[Beowulf] Re:Gaussian in parallel (Geoff Galitz)
Many of your questions may have already been answered in earlier discussions or in the FAQ. The search results page will indicate current discussions as well as past list serves, articles, and papers.
Donald Kinghorn kinghorn at pqs-chem.comTue Nov 30 14:04:08 PST 2004
- Previous message: [Beowulf] Gaussian in parallel
- Next message: [Beowulf] Cluster reviews
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Just a couple of comments; g03_C.02_w_LINDA is a 64bit build on the Opteron under linux. for example: file l502.exe l502.exe: ELF 64-bit LSB executable, AMD x86-64, version 1 (SYSV), for GNU/Linux 2.4.1, dynamically linked (uses shared libs), stripped Note that it uses the old thread libs i.e. it needs an LD_ASSUME_KERNEL=2.4.1 on Fedora Core2 I've installed it on several Opteron cluster running Fedora Core2. Runs well and scales good at least out to the 16 processes that I've tested. I don't know about big memory jobs since I've only done minimal testing to be sure that things are working correctly. Gaussian asks that we not benchmark it, and we don't. I'll be doing Mandrake10.1-x86_64 clusters soon and expect to do G03 installs on some of them too. Gaussian tests with SuSE 9.0 for the Opteron and had some difficulties with SuSE 9.1. They usually recommend SuSE 9.0. I've installed 32bit G03 on RedHat 8.0,9, Mandrake9.0,9.1,9.2,10.0 SuSE9.0, ... no real problems. Compiling is sometimes a nuisance but no more than you'd expect for a huge program that is supported on so many platforms. Cheers -Don > We do have it running on two clusters here, it was not > trivial. The hardest part was using a supported platform > and integrating that into our environment. We ended up > using Redhat 8 for it, not willingly. I chatted with the folks > at Gaussian and different people gave me contradictory > information as to whether it worked on more modern > Redhat variants (RHEL, Centos, WBEL and the like). > > I tested it on Centos and ran into problems. > > On RH8, we the trick was to make sure we were calling > the linda binaries and using scratch files that were accessible > across the cluster. > > Also, g03 (according to Gaussian) is still 32-bit in Linux > environments and may be memory bound as a result. -- Dr. Donald B. Kinghorn Parallel Quantum Solutions LLC http://www.pqs-chem.com
- Previous message: [Beowulf] Gaussian in parallel
- Next message: [Beowulf] Cluster reviews
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Beowulf mailing list
