AWS Thinkbox Discussion Forums

slave unable to restart deadline

Found a machine thats just unable to restart the deadline slave. The launcher goes just fine.

The launcher log:
2013-08-19 13:46:19: BEGIN - LAPRO0332\scanlinevfx
2013-08-19 13:46:19: Local version file: C:\Program Files\Thinkbox\Deadline6\bin\Version
2013-08-19 13:46:19: Network version file: \inferno2\deadline\repository6\bin\Windows\Version
2013-08-19 13:46:19: Comparing version files…
2013-08-19 13:46:19: Version files match
2013-08-19 13:48:30: Local version file: C:\Program Files\Thinkbox\Deadline6\bin\Version
2013-08-19 13:48:30: Network version file: \inferno2\deadline\repository6\bin\Windows\Version
2013-08-19 13:48:30: Comparing version files…
2013-08-19 13:48:30: Version files match
2013-08-19 13:49:11: Local version file: C:\Program Files\Thinkbox\Deadline6\bin\Version
2013-08-19 13:49:11: Network version file: \inferno2\deadline\repository6\bin\Windows\Version
2013-08-19 13:49:11: Comparing version files…
2013-08-19 13:49:11: Version files match

Then i tried forcing a resynch of all files by copying a version file from an old beta:

2013-08-19 13:49:44: BEGIN - LAPRO0332\scanlinevfx
2013-08-19 13:49:44: Deadline Launcher 6.1 [v6.1.0.52215 R]
2013-08-19 13:49:45: Local version file: C:\Program Files\Thinkbox\Deadline6\bin\Version
2013-08-19 13:49:45: Network version file: \inferno2\deadline\repository6\bin\Windows\Version
2013-08-19 13:49:45: Comparing version files…
2013-08-19 13:49:45: Version files match
2013-08-19 13:49:45: Launcher Thread - Launcher thread initializing…
2013-08-19 13:49:45: Remote Administration is now enabled
2013-08-19 13:49:45: Launcher Thread - Remote administration is enabled
2013-08-19 13:49:45: Launcher Thread - Launcher thread listening on port 17060

Slave still doesnt star. Can start it by double clicking the executable neither.

Does a slave log show up in the logs folder after you try to start it? If so, can you post it?

Also, can you try reinstalling the 6.1 client on this machine to see if that helps?

No, the slave log does not appear.

Yep, ill try the reinstall!

I’ve found a couple more machines doing this, and i noticed a mayabatch.exe process going full tilt on a core. As soon as i killed that process, deadlineslave could be started properly.

Weird…

I can confirm this behavior now… basically, i click on the slave app in the launcher, nothing happens. Go to task managr, find rogue mayabatch process going at 100% on a core, kill it, try slave again, now it starts.

Witnessed this first-hand today… was real weird. It had originally died on a python-related error (the dreaded Access Violation), so maybe it’s related to that whole stuff and will just go away with Ryan’s new fix.

Privacy | Site terms | Cookie preferences