[ warning: I just posted a similar message to the gridengine users mailing list. I expect to be flamed and/or corrected by the experts so my "diary" may updated or corrected quite a bit over the next few days ] Hi folks, After snneaking over from the Platform LSF world I've just built/configured my first SGE-based life science compute farm. Overall my reaction is hugely positive. Despite being a beginner I was able to install and configure SGE to do what the cluster owners wanted. The system is still going through its shakedown period but at first blush it is delivering everything that was promised. My one negative experience was reading in the quickstart documentation that in order to install execution hosts I was expected to physically login to each machine as root and manually run an interactive shell script. That was just too horrific for words for someone who is used to a systemimager-like world where I can deal with 1 "golden client" and have my changes pushed out to dozens or even hundreds of other cluster members. I ended up doing it the recommended way just to punish myself and then I sat down and wrote some Perl and shell scripts so that I'd never, ever have to do it again. The scripts essentially recreate the configured SGE environment using only 'qconf' commands run from the master host. Now all I need to do to rebuild the system from scratch is run the 'install_qmaster' script interactively. Once that is done three perl scripts take care of everything else. No need to touch the cluster compute elements at all (as long as their NFS mounts and startup scripts are cool...) The scripts and a "diary" from the build and configure process have been placed online at http://bioteam.net/dag/biocluster-diary-1/ Feedback, comments, corrections, etc. are welcome. Regards, Chris -- Chris Dagdigian, <dag@sonsorol.org> Life Science IT & Research Computing Geek- http://BioTeam.net Office: 617-666-6454, Mobile: 617-877-5498, Fax: 425-699-0193 Yahoo IM: craffi