Earlier I reported a problem I was having with recordings of adjacent programs failing on my slave backend, but not the master backend.
I discovered the source of the problem -- the ntp daemon was not running on the slave backend, and its clock had drifted. As such, the master was telling it to start recording too early, while the slave was still recording too much overage on the first show. So the slave would say "go away, I am still recording real show, it's not over yet" and abort the recording. Anyway, the fix is obvious -- make sure ntpd is running to keep your clocks in sync, but I thought I would post it so people could find the answer in searches. I suppose a longterm patch would be for the backend to include its idea of the date in its messages to the slaves (and vice versa) and to issue a warning if they start disagreeing by too much, effectively detecting teh clock mismatch, but I don't think this is a critical issue. It would have been more obvious if the clock had drifted the other direction as shows would be getting shorter and cut off, not longer.
_______________________________________________ mythtv-users mailing list [email protected] http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-users
