From: Alessandro Mascioni (masci002_at_umn.edu)
Date: Sun Jul 21 2013 - 12:24:23 CDT
Hello,
I have a question about restarting a simulation after a computer crash. For
some reason my computer rebooted spontaneously.
When I look at the restart files, they have a size of "0" and namd cannot
open them:
-rw-r--r-- 1 me me 38008708 Jul 21 09:44 protein.dcd
-rw-r--r-- 1 me me 39736364 Jul 21 09:44 protein.dcd.BAK
-rw-r--r-- 1 me me 0 Jul 21 09:44 protein.restart.coor
-rw-r--r-- 1 me me 1727500 Jul 21 09:44 protein.restart.coor.old
-rw-r--r-- 1 me me 0 Jul 21 09:44 protein.restart.vel
-rw-r--r-- 1 me me 1727500 Jul 21 09:44 protein.restart.vel.old
-rw-rw-r-- 1 me me 0 Jul 21 09:44 protein.restart.xsc
-rw-rw-r-- 1 me me 241 Jul 21 09:44 protein.restart.xsc.old
-rw-rw-r-- 1 me me 114068 Jul 21 09:44 protein.xst
I assume the computer crashed before/during the restart files were updated.
In this case should I restart from the .old files? Most of the
configuration files I found on this mailing list do not restart from these
.old files.
When I use the .old files as restart condition namd seems to work fine.
Also, for the firsttimestep, I used a script from this mailing list to
determine the last timestep saved from the interrupted simulation.
Should I set this timestep as firsttimestep and then cut the frames >= of
this timestep from the old trajectory before merging the two trajectories?
Thanks in advance,
Al
This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:21:28 CST