[mythtv-users] Transcodes Getting Stuck

Scott & Nicole Harris snharris99 at live.com
Wed Dec 1 01:05:58 UTC 2010

On Tuesday, November 30, 2010 05:09:54 pm Scott & Nicole Harris wrote:
>> I have had a few instances now, where transcode jobs get stuck and
>> would run presumably forever.  Oddly, the last 3 times have all been
>> with the same program (different episodes, obviously) and the
>> progress seems to get stuck in almost the exact same spot, around
>> Aside from the decidedly negative effect of this causing everything
>> else in the queue to take a permanent hiatus until this job is
>> stopped, it also has the nasty effect of filling up the backend log
>> with line after line of this...
>> 2010-11-30 18:29:55.710 Couldn't open file abort.dat
>> 2010-11-30 18:29:55.718 Failed to decode frame.  Position was: 0
>> I have had two occasions where the jobs were set up just before going
>> to bed for the night, and the jobs ran for 10 hours before I
>> discovered them, and the backend log was in the 100’s of MBs.
>> The only way I have been able to stop these jobs is to go into a
>> terminal, find the PID for mythtranscode and kill it.  Then find the
>> .tmp file for the job and manually delete it.  Any attempts to pause
>> it, stop it, etc from the Info Center are met with failure.
>> I only run lossless transcodes to remove commercial cuts.
>> Has anyone else encountered this issue?  More importantly, is there a
>> better way to kill these jobs?

Brian Wood wrote:
>killall will kill jobs by name. Something like htop will allow you to
>find the process easily and kill it.

I am not terribly concerned with finding the job to kill it, I am wondering 
if there's a better way.  Preferably from within Myth.  I guess the 
extension is, why would the job run on forever, without it being gracefully 
terminated at some point?

mythtv-users mailing list
mythtv-users at mythtv.org

More information about the mythtv-users mailing list