[dvd::rip] node in cluster timeout

Hello,
I'm using transcode version 0.6.14 and dvdrip 0.52.5 on debian
testing/unstable systems. I have two nodes in the cluster mode. At some point I
have always to stop the node "humunculus" and start it again, because it encodes
at a very slow rate (the master is running on the other node) When I do that,
everything goes fine. This is the relevant output of the log file. Any ideas how
to find the problem ?

Iván.

Tue Sep 13 20:24:32 2005 Successfully finished job (25): transcode video chunk 
11/15, pass 2, psu 0 on node humunculus
Tue Sep 13 20:24:32 2005 Project file saved to 
'/home/ivan/.dvdrip-master/projects/00000008-maid_in_manhattan.rip'
Tue Sep 13 20:24:32 2005 Starting job (30): transcode video chunk 14/15, pass 
1, psu 0 on node humunculus
Tue Sep 13 20:24:32 2005 execute command: dr_exec ssh -x ivan@humunculus 
"export LD_ASSUME_KERNEL=2.2.5; umask 0002; mkdir -m 0775 -p 
'/mnt/golem/big/maid_in_manhattan/avi/001/chunks-psu-00' && mkdir -m 0775 -p 
'/mnt/golem/big/maid_in_manhattan/cluster/001-00-00014' && cd 
/mnt/golem/big/maid_in_manhattan/cluster/001-00-00014 &&  transcode -H 10 -x 
vob,null -i /mnt/golem/big/maid_in_manhattan/vob/001 -w 829,50 -f 25 -R 1 -y 
xvid4,null -o /dev/null --print_status 20   -S 0 -W 
14,15,/mnt/golem/big/maid_in_manhattan/tmp/maid_in_manhattan-001-nav.log && 
echo DVDRIP_SUCCESS" (timeout=30)
Tue Sep 13 20:24:35 2005 Project file saved to 
'/home/ivan/.dvdrip-master/projects/00000008-maid_in_manhattan.rip'
Tue Sep 13 20:25:56 2005 Job has PID 2946
Tue Sep 13 20:25:56 2005 Command cancelled due to timeout
Tue Sep 13 20:25:56 2005 Aborting command. Sending signal 1 to PID 2946...