Hello Paul -
On Thu, 30 Sep 1999, Paul Thornton wrote:
> Hi Matthew / Mike,
>
> Can you please explain why we are starting to get this problem.
>
> schliebs 4 Start Tue Sep 28 16:52:16 1999
> schliebs 4 Stop Tue Sep 28 16:53:16 1999 User-Request
> schliebs 4 Start Tue Sep 28 16:53:47 1999
> schliebs 4 Start Tue Sep 28 16:55:17 1999
>
> As you can see from the above record, this user had a Start and Stop
> record a minute apart and then shortly after we received 2 more Start
> records from them? This causes our Accounting Server to re-add them to our
> online database which then continues to deduct their time.
>
This is curious, as the Starts are 90 seconds apart - I wonder if you lost a
Stop packet in between and the User has dialled up again? Or does this User run
multiple channels? If you are losing packets, you should check your network
performance.
You can also set up a SessionDatabase with strict checking, which will cause
Radiator to check with the NAS to verify if the User is still logged in (or
not).
> The have connected to a PM2 but this also happens on the PM3's regardless
> of where they have connected.
>
> Mike: If I create a dupinterval of say 180 would this have failed to pick
> up the Stop record above, considering it is only 1 minute apart.
>
The DupInterval parameter will only affect duplicate packets of the same type
(ie. Accounting Starts or Accounting Stops) so it will have no effect on a
Start followed closely by a Stop.
> Matthew: I know that on the Cisco's there is an option for setting some
> form of radius variable to restrict multiple instances of this aswell. Do
> the PortMasters have this as an option aswell?
Don't know about this.
hth
Hugh
--
Radiator: the most portable, flexible and configurable RADIUS server
anywhere. SQL, proxy, DBM, files, LDAP, NIS+, password, NT, Emerald,
Platypus, Freeside, TACACS+, PAM, external, etc etc on Unix, Win95/8,
NT, Rhapsody
===
Archive at http://www.thesite.com.au/~radiator/
To unsubscribe, email '[EMAIL PROTECTED]' with
'unsubscribe radiator' in the body of the message.
