[Wifidog] WifiDog has a bug

[Alexandre Carmel-Veilleux]

On Sat, Oct 02, 2004 at 05:26:51PM -0400, Benoit Grégoire wrote:
> 
> Why don't you just log to a remote syslog server?

	Do we have one? I would love to log to a remote ile sans fils
log server. It'd be great. And it'd take care of half of the stuff I
want.

	The other half is to have a startup script that notices that
wifidog has stopped and restart it. Combined with an existing log trail,
it'd be possible to diagnose problems and to minimize adverse impact on
the users.

	My latest thinking on this has been:

1) Implement a form of logging in WifiDog that would save to a number of
   log files, automatically rotate every time a watermark is reached and
   automatically clean the oldest log file when a second watermark is
   reached.
2) Modify wdctl to have a "start" command to start wifidog, like apachectl
   start. In addition, wdctl would fork a copy of itself in the background
   and call wait() on its child process.
3) When wifidog exits abnormally, wdctl automatically upload the log files
   to the central server and then cleans up logging and start wifidog
   again.
4) When wdctl stop is called a file /var/run/stop-wifidog is created and
   when wifidog exits, the monitoring wdctl process sees that file, deletes
   it and then exits.

	I think this would work. If we use syslog, set 1 is deleted and
step 3 is amended accordingly.

Alex

_______________________________________________
Wifidog mailing list
Wifidog at isf.waglo.com
http://isf.waglo.com/mailman/listinfo/wifidog_isf.waglo.com