Sorry to dig up an old topic, but I have some info to add to this.
I finally got annoyed enough with having to run mythfilldatabase manually that i dug around in the scripts to see what was going on. As far as I can tell, its from a bug in one of the scripts that causes it to mistakenly believe that an initial database fill is still running.
in /usr/pluto/bin/MythTvCommon.sh is function WaitUntilMythTvInitialRunXMinutesOld starting at line 149:
function WaitUntilMythTvInitialRunXMinutesOld {
STARTTIME=`date +%s`
if [ -r $FILLDBINITRUN -a -s $FILLDBINITRUN ] ; then
STARTTIME=$(cat $FILLDBINITRUN)
fi
This checks for a lockfile containing a timestamp left by the initial db fill script i assume, but what i dont understand is why it doesnt simply return control to the fill database script and let it continue if the lockfile isnt there. So I added a couple additional lines to do just that:
function WaitUntilMythTvInitialRunXMinutesOld {
STARTTIME=`date +%s`
if [ -r $FILLDBINITRUN -a -s $FILLDBINITRUN ] ; then
STARTTIME=$(cat $FILLDBINITRUN)
else
return
fi
So far it seems to do as it should, but I also have not given it much testing. I dont know what will happen if the initial db fill is still running.