XFBB archive - November 1999: Re: [xfbb] NbCh Problem with xfbb?

Re: [xfbb] NbCh Problem with xfbb?

Arno Verhoeven (terhi.victor@logonet.com)
Tue, 30 Nov 1999 13:52:08 +0100

Steve Fraser wrote:
> I've been unsuccessful trying to provoke it deliberately. However, both
> instances that I have investigated in detail had the number of users has
> exceeded the NbCh value. I dont know if this is a cause or a symptom though.
> However, most users on at the time had data in the receive queues,
> suggesting that the xfbb task had not serviced them - possibly because it is
> busy getting erroneous indications from the kernel!
>
> The system has run for 8-9 days between instances of the problem, so much
> "waiting" is required to do any tests.

I'm not sure if your problems with disconnects are the same as the
problem that I had some time ago.

Our station's main function was to act as BBS. It did not act as a node.
We found that users were disconnected from the BBS at the moment that
the BBS itself started to forward to a neighbour BBS. (in our case every
20 minutes) They could reconnect immediately. If they then issue the !
command, they could see that they BBS thought that the old connection
was still active.

Just set your BBS to a forward interval to 5 minutes and see if you get
disconnected every 5 minnutes.

If this is the case then the sollution (or should I say, work-around) is
to define a seperate port for outgoing forwards.
In our case all incoming connections are on port 1, and outgoing
connects are on port 2.

73, Arno pe1icq

_______________________________________________
xfbb maillist - gzlm@trasys.be
http://www.f6fbb.org/mailman/listinfo/xfbb