[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: out of memory

On Sun, Apr 01, 2007 at 01:18:07PM -0700, Francesco Pietra wrote:
> System; two dual-opteron, amd64 etch, 16GB ram, raid1.
> Heavy computation (memory either 1750 mb or 3750 mb
> per node) started with high % cpu usage and little
> usage of memory. Then, the two factors inverted, the
> HD led became lighted without interruption. The
> computation then closed "incomplete" with the warning
> message in the output file;
> ******************* ARMCI INFO
> ************************
> The application attempted to allocate a shared memory
> segment of 38731776 bytes in size. This might be in
> addition to segments that were allocated succesfully
> previously. The current system configuration does not
> allow enough shared memory to be allocated to the
> application.
> This is most often caused by:
> 1) system parameter SHMMAX (largest shared memory
> segment) being too small or
> 2) insufficient swap space.
> Please ask your system administrator to verify if
> SHMMAX matches the amount of memory needed by your
> application and the system has sufficient amount of
> swap space. Most UNIX systems can be easily
> reconfigured to allow larger shared memory segments,
> see http://www.emsl.pnl.gov/docs/global/support.html

> The http suggested above is of no help, referring to
> Linux kernel 2)
> This warning message was exaclty the same with the two
> different mem allocations. With smaller matrices (ie
> smaller molecules) the same type of computation ends
> OK.
> Thanks for suggestions how to tune the system.

Did you read about how to increate the shmmax parameter, and did you
change it?  It actually tells you how to change that setting.  It still
defaults to 32 MB, and you tried to allocate more (36 MB).

Anyway, use /etc/sysctl.conf to set it.


Reply to: