RE: (PM) Idle-Timeout = 0 overrides port setting(?) (fwd)

Brunette, Jason (jbrunette@excel.net)
Sun, 21 Mar 1999 12:02:16 -0600

> -----Original Message-----
> From: MegaZone [mailto:megazone@megazone.org]
> Sent: Friday, March 19, 1999 7:18 PM
> To: portmaster-users@livingston.com
> Subject: (PM) Idle-Timeout = 0 overrides port setting(?) (fwd)
>
> Once upon a time Brunette, Jason shaped the electrons to say...
> >Does an Idle-Timeout of 0 in a user's radius entry override
> a PM3's idle
> >port setting? I thought that it did not. I found this message from
>
> It shouldn't. If it does then that's a change.

I think it changed somewhere...

> Well, this doesn't really prove anything since folks have
> reported trouble
> with the idle timer in the past - users staying idle longer than the
> port idle timer even with no Idle-Timeout from RADIUS.

I thought of this. I ptraced the port for a while...there was definately no
idle-timer-resetting traffic going over it. There were LCP_ECHO_REQUESTs
and _REPLYs but they don't count.

> >Bug/feature? What is it really supposed to do? I think any
> Idle-Timeout
>
> It isn't supposed to. 0 values returned from RADIUS is
> normally treated
> as if that ACP had not been sent by the RADIUS server -
> telling the NAS
> to default. While I'd *like* it to override things, that isn't how it
> was designed.
>
> -MZ

Could someone else check this? This isn't critical, but I've only had
people reply "it can't" or "it shouldn't" when it does appear to ignore the
idle timeout set on the port.

============================================================================
Jason Brunette Excel.Net, Inc. Phone: (920) 452-0455
jbrunette@excel.net "Your excellent Internet provider" Fax: (920) 452-8118
http://www.excel.net/ Toll Free: (888) 489-9995
----------------------------------------------------------------------------
-
To unsubscribe, email 'majordomo@livingston.com' with
'unsubscribe portmaster-users' in the body of the message.
Searchable list archive: <URL:http://www.livingston.com/Tech/archive/>