[gmx-users] Segmentation fault while running MD simulation

[Mark Abraham]

rainy908 at yahoo.com wrote:
> Hi all,
> 
> I've encountered a problem while trying to run Gromacs on my 
> coarse-grained molecule.  I've already performed the initial energy 
> minimization.  I tried to run the inital MD simulation, but the job was 
> cut short due to a segmentation fault, as mentioned in the output file:
> 
> ^Mstep 0^Mstep 100, will finish Wed Dec 16 16:55:17 2009^Mstep 200, will 
> finish Wed Dec 16 07:04:11 2009^Mstep 300, will finish Wed Dec 16 
> 12:59:33 2009^Mstep 400, will finish Wed Dec 16 09:02:03 2009^Mstep 500, 
> will finish Wed Dec 16 12:12:01 2009^Mstep 600, will finish Wed Dec 16 
> 14:18:47 2009^Mstep 700, will finish Wed Dec 16 15:49:22 2009^Mstep 800, 
> will finish Wed Dec 16 13:29:16 2009^Mstep 900, will finish Wed Dec 16 
> 14:45:15 2009^Mstep 1000, will finish Wed Dec 16 15:46:03 2009
> 
> /opt/gridengine/default/spool/
> compute-0-7/job_scripts/182710: line 28: 18851 Segmentation fault      
> (core dumped) $MDRUN -v -nice 0 -np $NSLOTS -s npg3_4.tpr -o npg3_4.trr 
> -c confout.gro -g npg3_4.log -x npg3_4.xtc -e npg3_4.edr
> /opt/gridengine/default/spool/compute-0-7/job_scripts/182710: line 31: 
> -s: command not found

So what's on line 31 of your submission script? The above is just stderr 
and stdout, which are not the main output. What's in the .log file?

Mark