Lots to report on various fronts soon, but thought I would point to a new blog that I discovered by Allan Leinwand, one of the partners at Panorama Capital. Some great posts on Vyatta, an open source IP router player.
Allan clearly knows his stuff and has some great perspectives.
Add comment July 31st, 2007
I read a number of blogs on a regular basis. One category I watch fairly closely is storage, largely because it has a direct impact on the future of virtualization, but also because it is also going through some radical changes in terms of storage virtualization and commoditization. Storage virtualization is going to be vitally important to the future of utility computing and how we think about running operational infrastructure in the future.
The CTO of Hitachi, Hu Yoshida, has a great blog on storage and so does Dave Hitz, one of the founders of NetApp. I was reading Hu’s blog today where he was talking about the cost of running storage in terms of CapEx vs. OpEx and again I’m struck by the parallels to virtualization.
Hu makes a great argument regarding the Total Cost of Ownership (TCO) of storage. The key observation I had was that, in the enterprise, it’s easier for the finance folks to measure based on capital expenses (CapEx) than operational expenses (OpEx) because CapEx is an easily measurable expense. I think this is somewhat human nature. Folks simply walk away from deep analysis in those circumstances where qualifying and quantifying are difficult; settling for estimations and best guesses.
It’s not just in storage where this attitude prevails though. I see this playing out with Amazon’s Web Services (AWS) now; particularly the Elastic Compute Cloud (EC2). The commodity nature of the service means that it’s trivial and cheap to get compute cycles, but the cost to run those servers is not. Still, the uptake in terms of startups is tremendous. I’ve been talking to investors on and off for the past 6-9 months and knowledge of EC2 has gone from “yeah, we’ve heard of it” to “every one of our investments is leveraging it” during that time.
EC2 provides great leverage in terms of CapEx, but almost increases corresponding OpEx. It’s hard to use in terms of design, monitoring, and maintenance and while some folks are working on making it easier, it’s still not ‘easy’. Still, it’s trivial to quantify the CapEx costs in this case and AWS is getting great traction from this fact.
In the future, I expect reality to set in that the majority of the cost wasn’t in the hardware to begin with.
Add comment July 17th, 2007
Very cool, one of our competitors, RightScale, launched their new product, RightGrid. It’s an interesting solution they are advocating, although it doesn’t feel much different from a standard batch processing system. But if you are looking for this kind of solution, I would strongly suggest you check them out. Our own objectives are a bit larger in scope.
You may have noticed that the volume on this blog has picked up. Expect quite a few more postings in the next month or two. I’m committed to bring some of our perspective out from behind closed doors and into the blogosphere.
Add comment July 14th, 2007