RE: [Emerald] Work on SNMP Concurrency Checking

Duane Schaub ( (no email) )
Wed, 19 Apr 2000 08:33:17 -0500

I too see this - it has been suggested that the problem was that radius did
not recieve the stop packet. I personally feel that it is more of a problem
in that the stop packet is either not being generated or is fundamentally
different in some way so that a radiusnt or sql error occurs and the calls
table doesn't get updated. I haven't had the time to really debug this part
of the process, but it has rendendered login limits useless for three years.

SNMP may have been the solution for us in the past, but we are currently
moving to the SWBell VPOP on a Nortel CVX-1800 and to my knowledge, they
will not allow snmp traffic from a customer. They have a cool web page, so
it may be possible to write something that will parse the web page and
update the tables.... but again - I really don't have time to mess with


-----Original Message-----
[]On Behalf Of NCKCN
Sent: Tuesday, April 18, 2000 5:33 PM
Subject: Re: [Emerald] Work on SNMP Concurrency Checking


No. Probably 100 time a day users get "hung" on by the system and it still
thinks they are connected even though they have disconnected. The user
doesn't connect then on even though they are not accually connected. That is
what the SNMP Concurrency Check is supposed to correct.

I don't think it is the NAS as much as crappy modems. I have different NAS's
and they all seem to do it. I connect to any of my NASs without fail 100% of
the time, and disconnect 100% of the time. I have physically unplugged the
phone line, shut off the modem, and done just about any abusive thing I can
think of to get a connection to "hang" but I have never been able to
reproduce this error, yet I see a huge number of these "hangups" per day.
They are really a pain in the butt.

----- Original Message -----
From: "Randy Martin" <>
To: <>
Sent: Tuesday, April 18, 2000 5:22 PM
Subject: Re: [Emerald] Work on SNMP Concurrency Checking

> I don't mean to jump into the middle of this, but isn't the above the
> NON-SNMP concurrency doing it's job correctly? It did the radcheckonline
> and found the user already online. It doesn't seem to me that you should
> need both? I don't have the SNMP concurrency setup, and everything works
> just fine on our system. Is there only certain combinations of NAS's etc.
> that have a problem with the non-SNMP concurrency checking?

For more information about this list (including removal) go to:

For more information about this list (including removal) go to: