Using The Generic Submit Script

To try out the generic submit script, you should first make a copy of the script to a network drive accessible by all machines on your network. (Since this script calls itself as a render script, it must be accessible by all machines via the same path)

See the examples below for how to use submit-generic.pl to render Renderman RIB files.

General Use

  1. Copy the generic submit script from the rush directory to a network drive.

    In place of the following example, use an absolute path that can be accessed by your entire network:

    	   UNIX: cp /usr/local/rush/examples/submit-generic.pl /server/jobs/rush
    	WINDOWS: copy c:\rush\examples\submit-generic.pl \\server\jobs\rush
    	    

  2. Invoke the script from the network drive.

    The interface should pop up. If it doesn't, verify the first line in the script's path points to the correct location of your perl binary.

    Click the help button (buttons with '?') for any field you want to know more about.

    When the GUI pops up, note there are various 'Advanced Options' available if you scroll the window down.

  3. Select the 'Shell' you want to use.

    In this example, choose 'perl'.

  4. Enter the shell commands to execute.

    These are the commands you want to have executed for each frame on all the machines, and will be run in the 'Shell' you selected above.
    Since we've selected 'perl', enter some perl commands you know will work on both unix and Windows platforms:

    	    print "Working on frame $ENV{RUSH_FRAME}\n";
    	    system("netstat -an");
    	    exit(0);
    	    
    Note: If your commands assume you to be in a particular directory, you better include a chdir command to the full absolute path.
    As with all render scripts, you should return a proper exit code; 0=Done, 1=Fail, 2=Retry.

  5. Enter the cpus to use.

    In our case, lets use any 5 available cpus at the lowest priority:

      
    	    +any=5@1
    	    

  6. Set the directory where logs should be written.

    Scroll down to 'Log Directory', and set the absolute path to a network directory which all machines can access.
    In place of the following example, use an absolute path that is relevant to your network.

    	    //net/tmp/logs
    	    

  7. Submit the job by clicking the 'Submit' button.

    The job should submit, and an irush interface should pop up with the jobid already set.

    The next time you bring up the submit form, it will have all the values set from the last execution. Note that you can save and load you data entries for easy retrieval.

    For more info on the irush interface, see the irush tutorial.

  8. Monitor the job.

    Hit 'Frames' in irush to watch the progress of your job.
    If some frames are running, or got done already, click on some, and hit 'Logs' to see the output of the perl commands you entered.

  9. Dump the job when you're done.

    When you're finished with the job, hit 'Dump' in irush.

Example: Rendering RIB Files