[Dnsmasq-discuss] dnsmasq won't quit if listening on 'empty' bridge?

[dnsmasq at lists.bod.org]

Yes, run as root as part of the 'runparts cron.daily' that's standard on 
Fedora.

I know that a simple delay won't help. Even several hours later, the 
original instance hasn't fully quit, and is still listening on virbr0 
(according to netstat). To get dnsmasq working again, I have to manually 
kill the original pid, and start dnsmasq again.

I've never previously had any problem with restarting dnsmasq, and even 
now it works just fine when I manually restart it. It only chokes when 
it's run unattended by cron at ~4am local time (Pacific). I've certainly 
never needed to go to the lengths of hitting it repeatedly over the head 
with SIGTERM...

-- Paul

gypsy at iswest.com wrote:
> Quoting Paul Chambers <dnsmasq at lists.bod.org>:
>   
>> The strangest part is that I can't recreate this by typing the same 
>> command into the shell as is used by the script (service dnsmasq 
>> restart). Nor by running the script itself manually. It only happens 
>> when cron runs the script.
>>     
>
> The cron job is run by root?
>
> I have to put a 10 second delay between the stop and start, something like
> this:
>   sleep 1
>   killall dnsmasq       2>/dev/null
>   echo -n " 9 "
>   sleep 1
>   killall dnsmasq       2>/dev/null
>   echo -n "8 "
> --
> gypsy
>
> _______________________________________________
> Dnsmasq-discuss mailing list
> Dnsmasq-discuss at lists.thekelleys.org.uk
> http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
>