Sun Tries to Stay Ahead of Commoditization With Storage
Posted by Bob Warfield on September 8, 2008
I liked Jonathan Schwartz’s blog entry Fanning the Winds of Change in Storage. The seminal passage is here:
Granted, you can see an increasing focus on storage at Sun – the acquisition of MySQL is as much a storage acquisition, as an enhancement to Sun’s developer offerings. Discussions of flash memory, the economics of archiving, the Lustre parallel file system, all point to an increasing focus on what Sun sees as an exceptional opportunity for customers (and thus, investors). Storage and computing are converging – and we’re about to bring the trends that transformed the server industry a few years ago (mass engagement in open development communities, and scale achieved via clusters of commodity parts vs. proprietary technologies) to the historically closed and proprietary storage industry.
I read it a little differently than perhaps Jonathan intended, and perhaps I see Sun in a little different position than Jonathan would want to convey. From my perpsective they’re being increasingly pushed to being a software company. There is less and less advantage in their hardware innovations. The multicore crisis is makinig horizontal scaling a necessity and vertical scaling becomes harder and harder to justify at the limit. It’s become darned difficult for SUn to produce CPU’s that perform markedly better than Intel’s. Hence, their hardware business is steadily being commoditized. At the same time, the OS business is already pretty well commoditized.
Meanwhile, the advent of cloud computing has meant Storage as a Service has real legs. Yes, there has been a ton of storage technology available for deployment inside the firewall, and it’s good stuff. However, cloud computing really brings home the idea that you can start to forget about some of the complexities in storage. Services like Amazon S3 and Block Services for EC 2 really bring a lot of functionality without requiring much effort, thought, or cost to the consumers of these services.
At Helpstream, we converted backups for our SaaS Customer Service application to use S3 rather than tape a while back and have never looked back. The quality of protection these backups afford our users is better than we could afford to do any other way.
Jonathan is still very much focused on selling inside the firewall, with products like the “Thumpers” he mentions as having gone to $100M in sales with 80% year on year growth. The interesting question will be whether Sun or any other major vendor can achieve penetration in the cloud market. Thumper sounds great, but I still don’t see how it lets me do what I need to do as easily, cheaply, and efficiently as Amazon has.