[gmx-users] time steps and segmentation fault

[Justin A. Lemkul]

albitauro at virgilio.it wrote:
>  Hi all,
> 
> I carried out a MD simulation using the MARTINI CG force field and I had 
> some
> problems.  The simulation stops immediately, reporting a segmentation fault:
> 
> **********************************************+
> Making 1D domain decomposition 2 x 1 x 1
> starting mdrun 'FIB'
> 200000 steps,   6000.0 ps.
> step 0
> [beta:06822] *** Process received signal ***
> [beta:06822] Signal: Segmentation fault (11)
> [beta:06822] Signal code: Address not mapped (1)
> [beta:06822] Failing at address: 0x34321d0
> [beta:06822] [ 0] /lib/libpthread.so.0 [0x7f7e49b38a80]
> [beta:06822] [ 1] mdrun_mpi [0x66549a]
> [beta:06822] *** End of error message ***
> mpirun noticed that job rank 0 with PID 6822 on node beta exited on 
> signal 11 (Segmentation fault).
> 1 process killed (possibly by Open MPI)
> **************************************************
> 
> I have no WARNINGs in the output of the simulation.
> However, before, when running grommp I received 1 warning and 1 note:
> 
> ********************************************************
> WARNING 1 [file aminoacids.dat, line 1]:
>   For proper thermostat integration tau_t (0.1) should be more than an
>   order of magnitude larger than delta_t (0.03)
> ********************************************************
> and
> ********************************************************
> NOTE 1 [file ny6.mdp, line unknown]:
>   The Berendsen thermostat does not generate the correct kinetic energy
>   distribution. You might want to consider using the V-rescale thermostat.
> ********************************************************
> 
> I used a time step of 0.03 ps and a tau_t of 0.1 (suggested by the 
> authors of MARTINI ff even when using a
> delta_t=0.04 ps). In any case, modifying the tau_t and avoiding the 
> WARNING does not solve
> the problem.
> 
> The simulation starts and ends correctly if I use a delta_t=0.01 ps.  
> However,
> this time step is not recommended for use with MARTINI ff because the 
> range of
> time steps used for parametrization is 20-40 fs.
> Does anyone have any suggestion to solve the problem?
> 

In my tests with MARTINI (which are limited, granted), the largest timestep that 
I found to be stable was 20 fs.  Anything larger led to a crash.

-Justin

> Many thanks,
> 
> AM
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> gmx-users mailing list    gmx-users at gromacs.org
> http://lists.gromacs.org/mailman/listinfo/gmx-users
> Please search the archive at http://www.gromacs.org/search before posting!
> Please don't post (un)subscribe requests to the list. Use the 
> www interface or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/mailing_lists/users.php

-- 
========================================

Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================