My m0n0wall runs longer than 20 minutes before it flakes out and I have been
remote syslogging for several weeks...with all syslog output options
enabled.

FWIW, I'm running 1.2b8 (cd iso)...HTTPS management gui/HTTPS user auth
(with FreeRadius).

I have noticed this particular bug twice now.  On the first occassion port
8001 died outright.  On the second occassion port 8001 was alive but the
redirect never kicked in to take the user to the login screen.  After the
user reported it I logged myself out of the portal and tried to re-establish
a connection...no dice.  Please note I am running the captive portal without
the logout popup.

I really wish I could give you better feedback but the logging capabilities
are just not there, IMo, to give a full detailed report.  

The only thing I can remotely associate it with is user TIMEOUT associated 
with logportalauth.  One of the users helping me test it noticed the bug
just after he was timed out of his session.  

I've got a good amount of time to spend on testing and will try to recreate
the problem from that slant but please let me know how to recreate it if you
can Ugo. 

Respectfully,
John Burkhart


Matt Juszczak wrote:
> The new m0n0wall beta runs for 20 minutes for us and then crashes, stops
> responding to authentication.  We can still ping it, but it seems that
> possibly the webserver is dieing or some other process involved in the
> authentication procedures.
> 
> We brought our 1.2b2 box back up again and it works fine now.
> 

You should activate the remote syslog to see what it is logging.

> Any ideas?
> 
> Thanks!  Regards,
> 
> Matt