dvd::rip

->ABOUT / NEWS

->DOCUMENTATION
-TABLE OF CONTENTS
-INSTALLATION
-USING THE GUI
-CLUSTER MODE
-FAQ

->KEY FEATURES

->DOWNLOAD

->SUPPORT

->TRANSLATIONS

->MAILING LIST
-SEARCH ARCHIVE

->CHANGE LOG

->CREDITS

->TODO

->LINKS

http://www.exit1.org/


Re: [dvd::rip] Nodes in Cluster never return

Subject: Re: [dvd::rip] Nodes in Cluster never return
From: Dan Sherman <sherm126@xxxxxxx>
Date: Wed, 29 Nov 2006 21:15:18 -0500
After some more testing, it looks like transcode is hanging on the
remote boxes (but not the local one, odd?).  With -q 2 (debug) on, it
looks like it hangs after its done completely transcoding.  Last few
lines are:

[demuxer.c] (pid=10513) 0/12230 packets discarded 0:00:40, ( 0| 0| 5) 
(encoder.c) encoder closed1000],  53.81 fps, EMT: 0:00:41, ( 0| 0| 4) 
(decoder.c) import stop requested by client=-1211062080
(main=-1211062080) import status=1
(decoder.c) A/V import canceled (-1211062080) (-1211062080)
(decoder.c) video thread exit (ret_code=0) (status_code=4294967295)
(decoder.c) audio thread exit (ret_code=0) (status_code=4294967295)
(decoder.c) vframe_list_lock=0
(decoder.c) aframe_list_lock=0
... hang ...

Killing with kill -9 works.

>From reading around, LD_ASSUME_KERNEL=2.2.5 has given some help (also a
bit on this list about it).  However, in both Gentoo and Ubuntu 6.10,
adding the LD_ASSUME_KERNEL=2.2.5 just leads me to:
transcode: error while loading shared libraries: libpthread.so.0: cannot
open shared object file: No such file or directory

libpthread.so.0 is spread in many places around my machine, so I'm not
quite sure why it can't find it...
/lib/tls/i686/cmov/libpthread.so.0
/lib/libpthread.so.0
/usr/lib/libpthread.so.0

Any thoughts?

- Dan Sherman

On Wed, 2006-11-29 at 10:10 +0100, JÃrn Reder wrote:
> Dan Sherman wrote:
> 
> > Has this been addressed before?  Or shall I start digging further?
> 
> For me this looks like the recently discussed "transcode freezes" 
> problem, which seems to have a special relation to Ubuntu 6.10. 
> Unfortunately for now there is no solution resp. the exact source of the
> problem isn't located yet (whether it's transcode or dvd::rip which is 
> doing something wrong here, although I guess it's transcode... ;)
> 
> You said the CPU on your cluster nodes is at 0%, but ssh never returns.
> I assume transcode processes (transcode, tccat, tcdecode etc.) are still
> running resp. sleeping? Please post a corresponding "ps" output, just to
> be sure.
> 
> Which architectures (Athlon, Intel, 64/32 Bit) have your nodes? I 
> recently installed Ubuntu 6.10 to reproduce this on an Pentium III 
> notebook, unfortunately everything works well for me which makes 
> debugging this quite difficult...
> 
> Does non-cluster mode work for you?
> 
> Regards,
> 
> Joern
> 


<Prev in Thread] Current Thread [Next in Thread>
 

Archive powered by MHonArc. Search powered by ht://dig.

[ top ]