The setup:
Backend with 3x PVR250
P4 2.8Ghz, 1GB Ram
2x SATA Seagate Barracuda 300GB disk
The ringbuffer, music, videos, database, are all on one drive, recorded TV is on the other
No frontend on the backend server
The backend is connected to gigabit ethernet switch
2x frontends - no capture on the frontends frontends are connected via 100Mb FDX P4 2.x Ghz with 512MB RAM nvidia FX5200 video cards
All systems are running debian unstable with 2.6.9 kernels
From what I can tell so far the problem appears to be some sort of blocking problem on the backend - I have noticed that sometimes when this occurs mythcommflag is running on the backend, but can't say that it is for sure all the time. The recordings themselves are fine, as I can stop the playback, and return to it sometime later, and all is fine.
I am inclined to think this is not a frontend issue, because I can playback the shows later with no issues. I have looked through all of the relevant logs, and the only thing I see that arouses suspicion is the backend log has a number of IOBOUND messages in it from time to time.
The IOBOUND error seems odd to me for a few reasons:
1) the hardware part of this setup has not changed since 0.16 and I *never* had this problem with 0.16 - I am not recording any more or less streams now then I was on 0.16 - once I upgraded to 0.17 this started to show up right away
2) I can't imagine that there is nearly enough data moving around to saturate the SATA bus
3) I see no other evidence of io issues from any other logs, subsystems, etc on the system
Any thoughts would be appreciated, Troy _______________________________________________ mythtv-users mailing list [email protected] http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-users
