AWS Thinkbox Discussion Forums

Deadline monitor error

I was modifying the frame range of a job with a custom range such as:

1001-1130x2

It seemed to not set up the new task list properly, because the task panel only showed ~10 tasks (startin at task id 20 or something like that, so obviously the rest were somewhere around…)
I went to other jobs that had a lot of tasks, but the tasklist showed up empty.

I shut down the Monitor, to try and restart it, but now it doesnt start at all…

I tried restarting the launcher, and now thats not starting neither…

Launcher log:

2013-10-10 10:33:59: BEGIN - VCPRO1014\ScanlineVfx_user
2013-10-10 10:33:59: Local version file: C:\Program Files\Thinkbox\Deadline6\bin\Version
2013-10-10 10:33:59: Network version file: \inferno2.scanlinevfxla.com\deadline\repository6\bin\Windows\Version
2013-10-10 10:34:00: Comparing version files…
2013-10-10 10:34:00: Version files match
2013-10-10 10:35:13: Local version file: C:\Program Files\Thinkbox\Deadline6\bin\Version
2013-10-10 10:35:13: Network version file: \inferno2.scanlinevfxla.com\deadline\repository6\bin\Windows\Version
2013-10-10 10:35:13: Comparing version files…
2013-10-10 10:35:13: Version files match
2013-10-10 10:37:26: Local version file: C:\Program Files\Thinkbox\Deadline6\bin\Version
2013-10-10 10:37:26: Network version file: \inferno2.scanlinevfxla.com\deadline\repository6\bin\Windows\Version
2013-10-10 10:37:26: Comparing version files…
2013-10-10 10:37:27: Version files match
2013-10-10 10:37:31: Shutting down

Trying to start the monitor directly, i get this in the log:

2013-10-10 10:44:48: BEGIN - VCPRO1014\ScanlineVfx_user
2013-10-10 10:44:48: Deadline Monitor 6.1 [v6.1.0.53080 R]

The process does not start

Oddly enough, the slave starts up just fine…

Reinstall didnt fix it… helppppp

Rebooting did though. Weird! Could it be that having a lot of open handles starves deadline out? I had loads of applications, windows etc open…

Weird… really hard to guess what could have caused this. Glad to hear that a reboot fixed it though. If it starts to become a reoccurring problem, let us know!

I can repro this now.

Basically, if i have any framecycler sessions still open that were originally started from the deadline monitor, i will not be able to restart the deadline monitor.

So:

  1. open the monitor
  2. double click on a job, and use our default viewer with it (a batch file that starts framecycler with this command line:

C: cd \ C:\Python26_64\python.exe \\S2\exchange\software\managed\pythonScripts\framecycler_tools\launchFrameCycler.py %1 /BEAUTYJPG

  1. close the monitor

  2. try to open the monitor
    RESULT: it doesnt even pop up in the task manager…

  3. shut down the framecycler session

  4. try to open the monitor
    SUCCESS

You dont need framecycler for it though, you can also just make a default viewer batch file with simply these contents:

pause

Then use it to “view” an image. Shut down the monitor -> unable to start it again till you close the command line window with the pause message…

It doesnt matter if i start the processes with ‘start’ etc… very odd. Some subthread of deadline waits for those independent processes to quit?

If you guys get a chance to look into this, or maybe give a workaround, that would be great!

I can reproduce it on our end, and we’re currently looking into it.

Thanks!

  • Ryan

We figured it out. It will be fixed in beta 8.

Cheers,

  • Ryan

Great news thanks!

Privacy | Site terms | Cookie preferences