All OSPF routes disappear when only one of the participating ports is shut.

Chris Wilken · ‎10-26-2016

Hi There.

I feel as though I have something fundamentally missing in terms of my OSPF knowledge.

I have an OSPF LAB runnig on Packet Tracer 7.0.0.0306.

I have setup basic OSPF on 4 routers in a ring and the routing is working great.

The problem I have is that on any of the 4 routers, when I shut one of the two participating interfaces all OSPF routes disappear from the routing table.

When I do #show ip ospf neighbor , I can still see the neighbor on the "un-shut" interface, but no routes in routing table.

As soon as "no shut" the interface, the routes reappear.

Also, while the one interface is shut and all OSPF routes have disappeared, if I do #clear ip ospf process, the routes from the neighbor on the un-shut interface reappear even with the other interface still shut.

See attached my typology and configs

Georg Pauwen · ‎10-26-2016

Hello,

at first glance, you advertise the same network on all router, 192.168.0.0 0.0.255.255.. Try to advertise the networks that are actually configured on the interfaces:

network 192.168.1.0 0.0.0.255 area 1

network 192.168.2.0 0.0.0.255 area 1

network 192.168.3.0 0.0.0.255 area 1

network 192.168.4.0 0.0.0.255 area 1

network 192.168.5.0 0.0.0.255 area 1

Chris Wilken · ‎10-26-2016

Hi, Thank you for the reply.

I changed the networks on all 4 routers. See attached.

Seems like the problem persists.

Georg Pauwen · ‎10-26-2016

Chris,

I recreated the exact same topology you have in Packet Tracer. When I shut down the link to R3, the routing table indeed shows just the connected routes. When the 'clear ip ospf process' is entered, election of a new DR and BDR occurs, after which the full routing table is exchanged. The DR is R2 now, which is reachable. That would make sense, since the OSPF database only installs the relevant routes in the routing table if the advertising router is reachable.

chrihussey · ‎10-26-2016

This may be the case with Packet Tracer, but for several reasons I don't think it would be the case in a real environment.

Please understand, I'm not disagreeing with your assessment, and in the effort to learn can the connections be defined as point to point with Packet Tracer so DRs and BDRs would not be an issue?

Chris - I'll drop it if you are satisfied.

Georg Pauwen · ‎10-26-2016

Chrihussey, Chris,

the 'ip ospf network-type' is unfortunately not available in Packet Tracer. I agree that Packet Tracer has functional limitations, and is probably a good learning tool, nothing more, nothing less.

chrihussey · ‎10-26-2016

A valid point and something to consider.Thanks.

I'm on a project and expect to get my hands on some routers in the next few months. If possible I'll test this scenario and post the results to this discussion.

chrihussey · ‎10-26-2016

This wouldn't be the expected behavior and just taking a guess here. You are using 2800s and it might just be a code thing, but there is a chance that by shutting an interface you're causing the router's OSPF ID to be nullified, thus restarting the process establishes the router with a new ID and gets things working again.

By default OSPF uses the highest loopback IP, and in the absence of that, the highest interface IP as its OSPF router ID.

You could configure loopback IPs on the routers and restart the OSPF process so that they assume these as the router IDs, or simply define the router ID in the global OSPF (router-id x.x.x.x) process using one of the existing interfaces (or the loopbacks) to see if this changes things.

Chris Wilken · ‎10-26-2016

Hi chrihussey

I set up loopback addresses on all 4 routers

R1 - loopback address 10.0.0.1 255.255.255.0

R2 - loopback address 10.0.0.2 255.255.255.0

R3 - loopback address 10.0.0.3 255.255.255.0

R4 - loopback address 10.0.0.4 255.255.255.0

The problem persists.

R1#sh ip route

10.0.0.0/24 is subnetted, 1 subnets

C 10.0.0.0 is directly connected, Loopback0

C 192.168.1.0/24 is directly connected, FastEthernet0/1

O 192.168.2.0/24 [110/2] via 192.168.1.2, 00:04:50, FastEthernet0/1

O 192.168.3.0/24 [110/2] via 192.168.4.1, 00:01:23, FastEthernet0/0

C 192.168.4.0/24 is directly connected, FastEthernet0/0

R1#configure terminal

Enter configuration commands, one per line. End with CNTL/Z.

R1(config)#interface FastEthernet0/0

R1(config-if)#shutdown

R1(config-if)#

%LINK-5-CHANGED: Interface FastEthernet0/0, changed state to administratively down

%LINEPROTO-5-UPDOWN: Line protocol on Interface FastEthernet0/0, changed state to down

00:14:04: %OSPF-5-ADJCHG: Process 1, Nbr 192.168.4.1 on FastEthernet0/0 from FULL to DOWN, Neighbor Down: Interface down or detached

R1#sh ip route

10.0.0.0/24 is subnetted, 1 subnets

C 10.0.0.0 is directly connected, Loopback0

C 192.168.1.0/24 is directly connected, FastEthernet0/1

chrihussey · ‎10-26-2016

When you added the loopback IPs did you restart the OSPF process on each router and then see the loopback IPs as the router IDs when looking at the "sh ip ospf neighbors"?