Thanks Hugh -
We're not enforcing any simultaneous use limits -- I couldn't think of a
better word to describe sessions for 'packets'. The following seemed to
happen:
16:37:13 -> Auth-request (Identifier 225)
16:37:13 <- Access-Accept (Identifier 225)
16:37:13 -> Accounting-Request [start] (Identifier 218)
LOG ENTRY: "INFO: Duplicate request id 218 received from xx.xxx.xxx.xx:
ignored
.............
Only entry in the log prior to this was at 16:01:23 - an Accounting Request
(Identifier 218) for a stop packet for a different user.
.........................................................................
Mark Mackay,
Network Coordinator,
Orcon Internet.
> On Mon, 13 Nov 2000, Mark - Orcon Support wrote:
>> Had a problem where a start session of a customer was ignored because a stop
>> packet using the same identifier 25 minutes earlier.
>>
>> Thus I was wondering what time has to elapse between session packets before
>> it's classed as a new session? And is there any way to control this?
>>
>
> Could you give me a bit more detail please?
>
> A new session should never be ignored - the only thing that can happen is a
> rejection due to simultaneous use limits being incorrectly enforced because a
> stop record was missed.
>
> The way to control this is with the NasType parameter in the Client clause to
> have Radiator query the NAS to verify the existence of suspect sessions.
>
> hth
>
> Hugh
===
Archive at http://www.starport.net/~radiator/
Announcements on [EMAIL PROTECTED]
To unsubscribe, email '[EMAIL PROTECTED]' with
'unsubscribe radiator' in the body of the message.
