With its new DS line, IBM is back in storage in a big way. [Editor's note: see IBM's new arrays (Nov. 2004)] The giant admitted years ago that it was getting crushed trying to be all things to all people, especially outside the IBM server base. Now older and wiser, Big Blue is using its bundling muscle to retake a ton of share. The DS launch gives it much more solid products, which positions it to take aim at non-IBM platforms once again. I'll be interested to see how successful IBM can be against the likes of EMC's Clariion, which is growing at a staggering 50%+ year over year. If anyone can make a run at Clariion, it's IBM.
IBM caused a stir initially by renaming the Engenio-made FAStT lineup as the DS4000. Was that a first step toward replacing Engenio with IBM's own products? FAStT is IBM's most successful disk product by most measures, so some smarter IBM execs backpedaled on the new boxes because the channel is mighty happy selling the FAStT. While I fully expect IBM wants to wean you off non-IBM manufactured gear wherever it can, expect to see those boxes for years to come.
In the meantime, Engenio grabbed a bunch of new business by snagging one of the few remaining OEMs in Sun Microsystems. Say what you will about Sun, but it sells a truckload of disk arrays. I now think Engenio will not only survive the IBM threat, but may even prosper for the next several years until IBM shakes itself out.
If Hewlett-Packard (HP) were to realize that it's wasting money making its own disk arrays (it is, trust me) and focus on keeping customers healthy and happy, it would likely hop on board the Engenio bandwagon, too. Engenio is really the only OEM play available to the big guys--I doubt you'll see HP OEM the Clariion. RAID arrays are only strategic if that's your business and, clearly, they aren't strategic to HP (unless they start spitting out ink toner). So why not be like IBM and admit times have changed? HP has a high-end customer base that's second to none and, as long as it maintains that base, it can sell value into it. It just needs to realize that selling Oreos doesn't mean you have to raise milk cows, too.
Speaking of Oreos and milk, I've seen some great partnerships of late where one plus one equals three for customers. I love what NTP Software does for NetApp customers. It provides the same user-friendly quota management we're used to on Windows for filers. Seems trivial, but in a NetApp shop, it's a godsend. I like how Onaro's change-management software makes EMC's ControlCenter work the way it should and makes the whole platform pretty killer. AppIQ and Hitachi Data Systems (HDS) appear to be another match made in heaven, and HDS customers couldn't be happier.
Get ready for much higher software maintenance pricing. Forget the 20% annual contracts going forward. In the age of consolidation, vendors sell fewer licenses. They can either raise the cost of a license or recognize that support costs that used to be amortized over a billion servers can't be maintained over a hundred, and spread the cost to the back end. Oracle may not have to, but many broad-based software products will surely see maintenance costs rise. I'd plan on a 50% increase over the next few years to be safe.
Finally, with the Red Sox actually winning the World Series, and me somehow living through it, I'll need to find new analogies for sinking, choking, plummeting in September, falling from grace, etc. Any ideas would be appreciated. I mean, it just doesn't sound right saying, "That box crashed faster than the Yankees" or "That application sucked all the money out of us then started behaving badly--sort of like A-Rod." Or does it?
- Demystifying storage performance metrics –ComputerWeekly.com
- Containers and storage 101: The fundamentals of container storage –ComputerWeekly.com
- NVM Express, NAND Flash Energize Storage Vendors –SearchStorage.com
- Hybrid Flash: The Essential guide –ComputerWeekly.com