[mythtv-commits] Ticket #5278: Firewire Uses Bad Port Number and Fails
MythTV
mythtv at cvs.mythtv.org
Tue Jan 27 17:27:46 UTC 2009
#5278: Firewire Uses Bad Port Number and Fails
---------------------------------+------------------------------------------
Reporter: scottadmi at gmail.com | Owner:
Type: defect | Status: new
Priority: minor | Milestone: unknown
Component: mythtv | Version: 0.21
Severity: medium | Resolution:
Mlocked: 0 |
---------------------------------+------------------------------------------
Comment(by foom at fuhm.net):
Replying to [comment:17 anonymous]:
> I very hacked work around which has worked consistently for me was to
write a logsurfer script which basically monitors the backend for the
culprit error messages (there are two variations of this failure) and
reset the backend so at least recordings will try to resume. It doesn't
fix the problem, but it does help to catch recordings that would otherwise
be blank.
Me too. I use this script to run mythbackend for the last 4 months.
{{{
#!/bin/sh
sleep 3
echo -n "******* Restarting mythbackend at: " >>
/home/mythtv/logs/mythbackend.log
date >> /home/mythtv/logs/mythbackend.log
# restart mythbackend if it gets certain errors. and exit if a new
mythbackend is started.
tail -f -c0 /home/mythtv/logs/mythbackend.log| \
awk '/Error: Unable to get handle for port: -1/ {system("killall
mythbackend"); exit} '\
'/^\*\*\*\*\*\*\* Restarting mythbackend at:/ {exit}' &
/usr/bin/mythbackend --pidfile /var/run/mythtv/mythbackend.pid >>
/home/mythtv/logs/mythbackend.log
}}}
I also have a line:
{{{
m2:2:respawn:/usr/local/sbin/start-mythbackend
}}}
in /etc/inittab to monitor mythbackend and restart it if anything happens
(either it crashes by itself or I explicitly kill it, above)
It's been working reliably with that.
--
Ticket URL: <http://svn.mythtv.org/trac/ticket/5278#comment:18>
MythTV <http://www.mythtv.org/>
MythTV
More information about the mythtv-commits
mailing list