SGE 6.x still retains the flexible paradigm where users request the resources they need to be successful and (in most cases) the SGE scheduler will place the job into the appropriate queue. What is new in SGE 6 is the concept of cluster queues which can span multiple execution hosts. The feature is more of an admin or ease-of- management thing than any fundamental change in the flexible way that SGE matches job requirements to available resources. I wrote this document a LONG time ago but it still covers the differences in SGE 5/6/N1GE in some detail, including cluster queues: " Understanding the differences between Grid Engine 5.3, 6.0 and Sun N1 Grid Engine 6 (N1GE 6)" http://bioteam.net/dag/gridengine-6-features.html Regards, Chris On Aug 17, 2006, at 8:49 AM, Mark Hahn wrote: >> And then there's grocking the underlying paradigm; >> in sge5.3 the queue is something that describes a computing resource >> in sge6 (I believe) a queue characterises a job. > > this is a surprising shift to me - queues are less expressive than > some arbitrary collection of flags attached to a job. in fact, > this is one of the main things I've always disliked about LSF > (enough to write my own SQL-based system!) > > regards, mark hahn. > sharcnet > _______________________________________________ > Bioclusters maillist - Bioclusters at bioinformatics.org > https://bioinformatics.org/mailman/listinfo/bioclusters