[gmx-users] Re: Restarts: Truncation of file *.xtc failed
mark.abraham at anu.edu.au
Wed Sep 8 02:21:41 CEST 2010
----- Original Message -----
From: David de Sancho <daviddesancho at gmail.com>
Date: Wednesday, September 8, 2010 8:04
Subject: [gmx-users] Re: Restarts: Truncation of file *.xtc failed
To: gmx-users at gromacs.org
> Thanks Greg for your answer.
> Yes, my simulation is in fact pretty long and its files are considerably large. The size of the xtc files is approximately ~2 GB when I am trying to do the restart.
> I thought on the contrary that my problem was related to another post in today's list (http://lists.gromacs.org/pipermail/gmx-users/2010-September/053701.html). My xtc files were not synchronized (by very little). The answer to that other message was that in case the xtc files were not synchronized one would have to create new and synchronized tpr files using tpbconv and the trr files with positions and velocities.
4.0.x would write checkpoint files for REMD out of sync under some circumstances. One approach is to truncate all the files to a time when you have a checkpoint available for all replicas. Another would be to take the .tpr and .cpt for each replica on its own, adjust the final timestep with tpbconv so that you can synchronise each replica in a single-replica simulation, simulate, and once everything's in sync, restart REMD.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the gromacs.org_gmx-users