Re: restart a simulation

From: Ioana Cozmuta (
Date: Mon Apr 03 2006 - 13:37:36 CDT

Hi Lily,

Did you check the output file generated by your previous run? If the
previous run was terminated because of wall time limit, it might have
interfered with the writing of the information in your .coor file. This
means that even if your .restart.coor file is there, it could be
incomplete. A way to check really quick is to see if the size of the
.restart.coor.old and .resart.coor are the same. The same for
.restart.vel.old and restart.vel
Also sometimes the .xsc file could be missing because the program is
just preparing to open the new file handle, it has removed the old one
and the wall time kicks in.

Also make sure that in your configuration file you read in binary
coordinates if you say the way you output them was in binary format.

Hope this helps,

LEWYN LI wrote:
> Lily,
> The dcd file is not needed to restart a trajectory. To read the
> binary coordinates in the restart file, use the command
> "bincoordinates" instead of "coordinates" in your config file.
> Hope this help!
> On Mon, 3 Apr 2006, lily jin wrote:
>> Hi,
>> I have all the other files on the cluster machine except the dcd
>> file from the last run, and I want to restart from the last run.
>> From the configure file I cannot see the dcd file is loaded, but
>> when I sent the job it gave me the error messege as:
>> FATAL ERROR: Error reading binary
>> file../rotation/F1_ATP_in_beta_E_dp_restart_from_17508_64nodes.restart.coor
>> even the coor file is there.
>> do you know if the dcd file generated by the previous simulation is
>> used by the following restart simulation? Thank you!
>> Lily
>> ---------------------------------
>> New Yahoo! Messenger with Voice. Call regular phones from your PC for
>> low, low rates.

Ioana Cozmuta, Ph.D
Eloret Corporation
NASA Ames Research Center
Mail Stop 230-3
Moffett Field, CA  94035
phone (650) 604-0993

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:41:50 CST