I spent a bit more time looking into this one than it may be worth. But my look did reinforce some points made in this thread and in another thread started by John Neiberger and researched so ably by Nigel Taylor - that is, the nature and behaviour of secondary addresses.
Sorry I am unable to document everything I did here. It would take me writing a Jeff Doyle type chapter on RIP to get it all out and explained, with screen shots etc. To put things in terms of how I observed them: In the case of RIP, by default, advertisements are sent out an interface using the primary address of that interface as the source address. if another router on the segment is using and address that is not on the same subnet as the primary, that router will see messages like this: 01:46:25: RIP: ignored v1 update from bad source 172.29.101.1 on TokenRing0 01:46:30: RIP: ignored v1 update from bad source 172.29.101.2 on TokenRing0 01:46:35: RIP: sending v1 update to 255.255.255.255 via TokenRing0 (172.29.103.7) 103.1 was secondary address on my R1, 103.7 the address of my R3 You can see the error referring to 101.1 and 102.1 ( the address of another router on the segment ) I threw in a no ip split-horizon command on the interface of my R1, and lo and behold, it started sourcing rip packets from 101.1, 102.1 and 103.1 and all my RIP routes propagated from CCO: Note If any router on a network segment uses a secondary address, all other routers on that same segment must also use a secondary address from the same network or subnet. some of us already commented about issues with secondary routes among the various routing protocols. the point being that using secondary addresses can be tricky, and is probably not a good idea for newbies just trying to learn the basics. if you want to see how things work, use loopbacks. with secondary addresses, it is to easy to end up fighting with some complex issues beyond a beginner's understanding. in fact, there are some advanced students who find this topic complex and mysterious. best wishes. Chuck BTW, one of the implications of this study was a walk down memory lance. A guy named Bob Vance who used to hang here a lot and who was the progenitor of a number of interesting discussions once postulated that all stations on a segment will see the all F's broadcast, even if their layer three addresses are different ( i.e. seconday's ) the output above is something of a proof of that supposition. The router saw the RIP packets with the destination address of 255.255.255.255 ( MAC FFFF.FFFF.FFFF ), processed the packet, saw the source address as being on a different subnet ( even though on the same segment ) and rejected the packet. Interesting. Especially in that all subnets were part of the same Class B network. -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of Logan, Harold Sent: Friday, December 07, 2001 6:21 AM To: [EMAIL PROTECTED] Subject: RE: RIP routing (2 router lab) newbie [7:28327] It looks like Anil wants to get RIP to advertise the 193.9.200.0 network. A secondary address may work on one of the interfaces, but it would need to be on a different subnet. Notice from the config, he gave the secondary address the same IP as the primary addy. No matter what he does with the 193.9.200.0 network, those two routers will always show it as being "Directly Connected" instead of learned through RIP; DC routes have an administrative distance of 0, whereas RIP has an AD of 120. In the routing table, the router is only going o show the route with the best (lowest) distance. He could add a loopback on a different subnet on one of the routers, then add network statements for that subnet, and then he would see that network learned via RIP on the opposite router. Likewise Anil, if you had a 3rd router connecting to one of your two routers by the BRI port, that 3rd router would learn of the 193.9.200.0 network through RIP. (Granted, RIP wouldn't be your ideal routing protocol for an ISDN line, but that's going a little bit deeper than you need to for now) Try these configs, then look at your routing tables: > hostname rustya > ! > enable secret 5 $1$Ws8V$mRIwI97bc/Iv7PAEKFBVo1 > ! > interface Loopback0 > ip address 200.10.10.1 255.255.255.0 > ! > interface Ethernet0 > ip address 192.9.200.1 255.255.255.0 > ! > interface BRI0 > no ip address > shutdown > ! > router rip > network 193.9.200.0 > network 200.10.10.0 > ! > no ip classless > ! > line con 0 > line vty 0 4 > password cisco > login > ! > end > hostname rustyb > ! > enable secret 5 $1$JycL$W4sNa8kuL2.tppX2IYQJU/ > ! > interface Loopback0 > ip address 201.10.10.1 255.255.255.0 > ! > interface Ethernet0 > ip address 192.9.200.2 255.255.255.0 > ! > interface BRI0 > no ip address > shutdown > ! > router rip > network 196.9.200.0 > network 201.10.10.0 > ! > no ip classless > ! > line con 0 > line vty 0 4 > password cisco > login > ! > end hth, Hal > -----Original Message----- > From: Kane, Christopher A. [mailto:[EMAIL PROTECTED]] > Sent: Thursday, December 06, 2001 9:26 PM > To: [EMAIL PROTECTED] > Subject: RE: RIP routing (2 router lab) newbie [7:28327] > > > I'm not sure if RIP is the same as OSPF, but if so, you must have OSPF > running on the interface via the Primary address in order to have the > Secondary address also participate in OSPF. Did you try > adding your Primary > address (network statement) to RIP also? > > Chris > > > > -----Original Message----- > From: Chuck Larrieu > To: [EMAIL PROTECTED] > Sent: 12/6/01 5:45 PM > Subject: RE: RIP routing (2 router lab) newbie [7:28327] > > try using loopbacks instead of secondaries. Are your > secondary addresses > part of the RIP process via network statements? same subnet > boundary as > the > primary address? > > Chuck > > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of > anil > Sent: Thursday, December 06, 2001 2:03 PM > To: [EMAIL PROTECTED] > Subject: RIP routing (2 router lab) newbie [7:28327] > > > 2 C1603's connected to a hub. > It should be simple but I cannot see why RIP does not > update the routing tables (in either direction). > I added secondary addresses to both routers e0, and want RIP to send > these > routes. > > Any comments/suggestions welcome. > Many thanks > -Anil > > > --------- > rustya#sh ip route > > Gateway of last resort is not set > C 193.9.200.0/24 is directly connected, Ethernet0 > C 192.9.200.0/24 is directly connected, Ethernet0 > rustya# > ------------------------------ > hostname rustya > ! > enable secret 5 $1$Ws8V$mRIwI97bc/Iv7PAEKFBVo1 > ! > ! > interface Ethernet0 > ip address 193.9.200.1 255.255.255.0 secondary > ip address 192.9.200.1 255.255.255.0 > ! > interface BRI0 > no ip address > shutdown > ! > router rip > network 193.9.200.0 > ! > no ip classless > ! > line con 0 > line vty 0 4 > password cisco > login > ! > end > > > > -------------------------------- > rustyb#sh ip route > > Gateway of last resort is not set > > C 192.9.200.0/24 is directly connected, Ethernet0 > C 196.9.200.0/24 is directly connected, Ethernet0 > rustyb# > -------------------------------- > hostname rustyb > ! > enable secret 5 $1$JycL$W4sNa8kuL2.tppX2IYQJU/ > ! > ! > interface Ethernet0 > ip address 196.9.200.1 255.255.255.0 secondary > ip address 192.9.200.2 255.255.255.0 > ! > interface BRI0 > no ip address > shutdown > ! > router rip > network 196.9.200.0 > ! > no ip classless > ! > line con 0 > line vty 0 4 > password cisco > login > ! > end Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=28632&t=28327 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
