[mythtv-users] excessive crashing of 0.22 backend

[emarkert at aim.com]

 I've had a similar issue, but with the frontend. I recently upgraded to 0.22 and frequently use a frontend on the master backend to watch recordings.  My frontend constantly dies with a segmentation fault.

Like you, my -v all debug messages didn't produce anything that indicated an error.  However, I ran

dmesg | grep -i myth

and found that the frontend code appears to be using QT3 code instead of the requisite QT4 - even though I have both QT 3 & 4 installed.

 


 I haven't, yet, tried running a 0.21 frontend from my other computer to see if that causes a crash, but I suspect that it will cause some issues.

Though my backend has not had a problem, and continues to operate correctly, I have noticed that since upgrading that I've also had issues with mythweb not being able to stream and not updating the screen when I delete a recording.

My 0.21 system ran like a champ, since upgrading it's been interesting.  Actually considering downgrading to 0.21 because I haven't found any solutions, or clues of where to look further.

 

-----Original Message-----
From: Ronald Frazier <ron at ronfrazier.net>
To: Discussion about mythtv <mythtv-users at mythtv.org>
Sent: Thu, Jan 7, 2010 7:54 pm
Subject: [mythtv-users] excessive crashing of 0.22 backend


I recently upgraded my system to 0.22-fixes, and ever since then the
backend has crashed a remarkable number of times. Looking over my logs
of how often the backend has crashed, before the 0.22 upgrade it
crashed 7 times in 18 months. Now, since the upgrade it has crashed 20
times in 8 days.

Comparing to my frontend logs, I can see that all but 1 of them
happened within seconds of starting playback on the frontend. Some of
those were recordings, some were videos from an NFS mount, and some
were videos from storage groups. I have had a lot of times where I go
to watch a recording and it kicks me back to the watch recording
screen with an error that it can't connect to the backend.

My frontends are running on separate systems than the backend. What's
interesting is that there were a few crashes playing videos from the
NFS mounted drive. Those should have had minimal interaction with the
backend. I do have the "always stream" option enabled, so perhaps
that's where the problem lies (though for one of the times, the
frontend was playing from NFS rather than a storage group, and as far
as I can tell it didn't actually try streaming and seemed to play
right through the crash). I've since turned that option off to see if
it makes a difference. I had another backend crash today, but it was
while playing from a storage group, which even though the path is
available via NFS on the frontend, I think it still streams it from
the backend anyway (right?).

Any ideas what could be my problem or how to go about debugging? I
turned on "-v all" for the backend log. I've got a a few logs from
that. I'm still going through them (they're a bit messy, with
recording and commflagging going while the crash happened) but I don't
see any obvious problems so far.


-- 
Ron
_______________________________________________
mythtv-users mailing list
mythtv-users at mythtv.org
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-users

 
=
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mythtv.org/pipermail/mythtv-users/attachments/20100107/c8b11d48/attachment.htm>