[gmx-users] strange continued jobs

[Justin Lemkul]

On 9/11/12 3:11 AM, Albert wrote:
> hello:
>
>    I've restart my jobs with command:
>
> mpiexec -n 160 /opt/gromacs/4.5.5/bin/mdrun -nosum -dlb yes -v -s md.tpr -npme
> 20 -o md2.trr -g md2.log -e md2.edr -cpi
>
> The previous log file is:
>
>
> --------------------log---one---------------------------
> DD  step 113469999  vol min/aver 0.540! load imb.: force 40.5%  pme mesh/force
> 0.629
>
>             Step           Time         Lambda
>        113470000   226940.00000        0.00000
>
>     Energies (kJ/mol)
>            Angle            U-B    Proper Dih.  Improper Dih. CMAP Dih.
>      1.70355e+02    4.99014e+04    3.43419e+04    8.47175e+02 -1.29907e+03
>            LJ-14     Coulomb-14        LJ (SR)        LJ (LR) Disper. corr.
>      1.00912e+04   -5.66686e+03    1.58148e+04   -4.31489e+02 -4.61752e+03
>     Coulomb (SR)   Coul. recip.      Potential    Kinetic En.   Total Energy
>     -4.29387e+05   -9.64431e+04   -4.26678e+05    1.22078e+05 -3.04600e+05
>      Temperature Pres. DC (bar) Pressure (bar)   Constr. rmsd
>      3.03352e+02   -1.64273e+02    6.28299e+01    1.75083e-05
> -------------------------------------------------------------
>
>
> After I continued the job, I found my log file is the following:
>
>
> ------------------long----------two--------------------
>
> DD  step 86759999  vol min/aver 0.450! load imb.: force 38.0%  pme mesh/force 0.771
>
>             Step           Time         Lambda
>         86760000   173520.00000        0.00000
>
>     Energies (kJ/mol)
>            Angle            U-B    Proper Dih.  Improper Dih. CMAP Dih.
>      1.53725e+02    5.06175e+04    3.48843e+04    8.51459e+02 -1.35598e+03
>            LJ-14     Coulomb-14        LJ (SR)        LJ (LR) Disper. corr.
>      1.02049e+04   -6.48180e+03    1.63099e+04   -4.31662e+02 -4.62554e+03
>     Coulomb (SR)   Coul. recip.      Potential    Kinetic En.   Total Energy
>     -4.29154e+05   -9.60364e+04   -4.25064e+05    1.21298e+05 -3.03766e+05
>      Temperature Pres. DC (bar) Pressure (bar)   Constr. rmsd
>      3.01413e+02   -1.64844e+02   -7.07993e+01    1.74002e-05
> -----------------------------------------------------
>
>
> obviously, the continued job didn't wirte log based on previous time.
>

What's present in the .log file is rather irrelevant.  What's important is the 
contents of the .cpt file that was used as input for -cpi.  Use gmxcheck to find 
out what time from which the simulation should have continued.  It seems odd 
indeed that you would have gone back 50 ns, especially since the .cpt file 
should be written every 15 minutes.

> I am just wondering is it OK for this continued job when I am trying to merged
> it with previous after it is finished?
>

Given the information at hand, there's nothing immediately "wrong" about what 
you're seeing.  Inconvenient, perhaps, to have lost time, but the contents of 
the .cpt file used will confirm what should have happened.

> Also I found that the trjconv command doesn't work when I try to extract
> trajectory for analyis:
>
>
>
> trjconv -f md2.trr -s md.tpr -pbc mol -o md2.xtc
>
> -------------------------error-------------------------------------
> Select a group: 0
> Selected 0: 'System'
> trn version: GMX_trn_file (single precision)
> Reading frame       0 time 2000.000
> Setting output precision to 0.001 (nm)
>
> -------------------------------------------------------
> Program trjconv, VERSION 4.5.5
> Source code file: futil.c, line: 459
>
> File input/output error:
> md2.xtc
> For more information and tips for troubleshooting, please check the GROMACS
> website at http://www.gromacs.org/Documentation/Errors
> -------------------------------------------------------
>
> "Don't Push Me, Cause I'm Close to the Edge" (Grandmaster Flash)
>

Do you have sufficient disk space?  Read/write permissions?

-Justin

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

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

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