CCL:G: best Linux for clusters

[Adam Tenderholt <atenderholt-*-gmail.com>]

 Sent to CCL by: Adam Tenderholt [atenderholt^gmail.com]
 Hi Richard,
 I built and managed a Rocks cluster (based on CentOS) a couple of
 years ago, and found it quite easy to work with. Once the head node
 has been setup properly, compute nodes are added via network
 boot/install. Propagating new users to the compute nodes involves
 issuing a single command, software is shared over a NFS drive, and
 multiple queuing packages are available. The only commercial packages
 we used were Gaussian and ADF, and we had no special hardware.
  Adam
 On Fri, Oct 28, 2011 at 12:02 PM, Venable, Richard (NIH/NHLBI) E
 venabler()nhlbi.nih.gov <owner-chemistry*ccl.net> wrote:
 >
 > Sent to CCL by: "Venable, Richard (NIH/NHLBI) [E]"
 [venabler#%#nhlbi.nih.gov]
 >
 > We have been re-evaluating the choice of CentOS for our cluster, and would
 like solicit comments on which freely available Linux flavor seems to be the
 best for scientific computing, in terms of performance, stability, ease of
 administration, and support for more commercial software (QM packages, Matlab,
 etc.) and hardware (Qlogic Infiniband boards and driver software).
 >
 > Besides CentOS, we are looking at Scientific Linux and Debian, but comments
 on other distributions are welcome.  If you (or someone in your group) can
 provide some detailed remarks noting good or bad features of specific Linux
 distributions, please send them to me and I will summarize to the list.
 >
 > Thank you.
 >
 > --
 > Rick Venable     5635 FL/T906
 > Membrane Biophysics Section
 > NIH/NHLBI Lab. of Computational Biology
 > Bethesda, MD  20892-9314   U.S.A.
 > (301) 496-1905   venabler AT nhlbi*nih*gov>      http://www.ccl.net/cgi-bin/ccl/send_ccl_message>;  
    http://www.ccl.net/cgi-bin/ccl/send_ccl_message>;  
    http://www.ccl.net/chemistry/sub_unsub.shtml>;  
    http://www.ccl.net/spammers.txt>;
 >
 >