OSPF Troubleshooting

JUNKSA_2 · ‎09-15-2007

Dear Expert,

I am doing a migration between the 3660 routers and 7609 router.

I have an existing ppp multilink network, Main Headquarter and around 22 Remote Sites using 3660 routers and an ospf routing protocol.

When we are trying to migrate the main headquarter from 3660 routers to the new 7609 router, Its seems okay, it advertised the multilink address, serial interface address etc. but one thing I noticed is the fastethernet IP address is not advertised which is very important to our data servers (e-mail, servers, active directory etc.)

Any idea how can I fix this issue, since the fastethernet IP address in the router is not advertising, I put it back again to the old 3660 routers to access to their data.

Your input is highly appreciated.

Thanks

paul.matthews · ‎09-15-2007

What would help would be a copy of the config you are using so that we can see what is hapenning, in terms of troubleshooting, the first thing to do is check the remote router for an LSA for the route - if both routers are in the same area, all networks should have an LSA present in the LSA database. If it is not there you need to look at why it is not there, If it is there you need to look at why it may not be going to the routing table.

If the LSA is not there, send us the output of "sh ip ospf int" you then need to carefully check the network statements to make sure the interface is covered by a network statement. In reality as that is probably the most likely cause I would look at that first.

JUNKSA_2 · ‎09-15-2007

Dear Paul,

I will send you the config tomorrow.

Thanks

Richard Burts · ‎09-16-2007

Jhun

In addition to the configuration it would be helpful if you would post the output of show ip interface brief and of show ip ospf interface.

It would also be helpful to know when you were attempting to use the new router if the router was able to ping (or any other kind of access) to the servers on the subnet of the FastEthernet interface.

HTH

Rick

HTH

Rick

JUNKSA_2 · ‎09-17-2007

Dear Paul and Rick,

Guys, thank you. anyway I will attached the required config you requested.

To brief you, The rest of the remote sites was still connected to the existing 3660 router in the main headquarter, the 3660 router will be replaced with 7609 router as part of the migration.

I cannot disconnect the remote routers unless the 7609 router will work properly. Before I tried to disconnect all the interface from the old router and moved it to the new router (7609) but I failed, as I have said before I can ping all the interface, multilink, serial interface, tunneling but not the fastethernet IP which is very important for the administrator that managing the servers.

To avoid this issue, I connect only one remote site named Jizan (3660 router) was connected via E1 (G.703) to Riyadh (7609 router) which is also located in the main headquarter to start as the pilot testing site.

I will attached the config and other results that you requested.

I hope it will help and thank you for your time.

Note: Guys can you give me your e-mail add, I cannot send the information you requested.

cheers....

BR/Jun

paul.matthews · ‎09-17-2007

When you reply, next to the post button is a link to add attachments.

P.

JUNKSA_2 · ‎09-17-2007

Dear Paul,

See the attached config.

BR/Jun

JUNKSA_2 · ‎09-17-2007

Paul,

The config for 7609 router.

BR/Jun

Kevin Dorrell · ‎09-17-2007

Is anyone else getting a server error when they try and open an attchment, or is it just me?

Richard Burts · ‎09-17-2007

Kevin

I just pulled the attachments with no problem.

HTH

Rick

HTH

Rick

paul.matthews · ‎09-17-2007

Right, looking at these I note:

3660:

interface FastEthernet0/0

ip address 18.0.0.1 255.0.0.0

duplex auto

speed auto

router ospf 2

log-adjacency-changes

network 18.0.0.0 0.255.255.255 area 0

network 210.0.18.0 0.0.0.255 area 0

Which is a fas ethernet into area 0 of OSPF process 2.

On the 7609;

interface FastEthernet1/2

switchport

switchport access vlan 10

no ip address

!

interface FastEthernet1/3

no ip address

!

interface FastEthernet1/4

no ip address

!

interface FastEthernet1/5

no ip address

!

interface FastEthernet2/2

no ip address

!

interface FastEthernet2/3

no ip address

!

interface FastEthernet2/4

no ip address

!

interface FastEthernet2/5

no ip address

interface Vlan10

description Riyadh

no ip address

shutdown

No fast ethernet interfaces have an IP address. On is in VLAN10, but that does not have an IP address, and is shut down. The IP network that appears on the fast ethernet in the 3660 is not in this router at all.

Based on that, it appears the fast ethernet is not configured for IP, which means it cannot be advertised, and the same network is not included in either OSPF process.

Paul.

glen.grant · ‎09-17-2007

Makes sure your network statements covers the fast ether subnet otherwise it will not be advertised .

Richard Burts · ‎09-17-2007

Jhun

Thank you for the additional information that you posted. I believe that it demonstrates that the FastEthernet is being advertised. But advertised only on one interface.

I notice in the config that you have configured 2 OSPF processes. I am not clear why you did this but I believe that it is the cause of the apparent problem here. Note that each OSPF process maintains its own version of the Link State Data Base.

I asked for the output of show ip interface brief and of show ip ospf interface as easy ways to find whether the interface was up/up (and it is), if it has a valid IP address (and it does) and if it is included within an OSPF network statement (it is because it is included in the output as an OSPF interface).

Here is part of the output and it includes a significant clue:

FastEthernet0/0 is up, line protocol is up

Internet Address 18.0.0.1/8, Area 0

Process ID 2

note the process ID is 2 where most of the other interfaces have process ID of 1.

There are only 2 interfaces configured in OSPF process 2 - FastEthernet0/0 and Tunnel 0. So the FastEthernet is being advertised over the tunnel 0. But it is not advertised out any other interface because OSPF will only advertise routes that are from the Link State Data Base of its own process. If you check the router at the other end of Tunnel 0 I am sure that you will find that 18.0.0.0/24 is being advertised over the tunnel. But it is not advertised over any other interface.

HTH

Rick

HTH

Rick

JUNKSA_2 · ‎09-17-2007

Hi Rick,

I really appreciate your comments and observation.

Regarding the 2 OSPF process, When I take over the project the configuration is already there and the network is running well in that configuration up to now. I think I will send you the existing config from the other remote router that is connected to the main head quarter for comparison.

Thank you.

BR/Jun

JUNKSA_2 · ‎09-18-2007

Hi Rick,

I checked the other remote site that is connected to the main site.

I think the reason why he configured two (2) OSPF processes in his recent configuration, in the main site we have two (2) 3660 routers, 9 remote sites are directly connected to the first router and the other 7 remote sites are connected to the other router for load balancing and there is a tunneling between the two (2) routers.

Now, since we have only one (1) 7609 router in the main site, I think, no need to configure two (2) OSPF process in this scenario.

This morning, I removed the other ospf process on both ends (Jizan and Riyadh Site) and I add the statement for the fastethernet and the tunneling 0 in OSPF 1 on both ends and its work, when I checked the IP route, the fastethernet and tunneling 0 are advertised. I can ping the fastethernet IP address.

What is your opinion?

BR/Jun