Mailing List Archives
Authenticated access
|
|
|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[condor-users] New user questions about condor_submit
- Date: Fri, 9 Jan 2004 12:15:54 -0500
- From: Brian Baer <baerb@xxxxxxx>
- Subject: [condor-users] New user questions about condor_submit
Hello,
I've recently installed Condor 6.6.0 onto a couple of linux (Redhat
9) computers to tests to see if Condor will work with an application
that we have developed here.
Condor seems to have installed correctly and I've made some
successful test runs. However I have three questions:
1) We often make a number of runs of the same program with the only
difference being a different random number seed. If we were running
the program from the command line the syntax would be:
primitive -s $seed
inside of our condor command file we could do something like:
executable = ./primitive
arguments = -s $(Process)
queue 10
However sometimes we don't want to start the sequence from 0. Is
there someway to change the starting value of the Process variable?
Or is there someway to do some basic math inside the command file
like:
arguments = -s $(Process)+$(Some_Offset)
or
arguments = -s `echo "$(Process)+$(Some_Offset)" | bc`
2) Our program outputs a number (100's or 1000's) of small files that
are saved to subdirectory. In addition the program creates half a
dozen larger files in the main directory. When I've run the program
under Condor the files from the main directory are brought back to
the submit machine but the subdirectory and the small files that are
in that directory do not not get returned to the submit computer. Is
there a flag to tell condor_submit that I want to bring back
subdirectories as well as individual files?
3) The output files for each run have the same names. Therefore we
need to have different directories for each run. I've used the
statement :
initialdir = testout.$(Process)
However I've had to create each of these directories by hand. Is
there any way to have Condor create these files automatically?
Thank you advance for any advice you can give me.
Brian
--
==============================================================================
Brian Baer baerb@xxxxxxx
Program Analyst
Microbiology and Molecular Genetics (517) 355-6463 x 1602
6176 Biomedical Physical Sciences Building (517) 353-8957 (Fax)
Michigan State University
East Lansing, Michigan 48824-1325
==============================================================================
Condor Support Information:
http://www.cs.wisc.edu/condor/condor-support/
To Unsubscribe, send mail to majordomo@xxxxxxxxxxx with
unsubscribe condor-users <your_email_address>