Hi Uddhav,
I always used multiple restart files when I did long simulations.
First because most SuperComputerFacilities have only time limited queue's
(24-48 hours) and/or disk quotas, second I found it saver in regard to any
possible down-time of the computer (crash or what's more common,
disk/quota is full, and then you know where exactly you restart your
simulation) and third the backup of big files is sometimes a problem (with
small(er) files you can gzip/compress after each restart and especially
use CD's instead of tapes).
The analysis (carnal,..) can handle multiple trajectories files, so no
problem there.
Cheers
Johannes
-----------------------------------------------------------------------
Dr. Johannes Zuegg
Computational Research Specialist
Research Computing Services
Nathan Campus, Griffith University
Brisbane, QLD 4111, Australia
Tel: +61-[0]7-3875 6603
Fax: +61-[0]7-3875 6650
EMail: J.Zuegg.griffith.edu.au
-----------------------------------------------------------------------
"uddhavesh bhaskar sonavane" <ubs3.rediffmail.com>
Sent by: owner-amber.scripps.edu
18/09/2003 16:50
Please respond to amber
To: amber.scripps.edu
cc:
Subject: AMBER: Suggestion regarding long simulation.
Hello,
I am planning to do long simulation about 30ns. What care
should I take to carry out simulation smoothly? Shall I go for
continuous long simulation with only one trajectory and restart
file or break it in parts with multiple trajectory and restart
files. For long simulation, file formats for trajectory and
restart file or file handling should not collaps. Welcome any
suggestion regarding this.
Regards
Uddhav
___________________________________________________
Click to Order Your Prints For Delivery Anywhere
in India at
http://www.kodakexpress.co.in
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Fri Sep 19 2003 - 00:53:01 PDT
