June 26, 2012
Earlier this month, NASA CIO Linda Cureton detailed several changes to the agency's cloud strategy in a blog post. The organization migrated a number of enterprise applications to Amazon Web Services as well as deploying their "Be a Martian Project" on Windows Azure in what seemed to be a departure from OpenStack. HPC in the Cloud spoke with Cureton about the new direction and NASA's standing with the cloud platform they co-founded.
HPC in the Cloud: Can you tell us why NASA suspended development on OpenStack?
Linda Cureton: It's not NASA's role to develop software, especially software that has a commercial application. That's the way we do other technology, like our space technology. We may work in the early points of the lifecycle to develop the technology, but then at some point when it becomes commercially viable, it's not really appropriate for NASA to continue development. So many other entities can do it much better than NASA, it's their business.
Coincidentally this week, NASA launched our technology transfer portal, technology.nasa.gov, which talks about larger technologies that we initially worked with and then turned over to others to continue development. Our work with OpenStack is really no different than that.
HPC in the Cloud: What drove the Jet Propulsion Laboratory (JPL) to take on Windows Azure for the "Be a Martian" program?
Linda Cureton: The decisions were made early on that, and I think that they were the right decisions at the right time. One of the interesting features about this space is that it changes so fast. You continue to see enhancements, the cost comes down, flexibility goes up, and security gets better. All these things happen at quite a fast rate. So the decision about JPL and Be a Martian isn't a decision against OpenStack, it's a decision on what's the best solution for them, and they made that decision at the time.
HPC in the Cloud: In regards to the Amazon migration, you mentioned in your blog that the agency would save $1 Million annually. Can you elaborate on how those savings were achieved?
Linda Cureton: We did see some real savings in the case of Amazon. We were getting the services from a commercial datacenter. Doing a comparison in terms of the commercial offerings that are out there, Amazon appeared, at least from where we stood, in working with our primary contractor who worked with this. It seemed like this was the best way to go and the best solution.
I want to make a point here, that a decision for Amazon is not a decision against OpenStack. It's just that the commercial offerings are very competitive, and that's what we found when we did make the decision to put this out in the open source community, because we couldn't keep up with it and we found that we were just putting features that other offers like Microsoft and Amazon could do better, faster, cheaper, all that stuff. For us to compete with them with our internal instantiation of OpenStack didn't seem like a good business deal.
Our computing strategy for cloud now, is to choose the right cloud strategy for the right environment. Everything is not OpenStack. OpenStack, I think, has a great value when you look at features that aren't necessarily there in the commercial providers. For example, if you have some very specific security requirements or if you are dealing with extremely large datasets, looking at a solution like OpenStack starts to tip the scale to that. But your run of the mill, general purpose type web servicing, a lot of commercial providers have been doing that for a while, and they do that very effectively and so OpenStack needs to compete with that, but I think the big value for OpenStack is with their customized needs.
HPC in the Cloud: Who was the commercial datacenter provider prior to switching to Amazon?
Linda Cureton: All I can say is that these were services we were getting from a vendor called E touch. As part of that contract, they were providing hosting services, so this was an internal efficiency that they were able to realize by doing this.
This is internal to what our integrator is giving us. This isn't a specific thing where we get in and make decisions. We work closely with them as partners and this wasn't a decision to be made in a vacuum of course. Being the primary contractor, they are certainly incented to find the best value that gives us the right kind of service and where they can keep their cost down. In this particular situation, the cost savings were passed onto us.
HPC in the Cloud: Does NASA have any current OpenStack projects?
Linda Cureton: Yes, absolutely. There are pilots going on at several centers. At Ames research center, I think there are two pilots, and there is at least one at Goddard Space Flight Center in Greenbelt, Maryland. And there are a few others across the agency and I think that those pilots will probably bear fruit. It's too early right now to know exactly what's happening and how well it's going, but I think they will see more activity as time goes on. The only caveat there, I would say, is that these commercial providers are fiercely competitive and they are really putting a lot of energy and emphasis in their cloud offering. They are recognizing that the federal government is a very big customer and a very important customer, so they are not sitting back and letting things go by. I expect that some may go to OpenStack, maybe there are some features that come up in one of the other providers that might appear more attractive, but the beauty of this, I think, for our environment, is to have a cloud computing strategy for the agency that's flexible, that is fiercely dedicated to the mission needs and not really sacrificing technology or technology preference for what the mission needs are.
May 23, 2013 |
The study of climate change is one of those scientific problems where it is almost essential to model the entire Earth to attain accurate results and make worthwhile predictions. In an attempt to make climate science more accessible to smaller research facilities, NASA introduced what they call ‘Climate in a Box,’ a system they note acts as a desktop supercomputer.
May 16, 2013 |
When it comes to cloud, long distances mean unacceptably high latencies. Researchers from the University of Bonn in Germany examined those latency issues of doing CFD modeling in the cloud by utilizing a common CFD and its utilization in HPC instance types including both CPU and GPU cores of Amazon EC2.
05/10/2013 | Cleversafe, Cray, DDN, NetApp, & Panasas | From Wall Street to Hollywood, drug discovery to homeland security, companies and organizations of all sizes and stripes are coming face to face with the challenges – and opportunities – afforded by Big Data. Before anyone can utilize these extraordinary data repositories, however, they must first harness and manage their data stores, and do so utilizing technologies that underscore affordability, security, and scalability.
04/02/2012 | AMD | Developers today are just beginning to explore the potential of heterogeneous computing, but the potential for this new paradigm is huge. This brief article reviews how the technology might impact a range of application development areas, including client experiences and cloud-based data management. As platforms like OpenCL continue to evolve, the benefits of heterogeneous computing will become even more accessible. Use this quick article to jump-start your own thinking on heterogeneous computing.