Hi Sadact, just saw your post. If you haven't resolved your issue with some of the suggestions in this therad, then use this procedure forthwith..
Firstly I'm not sure (very skeptical) if the whacking the Compressor.app preferences or using some 3rd party app has any result. It's like throwing a ball over the street because its raining. you lose all your settings and layouts and keyboard sortcuts.. so so good.. and has nothing to do with resolving the issue in this case..
In Compressor V4.1 (dec 2013) it appears the state information can be managed/maintained in a few places.
Of late, I've had compressor.app V4.1 users tell me of :
- "run-away" jobs that have been submitted to a "cluster" and they cannot FORCE them out thrugh compressord subtasks.. they just resume
- jobs status quickly swap between ACTIVE and COMPLETED - accomo[anied by endless (each 5 seconds) Qmaster symptom dump crashes (address exceptions)
- compressor machine nodes mounted and fail the job when its submitted and cant get rid of the status (as you describe)
- compressor jobs RESUME after a host startup and run endlessly.
- sounds familiar?
Obvious and simply, your Compressor.app HISTORY and SETTINGS etc are maintained in your home directory ~/Library/Application Support/Compressor :
- CompressorConfig.plist
- CompressorConfig.plist.default
- CompressorPID.plist
- History
- Layouts
- Settings
- Storage
- job4123_distribution_H.264.compressor
- my_compressor_encodes_sasraid01.destinationTemplate
- exports
- jobcontroller:com.apple.compressor.cluster.jobcache.db
TRY this #1: quit compressor.app V4.1 and delete the folder ~/Library/Application Support/Compressor however it probably wont help....!!
Because: In compressor 4.1 (dec 2013) this STATE information seems to be maintained in one of two places and DEPENDS on your configrations setup depending if you specified the following:
Disabling File Sharing: (I leave if OFF - unticked) or Enabling (Default)
and
if you have your HOST (MAC) enable for others to have your machien process their batch jobs. (I set it to OFF for my work machine MACPRO and ON for the mac minis to act as nodes)
![]()
Teh STATE information is maintained for that host (where this information is maintained causing your headache) is where the host compressor.app v4.1 is set as
File Sharing is OFF on all machines: home directory: .. example only ....
~/Library/Application Support/Compressor/Storage/771C3C1A-AB1A646C:
~/Library/Application Support/Compressor/Storage/771C3C1A-AB1A646C/jobs:
- 0E29E980-D037-4E06-8CC6-7E1D2BBAFC86
- directory
- requests
and
/Users/Shared/Library/Application Support/Compressor/Storage/771C3C1A-AB1A646C:
shared
File Sharing is ON in machines: this will be populated.... because it SHARED by defauolt and mounted on those hosts as "SHARED"
/Users/Shared/Library/Application Support/Compressor/Storage/771C3C1A-AB1A646C:
shared
SO..... THIS IS WHERE THIS INFORMATION IS MAINTAINED and so this is you should direct you whacking!
So..
TRY THIS #2: steps as follows.
- Force OUT or STOP compressor.app V4 on each host where you have it running
- delete | purge bothe of these folders using the unix rm command. (Note: folder trash may not work until completely deleted due to UID on file hander... you may NOT be able to empty the trash because the serialisation of the tasks running).
- /Users/Shared/Library/Application Support/Compressor/Storage
- rm -rf /Users/Shared/Library/Application Support/Compressor/Storage
- your home directory's ~/Library/Application Support/Compressor/Storage
- rm -rf ~/Library/Application Support/Compressor/Storage
- launch Compressor.app V4.1 on the hosts. (if you deleted the copressor folder you'll need to set up Compressor and clusters again.
- all should be calm and ready for work.
- submit your batch and hav a cuppa!
Post your results for others to see.
Warwick
Hong Kong