[gmx-users] Re:Re: Incomplete header: nr 3078 time 9408

[Mark Abraham]

Dechang Li wrote:

>> start by testing the trajectory with gmxcheck
> I did what you said. Following is the message:
> 
> Reading frame    3000 time 9254.000   Warning at frame 3077. Velocities for atom 11896 are large (52568.4)
> Warning at frame 3077. Velocities for atom 11896 are large (7.68295e+31)
> Warning at frame 3077. Velocities for atom 11896 are large (2.75578e+23)
> Warning at frame 3077. Velocities for atom 11901 are large (9410)
> 
> WARNING: Incomplete header: nr 3078 time 9408
> 
> 
> Item        #frames Timestep (ps)
> Step          3078    2
> Time          3078    2
> Lambda        3078    2
> Coords        3078    2
> Velocities    3078    2
> Forces           0
> Box           3078    2
> 
> Did it mean my simulation was failed? But I have checked the energy file "ener.edr", too. The energy file
> seems fine. How did it happen? Can I use the date after time 9254ps?

Use gmxcheck on it too... What we're seeing above is that no frames were
written after 3078. Either your simulation crashed, or your disk became
full. Check your stdout, stderr and .log files for information you've
(apparently) missed about this.

See http://wiki.gromacs.org/index.php/Doing_Restarts for info about
achieving a restart here.

Mark