> As Sun recently released its entire application > stack, it is now viable to run the enterprise > availability tools right in one's rack at home.
Most of these products have been available for free download for education and research use for some time. > There are two products in the aforementioned > application stack - namely the clustering SW and the > N1 Grid Engine. > > My goal is to build a high-availability > infrastructure, for example, build web servers that > automatically fail over to the other node (located on > another continent). Additionally, besides Solaris > I'll also have HP-UX servers. > > Should I use the Sun Cluster or the N1 Grid Engine > for the above? > > What is the difference between the N1 Grid Engine and > the SunPlex clustering SW? In general, the grid engine is a loosely-coupled cluster while Sun Cluster is a tightly-coupled cluster. Traditionally, one would use grid engine for managing compute services where the jobs have a beginning and an end. OTOH, Sun Cluster would be used for continuously operating services. That said, there are exceptions to every rule and some people do use the technologies for multiple purposes. Sun Cluster also has a geographic edition which is designed for long distance separation, and commonly used for disaster recovery style of clustering. However, unlike the grid engine, Sun Cluster will not run on HP-UX. -- richard This message posted from opensolaris.org _______________________________________________ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
