From: Sam.MicroGate@xxxxxxxxxxxxxx
Date: Mon Apr 22 2002 - 12:25:51 GMT-3
Nigel,
Getting the 172.16.80.0/25 and 172.16.80.128/25 is an issue and has a big
benefit. Suppose that router r3 is talking RIP V1
with another router (say r1) and router r1's loopback address is
172.16.70.1/24. r3 is distributing between OSPF and RIP.
Now you have two routing domains (RIP and IGRP) advertising the same major
network (172.16.0.0/16) And that is a routing loop.
Sam
-----Original Message-----
From: Nigel Taylor [mailto:nigel_taylor@hotmail.com]
Sent: Sunday, April 21, 2002 4:27 PM
To: aamercado31@yahoo.com; ccielab@groupstudy.com
Subject: Re: FLSM has longer mask than VLSM Probl.
Tony,
The question is "why exactly do you need the "172.16.80.0" route in
the r6 RIB? You should have the classfull "172.16.0.0" which should provide
connectivity(i.e ping, traceroute)to r8. What requirement/benifit would
this achieve. I would think the issue here is getting the
r6-r3(172.16.100.0/27) into the r8 RIB to achieve full connectivity. Now
getting this to happen would be interesting.. :->
Can you from r8 ping the s1 interface of r3? See Inline...
Nigel
>From: hong tony <aamercado31@yahoo.com>
>Reply-To: hong tony <aamercado31@yahoo.com>
>To: ccielab@groupstudy.com
>Subject: FLSM has longer mask than VLSM Probl.
>Date: Sun, 21 Apr 2002 12:06:33 -0700 (PDT)
>
>HI
>
>I search the archives and can't find an answer for
>this one.
>
>\lo0
>r8----r6-----r3
>/lo1
>
>r8/r6 is IGRP with r6 as the redistributing router for
>the OSPF on r6/r13
>
>r8 ip address:
>lo0 = 172.16.80.1/25
>lo1 = 172.16.80.129/25
>s0= 172.16.86.8/24
>
>r6 ip address:
>s0 = 172.16.86.6/24
>s1 = 172.16.100.6/27
>
>r3 ip address:
>s1 = 172.16.100.3/27
>
>My question is - How can I can the 172.16.80.0 network
>into r6 routing table?
>
>If I put "ip route 172.16.80.0 255.255.255.0 null0"
>the route would propagate to r6/r3. However, obviously
>I do not want to do statics...so here were my
>alternative attempts.
>
>1. default-network - Can't do it cuz of the classful
>nature of this command which would propagate a static
>route into r8.
>
>2. Summarizing - Nope cuz the IGRP (FLSM) has a longer
>mask than OSPF (VLSM)
NT: In a non-production enviroment you might consider disabling the "ip
classless and check out the effect on the longer FLSM you mentioned.
>
>3. secondary address - Because of the 172.16.80.x/25
>mask is using up all the subnetworks for 80.x, I don't
>have any other address to use for secondary
NT: the emphasis here would not be in using the x.x.80.x network but using
the same mask so that the interface will pass route info with the /25.
>
>4. tunnelling - same problem as #3
>
>5. policy routing - I can't see this as applicable
NT: I wouldn't rule that out. The issue here is all the connections are on
the 17.16.0.0 classfull boundary. OSPF won't have an issue with this but
IGRP will as you've noticed.
>
>Is this possible or am I stuck to the null 0 option.
>
>Thanks
>
>
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 10:58:16 GMT-3