Open Source Cloud Computing Epiphanies and the Structure 09 Conference
Yesterday, GigaOM's Structure 09 conference took place in San Francisco, focused on cloud computing. At GigaOm.com, you can find a number of good posts on what everyone from venture capitalists to Salesforce.com CEO Mark Benioff had to say about cloud computing. Notably, more than a few of the speakers at the conference expressed caution toward the cloud, and, as we've pointed out on a number of occasions, quite a few of the cloud pundits clearly recognize that open source solutions and more favorable economics are essential to a healthy cloud ecosystem.
Data Center Knowledge did some nice analysis here on comments at Structure 09 from WordPress' founding developer Matt Mullenweg, regarding open source. Mullenweg pointed to the open source nginx web server, which does load balancing, as being responsible for significant cost savings in operating WordPress. He also said this:
"When I have to go to the cloud, I consider that a failure. The thing that’s been most exciting to me is how the open source tools have evolved."
Notably, Mullenweg said that WordPress' decision to use Amazon’s S3 storage service to host static files came about simply because there was no appropriate open source solution to compete with it. That's definitely going to change over time. Eucalyptus Systems and many other players have their eyes on the cost efficiencies and peace of mind that open source cloud solutions can offer.
I also noted Liz Gannes' story from Structure 09 about how venture capitalists talk about the cloud all the time, but when asked if they use it, they tell a different story. She points to a panel of VCs at the conference who were asked if they were using cloud services themselves, at which point very few hands went up. They cited concerns over "losing control" as part of why they are cautious with their own use of cloud computing resources.
That issue of control is an important one. What if I put all my files out in the cloud, and then they aren't available when I need them? To protect against that, should I pull my hair out retaining constantly changing backup versions of the files?
Open source has the potential to provide the long-term answer to this issue of control. In my recent conversation with Eucalyptus Systems' founder Rich Wolski, I asked him what companies are doing with the free, open source Eucalyptus cloud infrastructure software (which duplicates Amazon's functionality). He said this:
"They're doing a variety of things, but a lot of them are basically interested in Eucalyptus for doing the same kinds of things they're doing in Amazon AWS, such as business logic applications, where part of the attraction of Eucalyptus is that they can use it as a platform for seamlessly running their public cloud applications and their on-premise cloud apps."
Exactly. If companies can seamlessly use public and private cloud resources in tandem, served by free, open source software conduits such as Eucalytpus, then they don't have to give up full control of what they put in the cloud. Just as critical corporate data gets automatically backed up and mirrored in separate versions, companies taking advantage of a tool like Eucalyptus can leverage public and private cloud efforts in parallel.
There are lots more interesting thoughts on cloud computing, from Structure 09, over at GigaOm.