cisco/EMC. The move by enterprise IT to virtualization and by inference, the cloud, drives this integration. And one interesting and perhaps intended consequence is this: In the world of the integrated processing stack, all becomes visible. A vendor like Cisco supplying new data center network infrastructure to a customer (Data Center Ethernet for example) can also see and respond to needs for upgraded storage (from partner EMC), new server connection devices like NICs and HBAs, upgraded management software (again from partner EMC). Virtualization gives stack vendors a sales opportunity vantage point that plays directly to the integrated stack strategy. Should smaller point-product vendors be pursuing an integration strategy with them? You bet.

There once was a point in time when the storage industry stood proud, broad-shouldered, and fiercely independent. Remember when storage mavens scoffed at Scott McNealy's "storage is a mere feature of the server" pronouncement? Sorry, but those days are gone. Integrate and prosper.

[splunker2]

Take a hard look at IBM's offering of XIV storage. The design uses less expensive disk in a configuration that offers the performance of much more expensive SAN's. It also utilizes off the shelf components such as CPU and RAM. As you add shelves of disk storage to the SAN you are in addition adding more processor and memory. By spreading out the resources in this way and building in intelligence to dynamically move data that requires high I/O to faster disk areas of the SAN it prevents hot spots. No more carving up the SAN for RAID10 vs RAID5. It's all dynamically managed. If anyone doubts the merits and usage of virtual technology see how far VMWare has come.

Obviously there is still a place for your dedicated Tier 1 storage. When it comes to most business needs you simply can't justify the cost of high end storage for all of your solutions when considering the ROI for an XIV SAN without sacrificing the performance of a lower end system. Especially when it comes fully licensed out of the box with licensing for snapshots, performance monitoring, and migration tools. Typically this is where you get dinged on additional licensing costs. This virtual design also reduces the cost of SAN management where a SAN administrator must monitor and tune the resource.

An interesting point, at least word off the grapevine, is that the designer of this technology once worked for EMC and was one of the brain childs of their high end storage. As the story goes EMC saw this cheaper technology cutting into their profit margin on more expensive SAN offerings so decided to shelf it. At least until the designer left to start his own company which IBM then bought a couple years ago. I'm not sure on all those details but you have to wonder why EMC let this slip away and into a competitor's aresnal to cut into their market share.