NCL threads?

From: <Oliver.Fuhrer_at_nyahnyahspammersnyahnyah>
Date: Wed, 11 Feb 2009 15:09:20 +0100

Hi all!
 
We've been investigating NCL memory usage in order to have several
instances of NCL run in parallel to produce some plots. While monitoring
NCL with the Linux top command, I've noticed something rather strange...
When an NCL script is executed that produces several plots into PS/PDF
files that are named differently, I do not see only a single "ncl"
instance in the process list. Sometimes (it seems that everytime a wks
is closed) NCL seems to fork and create a new instance of the same
process running. When several plots are produced consecutively,
sometimes there are even three ncl instances visible in the process
list. Since the "forked" NCL instances are exact duplicates of the
original NCL instance, they use exactly the same amount of memory. I am
not sure wether this is actually the case, but it seems that NCL forks
off a process everytime a plot is being finished and written to disk.
 
Our plan is to run several NCL's executables with different scripts in
parallel. If these happen to fork off one or even two children at the
same time, we can run into memory problems (the operating system will
start swapping).
 
It would be interesting to hear why NCL is doing this and if this
behaviour could be avoided by setting some kind of flag.
 
Cheers,
Oliver
 

________________________________________

Oliver Fuhrer
Numerical Models

Federal Departement of Home Affairs FDHA
Federal Office of Meteorology and Climatology MeteoSwiss

Kraehbuehlstrasse 58, P.O. Box 514, CH-8044 Zurich, Switzerland

Tel. +41 44 256 93 59
Fax +41 44 256 92 78
oliver.fuhrer_at_meteoswiss.ch
www.meteoswiss.ch - First-hand information
 

 

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Wed Feb 11 2009 - 07:09:20 MST

This archive was generated by hypermail 2.2.0 : Thu Mar 05 2009 - 15:48:48 MST