From: Brian Dennis (brian@5g.net)
Date: Wed Mar 12 2003 - 04:49:39 GMT-3
Without the configurations it's hard to tell what the exact problem is
but a possible problem is with RIP telling OSPF about the status of the
ISDN subnet through redistribution. Try taking off the redistribution of
RIP into OSPF and see if the link stays quiet.
If the link does stay quiet after redistribution is removed, you should
look into using a route-map and filter off the ISDN subnet from being
redistributed into OSPF from RIP. Basically the problem is that every
time the ISDN link drops RIP informs OSPF through redistribution that
the ISDN subnet isn't available. Since this causes a change in the OSPF
database, OSPF has to inform its neighbors. This means that a call
across the demand circuit has to be placed to notify the OSPF neighbor
on the other end. Once the link comes back up for OSPF to notify its
neighbor RIP informs OSPF that the ISDN subnet is again available. All
is fine till the idle-timeout is reached for the ISDN link and the link
drops again. This isn't a problem for OSPF because this is a demand
circuit but now that the ISDN link is down RIP again informs OSPF that
the ISDN subnet isn't available. This of course causes another change in
the OSPF database and another call must place to inform the OSFP
neighbor about the change. This is just a never ending circle as you can
see.
You should also look into filtering off any /32 host routes that might
have been installed via PPP from being redistributed into OSPF from RIP.
Since RIP also supports host routes, RIP will redistribute them into
OSPF. You can use the "no peer neighbor-route" command if it's not
already on to stop the /32 host routes from being generated or just
filter off the /32 in the route-map used for redistribution.
In the future if you ever have to configure an OSPF demand circuit on an
ASBR watch out for the problems described above ;-)
Brian Dennis, CCIE #2210 (R&S/ISP Dial/Security) CCSI# 98640
brian@labforge.com
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
tria_ka@gmx.net
Sent: Tuesday, March 11, 2003 10:30 PM
To: OhioHondo
Cc: jhays@jtan.com; ccielab@groupstudy.com
Subject: RE: isdn - backup - challenge - authentication ???
hi,
> Can it be that changes in the OSPF topology have occured?
yes that is also what i understand but i don't know why there is a
change
every time.
if i have a look on ip ospf database normally it's really stable. when i
configure bri on both sides for isdn the link comes up changes lsa, the
go
down....and up again....every time...no stopping !!..;-(
> When using OSPF on demand, it is my understanding that the ISDN link
comes
> up initially and learns the network.
>
> I also have read that every time a change in the topology is sensed,
the
> ISDN circuit comes up again to keep itself current with an accurate
> topology. (When the primary goes down, network operation depends on
the
> ISDN
> routes to be an accurate view of the network. Imagine a primary that
> hasn't
> gone down in months. If the ISDN didn't keep updated, it would be
> worthless
> as a backup that is meant to give an IMMEDIATE alternate route when
the
> primary goes down.)
>
> I have also read that OSPF On Demand is not recommended in parts of
the
> OSPF
> network where changes occur frequently --- because the ISDN link will
keep
> going up and down. It is recommended that it be used on Stub links.
uuppss that is what i don't try. i also heard there is sometimes a
problem
with virtual links ? so there is also a virtual link !!
keep it the isdn line up ?
thanks
udo
>
> -----Original Message-----
> From: nobody@groupstudy.com [mailto:nobody@groupstudy.com]On Behalf Of
> Jonathan V Hays
> Sent: Tuesday, March 11, 2003 2:21 PM
> To: ccielab@groupstudy.com
> Subject: RE: isdn - backup - challenge - authentication ???
>
>
> > -----Original Message-----
> > From: tria_ka@gmx.net [mailto:tria_ka@gmx.net]
> > Sent: Tuesday, March 11, 2003 1:47 PM
> > To: Jonathan V Hays
> > Cc: ccielab@groupstudy.com
> > Subject: RE: isdn - backup - challenge - authentication ???
> > Hi Jonathan,
> > thanks. That is what I also know ..,;-)) I can read the debug
> > output...
> > But I don't understand WHY this MC packet trigger my isdn
> > line to go up.
> > Both isdn bri if's are in the ospf state 'running on demand' !
> > any comments ? help ??
> > thanks
> > udo
> >
> > > The dial reason is 224.0.0.5 which is one of the OSPF multicast
> > > addresses.
>
> Udo,
>
> Possibly something is not allowing a complete OSPF database
> sychronization between the two routers. Or maybe it's OSPF
> authentication that hasn't been configured across a virtual link. It's
> hard to tell with what little information you have provided.
>
> Please post your configuration files for both routers.
>
-- +++ GMX - Mail, Messaging & more http://www.gmx.net +++ Bitte ldcheln! Fotogalerie online mit GMX ohne eigene Homepage!
This archive was generated by hypermail 2.1.4 : Sat Apr 05 2003 - 08:51:37 GMT-3