Autonomic computing - a building block of pervasive computing - continues to gather a strong following. IBM, HP, Sun, Microsoft, Xerox's Palo Alto Research Center (PARC), Carnegie Mellon, The Massachusetts Institute of Technology (MIT) and now Novell are all talking about their autonomic computing strategies, projects, research and/or product offerings and how they will fit into a society saturated with a hopefully more transparent layer of ubiquitous computing devices.
As more and more 'players' begin to weigh in on 'autonomic computing' - and what it means to them - you will find a number of different definitions or interpretations of this concept emerging.
At Computerworld's 20th annual BrainShare Conference in Salt Lake City Matt Hamblen writes that -- Novell's Messman, Stone weigh in -- on this and other future vision topics:
...Novell CEO Jack Messman and Chris Stone, vice chairman in the office of the CEO, talked about Linux at Novell, grid computing and other topics in separate interviews Monday with Computerworld's Matt Hamblen at the company's 20th annual BrainShare conference, which has attracted 6,000 users in Salt Lake City.
Q: What will Novell be offering customers in the autonomic or grid computing space?
There is a lack of clarity around this topic. Everyone has a name for virtualizing data, processes and applications in the computing environment. You can call it grid, autonomic, utility computing, you can call it on-demand, demand more, the list goes on. Those are market terms, but I don't care what you call it. It really is the ability to provision a storage service or an application or a print engine across an array of devices. So, if this computing utilization is at 80 percent on this CPU, I'm going to take the unused print server and place it over here and provision the 80 percent process instantly. And when I do this, the file system knows the move is taking place, the control lists all move along with it, the management environment knows that provision has taken place. ... It's not that big of a deal technically.
But to customers, it's a big deal and here's why: They've got Unix machines or mainframes or whatever and run these things at 30 percent of capacity and they have no way to provision more stuff that's running only at 30 percent. That's why this process makes sense. You save a ton a money. The challenge for companies like Novell is to build a better management platform for grid or utility computing, and when you provision a service to happen, you don't drop stuff like Access Control Lists when you move a process. For example, if I move a file serve process from one blade server to another, I need to made sure the Access Control Lists move, too.
Q: Will you have a product for grid computing at Novell?
Stone: All I'm saying is we understand what it is. We don't have a name yet. We have some technology to build out what I described for you. Pay attention: Watch this space. We'll be participating in providing a solution there shortly, but that's really all we're saying.
Messman: We've been doing some of the dynamic allocation of computing kind of work for years with our resource management product, called ZENworks, Zero Effort Networking, where we manage desktops, download software and keep track of things on the desktops. We've been doing storage-area networks and network-attached storage already for a number of years and that's all coming together in the whole concept sometimes called autonomic computing.
The one thing we'll do that's different from other vendors is the dynamic allocation of applications. We're not so much focused on dynamic allocation of CPUs and storage and saving the hardware costs. We're interested in dynamic allocation of applications, the provisioning of applications to the right people. That's where our identity management product comes into play. As far as a product goes, it might ultimately turn into a product or a version of ZENworks. It's in the engineering stage. There will be other advances in clustering. Those kinds of technologies are all coming together...
K-Collector Topics: Autonomic Computing Hardware platforms Productivity research Writing IBM Microsoft Sun Microsystems