General script run from Monitor or launcher

Hello,

I wrote a submission script. It is stored in ‘DeadlineRepository\scripts\Submission’.
I modify it directly on my shared NAS.
I noticed two different behavior when I run my script from ‘Monitor’ or ‘Launcher’ sub menu.

Imagine I use a version of my script that works well :
Both ‘Monitor’ and ‘Launcher’ run the script properly.

Now I modify my script (Clear its entire content) :

  • ‘Monitor’ fails, ‘Could not run script …’, that is good, it is the right behavior
  • ‘Launcher’ succeededs, it run the previous version (before I broke its code),
    and that is not what I was expecting.

The only way to correct this, is stop ‘Launcher’ and run it again.
I suspect something is put in a cache somewhere, but I found no trace, while I was
looking in :
C:\ProgramData\Thinkbox\Deadline
C:\Users[username]\AppData\Local\Thinkbox\Deadline

Did I miss something ?
Because, I thought ‘Launcher’ sub menu was very helpful for debug.

Thanks

The Launcher does cache these scripts, so that definitely explains the behavior you’re seeing. In Deadline 6, we’ll actually be reverting back to when the Launcher didn’t have these script menus. The reason is that the Launcher was always meant to be a lightweight application with little to no repository interaction, and we slowly moved away from that, so we’re trying to strip it down back to its necessities.

For quick debugging, you could use deadlinecommand.exe instead:

deadlinecommand.exe -ExecuteScript "\\path\to\my\script.py"

Cheers,

  • Ryan

Ok
Thanks for the tip