Greetings DCPoMatic mailinglist, I'm encountering an odd bug(?) during "cluster" encoding with 4 dcpomatic2_cli_server encoding servers and one master. The setup worked for years without problems till now. Complicating bug detection is that it works some times and gets stuck other times. With the same source data! Some Information about my setup all using dcpomatic 2.9 (deb) Running ubuntu 14.04 master (headless) + enc1 (kvm VM) * amd64 8core 4GHz * prep is done on different Desktop using same path structure or with dcpomatic2_create * dcpomatic2_cli is running with "only Enc Server" options and there is a dcpomatic2_server_cli -t 8 or -t14 on the same VM (results in much better frames/s when it works compared to encoding it inside "dcpomatic2_cli" on the master during cluster encode. enc2 / enc3 / enc4 all ubuntu 14.04 * 4cores / 4cores / 2cores Hardware machines testing different thread counts for cli_server with -t 8 / -t 4 / -t 2 storage * nfs * debian 7.x (kvm VM) All are conected with a hp pro curve 1GB lan The syslog files on all machines only show normal messages! The crazy part is that in about 95% of the stuck cases i can encode the movie with only changing it with running "dcpomatic2_cli -r XXX" I have included two dcpomatic log files with max debug options ... -- Markus Kalb 06151-16-27 27 0 0177-240 98 95 mk(a)filmkreis.tu-darmstadt.de Studentischer Filmkreis an der TU Darmstadt e.V.