From: Brian Bennion (brian_at_youkai.llnl.gov)
Date: Fri Jan 14 2005 - 10:20:49 CST
Hello Bora,
I looked through your files and have some suggestions but no solid
answers.
First, the number of cpus you are using (16) is probably to many for the
number of atoms you have 14K using a ethernet connection. Again, not a
solution, but something to keep in mind.
Second, do you have a "core" that gets created when the system crashes?
Third what do you use to submit your jobs (ie bsub,psub,qsub etc...).
Fourth, does the job always die in the same spot?
And finally, have you run the job with ++debug switch and/or
memory_paranoid?
My first guess is that the ethernet traffic is overwhelming the ability of
NAMD to keep track of atoms properly. The traffic is constant and is only
carrying a small amounts of information which is really hard on most
ethernet cards/networks.
Regards
Brian
On Fri, 14 Jan 2005, bora erdemli wrote:
>
> Hi;
>
> I am sending you my configuration and log file plus
> the error file.
> My simulation was NVT ensemble with periodic boundary
> conditions. I applied Langevin Dynamics and PME and
> 10000 step minimization before the simulation.
>
> I am waiting for your response because I stuck at that
> point.
>
> Thank you for your consideration and responsing me
> everytime I send an e- mail...
>
>
>
>
> --- Brian Bennion <brian_at_youkai.llnl.gov> wrote:
>
> >
> > Hello,
> >
> > I have searched the email archives for your initial
> > email Bora but have
> > been unsuccessful (ie searching for bora and
> > segmentation and fault).
> >
> > Can you restate the problem again, with all the
> > details that you can?
> > Log file snippets are nice as are details of the
> > system configuration (ie
> > type of machine-cluster or standalone, OS, batch
> > queueing system etc...)
> >
> > Thanks
> > Brian
> >
> >
> > On Thu, 13 Jan 2005, bora erdemli wrote:
> >
> > > Hi all;
> > >
> > > I did send an e mail two weeks ago almost with the
> > > same topic and I couuld not find a solution to my
> > > problem even in NAMD web site and mailing list.
> > >
> > > I got this error at the middle of an simulation. I
> > did
> > > NVT ensemble simulation with periodic boundary
> > > conditions.
> > >
> > > Do you have any suggestions?
> > >
> > > Best regards
> > >
> > >
> > >
> > > =====
> > > Sabri Bora Erdemli
> > > Koc University
> > > Computational Science and Engineering
> > > Research and Teaching Asistant
> > > Koc Universitesi pk.218 34550
> > > sariyer Istanbul/TURKEY
> > > tel no: 02123381736
> > > 05326512523
> > >
> > >
> > >
> > >
> > >
> > >
> > > __________________________________
> > > Do you Yahoo!?
> > > Yahoo! Mail - now with 250MB free storage. Learn
> > more.
> > > http://info.mail.yahoo.com/mail_250
> > >
> >
> >
> *****************************************************************
> > **Brian Bennion, Ph.D.
> > **
> > **Computational and Systems Biology Division
> > **
> > **Biology and Biotechnology Research Program
> > **
> > **Lawrence Livermore National Laboratory
> > **
> > **P.O. Box 808, L-448 bennion1_at_llnl.gov
> > **
> > **7000 East Avenue phone: (925) 422-5722
> > **
> > **Livermore, CA 94550 fax: (925) 424-6605
> > **
> >
> *****************************************************************
> >
> >
>
>
> =====
> Sabri Bora Erdemli
> Koc University
> Computational Science and Engineering
> Research and Teaching Asistant
> Koc Universitesi pk.218 34550
> sariyer Istanbul/TURKEY
> tel no: 02123381736
> 05326512523
>
>
>
>
>
>
> __________________________________
> Do you Yahoo!?
> Yahoo! Mail - Helps protect you from nasty viruses.
> http://promotions.yahoo.com/new_mail
*****************************************************************
**Brian Bennion, Ph.D. **
**Computational and Systems Biology Division **
**Biology and Biotechnology Research Program **
**Lawrence Livermore National Laboratory **
**P.O. Box 808, L-448 bennion1_at_llnl.gov **
**7000 East Avenue phone: (925) 422-5722 **
**Livermore, CA 94550 fax: (925) 424-6605 **
*****************************************************************
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:40:29 CST