> > I would bet your disks are plenty fast and you have some sort of > software / driver issue, but perhaps the below will help you figure it > out. Keep in mind that your recordings are at most a few megabytes per > second and a single disk should be able to do at least 20-30 MB/sec. > > Are your two lvm disks configured as a stripe (raid0) or just a plain > concatenation? Stripes are obviously faster as blocks read/written are > more balanced across the physical devices. See lvcreate(8), -i option. > > For optimal performance, the disks should be on separate IDE cables (but > you say they're SATA, so this doesn't apply to you, but maybe someone > will find this tip useful). > > While 99% of the time it's not the problem, double-check that DMA is > enabled on the disks using 'hdparm' (forgive me if that doesn't apply to > SATA; I have yet to use one). > > Are there any syslog messages or mythbackendlog messages? If you > carefully correlate the time you see dropped frames in your recording > (look at the current position with "i") to the times in the syslog, you > might be able to find the appropriate area in the logs to look for > abnormal messages.
Jan 29 11:39:14 localhost kernel: ivtv2 warning: ENC Stream 0 OVERFLOW #2140: Stealing a Buffer, 512 currently allocated Jan 29 11:39:14 localhost kernel: ivtv2 warning: ENC Stream 0 OVERFLOW #2141: Stealing a Buffer, 512 currently allocated Jan 29 11:39:14 localhost kernel: ivtv1 warning: ENC Stream 0 OVERFLOW #703: Stealing a Buffer, 512 currently allocated Jan 29 11:39:14 localhost kernel: ivtv1 warning: ENC Stream 0 OVERFLOW #704: Stealing a Buffer, 512 currently allocated and that keeps on going. if i schedule 3 recordings load goes to the sky. It doesn't crash. With or without the nvidia module loaded. thanks -- Met vriendelijke groeten/kind regards, Chris Holleman _______________________________________________ ivtv-users mailing list [email protected] http://ivtvdriver.org/mailman/listinfo/ivtv-users
