[mythtv-users] Mythfrontend and backend crashes on yesterday's CVS

[nowhere]

Hi everyone,

Chances are I already know the answer but I would like some
confirmation. The mythfrontend is crashing very very frequently,
dropping out to the X desktop when I try to do many different things
like start a video, delete videos, use the skip function while watching
a video, starting LiveTV, calling up the OSG etc.

Quite frequently I will have the backend crash too in what I believe is
association with the frontend crashes but sometimes I walk up and find
the back end dead (or not responding and needing a restart).

There are no errors in dmesg, or /var/log/messages and the backend logs
show no errors either, but yesterday's crash has a huge hole in the log
with no entries from about 1am to 7pm on 11-02. 

I did notice that in the messages file at 7:06am (during that gap above)
I find the following...

Nov  2 07:06:53 localhost kernel: hda: dma_intr: status=0x51 {
DriveReady SeekComplete Error }
Nov  2 07:06:53 localhost kernel: hda: dma_intr: error=0x40 {
UncorrectableError }, LBAsect=2158733, high=0, low=2158733,
sector=2158733
Nov  2 07:06:53 localhost kernel: end_request: I/O error, dev hda,
sector 2158733
Nov  2 07:06:57 localhost kernel: hda: dma_intr: status=0x51 {
DriveReady SeekComplete Error }
Nov  2 07:06:57 localhost kernel: hda: dma_intr: error=0x40 {
UncorrectableError }, LBAsect=2158733, high=0, low=2158733,
sector=2158733
Nov  2 07:06:57 localhost kernel: end_request: I/O error, dev hda,
sector 2158733

So. I'm betting I know what this means but would like to hear it from
someone else before I spend more money. Keep in mind we had a power
failure of some kind last month that trashed my filesystem requiring me
to format and start over with the root filesystem (my /video partition
survived that).

I was suspecting that the hack of a script I wrote to re-insert all the
videos that survived the crash back into mysql was causing the
instability but I have since watched and deleted them all and have found
the had error in the logs now.

Please let me know what you all think...

Thanks,
Eric