Greetings and Salutations,
One of the articles I’d like to make a regular feature of on this blog are my thoughts on the future of IT technologies that, in my not-so-humble opinion, are hot and/or going to play a major role in our lives over the next three to ten years.
So, in that spirit, I’d like to present the first of these …
It’s a term being bandied around a lot at the moment. Everyone is touting it as “the next big thing” and asking if “you’re ready to join the grid“?
It’s a known state of affairs, suppliers and other major players in the IS&T environment are constantly creating new buzzwords in the hopes of creating greater market awareness, demand and some form of competitive advantage.
Grid Computing is the newest addition to the ever growing list.
Even though in my opinion, many of the “new technologies” emerging are just another evolutionary step, the ever increasing list of buzzwords gives the impression that there’s something new and exciting and altogether different from previous technologies.
You would have heard all of them as well, phrases and words such as adaptive computing, autonomic computing, on-demand computing, utility computing and N1 all discuss the same basic concept.
So, in this rant I will use the term Grid Computing to cover all of these different terms.
[siderant.1] The use of all of these terms has left the impression that the technology is complex and difficult to understand. It’s also alluded to the fact that each supplier is doing something different, and thus, the whole concept is not ready to support mission critical implementations. This simply isn’t true. In my opinion, being able to apply technology to mainstream business requirements, not the technology itself, is the true sense of its value. [/siderant.1]
So What is it?
In essence, the technology enables the use of distributed computing resources (such as processing, network bandwidth and storage capacity) to create and grant it’s users and applications seamless access to vast IT capabilities. The technology is basically based on an open set of standards and protocols (check out the OGSA – Open Grid Services Architecture) that enable communication across (potentially) geographically dispersed environments.
The idea is that with grid computing, you can optimise computing resources, pool them for large workloads, share them across networks and enable collaboration from anywhere across the planet.
Evolution, not revolution
The technology isn’t new, per se, in fact, most of it is built on previous technologies and in some ways can be seen as the latest evolution of other developments. These technologies are all familiar to you, such as distributed computing, World Wide Web, peer-to-peer computing and virtualisation.
Here are the analogies I’d like to use to demonstrate:
Distributed Computing: Grid computing is about bringing computing resources together, not unlike distributed computing, however, unlike clusters and distributed computing, there is no requirement for physical proximity nor operating homogeneity.
World-Wide-Web: For most users, the complexity of the web is hidden, they enjoy a single, unified experience. While the Web is mainly designed to enable communication and information sharing, grid computing enables the ability of full collaboration toward common business goals.
Peer-to-Peer: Most are familiar with the p2p model that allows users to share files. Grid computing ramps that up and allows many-to-many sharing of files and other computing resources as well.
Virtualisation: Just like most virtualisation technologies, grid computing enables the abstraction of IT resources, however, grid computing enables the virtualisation of vast and disparate IT resources, not just a single system.
What does it mean in the corporate environment?
Well, most large vendors are already on board, each offering different value proposition services.
IBM has been heavily involved in the OGSA Global Grid Forum, and with its generally evangelising activities, it is fair to say, has been instrumental in bringing grid computing out of the labs and white papers and onto the mainstream IT floor.
IBM, HP and Sun expect grids will blend into the ordinary IT fabric. They believe it will follow the trajectory of other HPC technologies, such as RISC, SMP, parallel computing and CC-NUMA.
IBM Global Services, CSC and other such professional services and consulting corporations will target global, enterprise, R&D, business analytics, engineering and design grid opportunities. Meanwhile, while CPU cycle scraping doesn’t have much of an enterprise grid future, the per-GFLOP and per-hour metering charge systems do and are firmly in their respective cross hairs.
Unsurprisingly, Microsoft hasn’t had much to say about its grid plans. although it is a sponsor and partner of OGSA and Globus, there isn’t enough of a volume push to force their hand in this space yet. Indeed, I’m wondering if we will hear too much before it completes its Web services push. However, that’s not to say it’s not part of their long term plans. Microsoft’s plan is to have all its products running on .NET within five years. Part of this road-map is its services-oriented Business Framework, which will provide things like transaction and work-flow utilities – they sound very grid-like to me.
For Further Information
- Grid Computing Info Centre
- IBM’s Grid Computing Page.
- Sun’s Grid Computing Solutions Page.
- HP’s HPC Grid Computing Page.
- Microsoft don’t have a dedicated page – see what’s under the research site or read Jim Gray’s (Engineer and Manager Microsoft Bay Area Research Center), August 2002 paper “What should we do about Grid Computing?“
- Oracle Technology Network’s Grid Technology Center.
- OGSA– Open Grid Services Architecture.
- Join the World Community Grid.