Re: what is its relationship to clustering (e.g. JBoss)?
Thank you for your comments and question.
Indeed, JPPF is not a cluster, but rather a computational grid framework.
The generally agreed upon difference between clusters and grids is that clusters are made of a set of tightly coupled, homogeneous systems (hardware/OS/software) in a single location or complex, whereas grids are loosely coupled, heterogeneous and widely distributed in nature.
In a larger definition, a grid provides access to computational and storage resources across organizational boundaries. In effect, it virtualizes these resources from the user's perspective, as the user doesn't need to know what these resources are made of or where they are located to use them.
A J2EE application server such as JBoss provides clustering capabilities. Its main goal is to serve a large number of relatively short-lived interactions (or transactions) with a large number of users.
On the other hand, a computational grid will serve requests for long-lived intensive computations that can take hours or days or even longer to complete.
In short, frameworks like JBoss and JPPF provide capabilities in different areas, where they do not compete with each other. It is my belief that they instead complement each other. To demonstrate this, we (JPPF team) have developed a connector between JPPF and the major J2EE application servers in the market, including JBoss.
I invite you to find more about it on the JPPF web site at http://www.jppf.org
I hope this answers your question,
> This project sounds very interesting. My
> question is: what is its relationship
> with clustering setup for example in
> JBoss? Is it a similar thing but in a
> new development? or is it complete a new