November 26, 2007
I like to spend my holidays thinking about anything but work, so I was a little upset over Thanksgiving when thoughts of cloud computing kept penetrating my mental barriers and infecting my mind.
“Isn’t cloud computing great?” I would think. “Isn’t it bound to be the next big thing in enterprise distributed computing?” Of course, the skeptic in me also posed some questions, such as “Haven’t we learned anything about the risks involved in giving new technologies catchy labels that mean different things to different people, and nothing to others?”
As an example of the latter thought, I read a couple of blogs discussing IBM’s upcoming Blue Cloud solutions in which the authors praised the notion of cloud computing for small- and medium-sized companies that want access to computing resources while avoiding the hassle inherent with managing their own datacenters, but questioned whether large companies would risk losing their datacenter-derived competitive advantages and becoming part of the greater IT homogeny. It’s a fine take on the future uptake of cloud computing, but it also is reliant on a premise that I don’t think holds much water: Cloud computing is analogous to outsourcing.
After reading IBM’s announcement, I was in no way under the impression that IBM is about to offer businesses access to resources from any source other than their own datacenters. To me, the message IBM seems to be sending is that its Blue Cloud solutions will allow businesses to form a cloud from the business’s collection of resources and that those resources will then be available to users in any location without worry of where their jobs are running. In addition, just like the architecture mimics what we are seeing with the Web, the cloud also can be used to host services that will be available company-wide in the same way the services hosted by Google and Yahoo are available over the Internet. It is very similar to what we already have seen offered by companies like Cassatt, and it does not involve accessing computing resources outside the corporate firewall. In fact, while cloud computing certainly can be an external solution (see Amazon Elastic Compute Cloud), I foresee its early success being mainly on the in-house front.
And, as we have seen with grid computing, it’s not only an issue of how specific vendors or users define a technology, but it also is an issue of whether entities even decide to use the term. Aside from “cloud computing,” terms like “fabric,” “application virtualization,” “datacenter virtualization,” “grid-based application platforms,” etc., all are used to describe similar, if not nearly identical, approaches. One example of this is my recent discussion with Yahoo about its new university partnership program. Whereas “cloud computing” was everywhere when Google and IBM announced their university project, it did not arise once with Yahoo, which seems content referring to Internet- or Web-scale computing.
Speaking of Yahoo’s new university partnership program, I was told by Yahoo vice president of worldwide research operations Ron Brachman, who also heads up academic relations for the company, that the partnership with Carnegie Mellon University is just the first step in Yahoo’s mission to vastly increase its R&D relationships with universities across the globe. “We know quite well we are not the only ones to advance the science and technology of the field,” said Brachman, which is why it is important to work hand-in-hand with the tremendous research talent that academia has to offer. In regard to this specific program, other universities will be brought on board once testing to determine whether the infrastructure is “up to snuff” is complete, but the company also has plans to work with universities to advance open source Web-scale computing in many other ways, including the possibilities of fostering a Hadoop summit meeting or funding specific research projects.
“It is very much our intention to go global with this idea because as we’ve seen with other open source work, the broader the community, the more robust the outcome,” said Brachman.
Of course, the program goes beyond what Brachman calls a “deep partnership” around joint research, as it also is intended (as is the Google/IBM project) to alter the computer science curricula at leading universities with the aim of producing future software engineers familiar with the kinds of computing that are done at Yahoo. According to Jay Kistler, vice president of the engineering for the systems tools and services group at Yahoo, there is a “bit of a shortfall” in current curricula when it comes to teaching this style of computing, and most students -- both undergraduate and graduate -- are not exposed to it in any meaningful way.
Brachman echoed this sentiment, adding that, “As we look to hire new graduates, both at the undergraduate and graduate levels, we find that in most cases people are coming in with a good, solid core computer science traditional education … but not a great, broad-based education in all the kinds of computing that are near and dear to our business.”
However, noted Brachman, Yahoo is seeing a lot of excitement among university faculty and students around the real-world data and computing “that Yahoo lives and breathes every day,” and although still isolated, instances of universities updating their curricula to address Internet computing are increasing.
In regard to the similarities with Google’s recently announced program, Brachman didn’t go into much detail other than saying that there are some differences in the approaches of the two companies (although they are not glaringly obvious to me), but in the end, the more people who are involved in spreading Web-scale knowledge, the better off everyone will be.
Moving on to this week’s issue, we see something similar to cloud computing pop up in the storage space with Seanodes’ new virtual storage pool solution, Exanodes. And although Thanksgiving makes for a slow news week, there are several announcements worth reading, including: “Digipede, Zeliade Deliver High-Performance Financial Apps”; “Sun, Zeus Team on Application Traffic Management”; “iTKO Announces Service-Oriented Virtualization”; and “DataSynapse Appoints New President.” Also, be sure to check out our Q&A with Layered Technologies president and COO Todd Abrams, who enlightens us on DynaVol, the company’s new online, virtual grid-based storage solution.
Comments about GRIDtoday are welcomed and encouraged. Write to me, Derrick Harris, at email@example.com.
Posted by Derrick Harris - November 26, 2007 @ 11:27 AM, Pacific Standard Time
Derrick Harris is the Editor of On-Demand Enterprise
No Recent Blog Comments
10/30/2013 | Cray, DDN, Mellanox, NetApp, ScaleMP, Supermicro, Xyratex | Creating data is easy… the challenge is getting it to the right place to make use of it. This paper discusses fresh solutions that can directly increase I/O efficiency, and the applications of these solutions to current, and new technology infrastructures.
10/01/2013 | IBM | A new trend is developing in the HPC space that is also affecting enterprise computing productivity with the arrival of “ultra-dense” hyper-scale servers.
Ken Claffey, SVP and General Manager at Xyratex, presents ClusterStor at the Vendor Showdown at ISC13 in Leipzig, Germany.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?