OSPF troubles

[netops01]

I hope someone can help me with this. I have a 6506 with two supervisor/MSFC's. They are configured as Active/Passive, both having identical configs, one having all Vlan's shut down. I use OSPF between the active Vlan to my Nokia firewall and everything works fine.

However when I shutdown the working Vlan and bring up the identically configured vlan on my other MSFC OSPF gets stuck in EXSTART/DR. Here is a sample of deb ip os adj.

Feb 7 06:22:06: OSPF: Neighbor change Event on interface Vlan11

Feb 7 06:22:06: OSPF: DR/BDR election on Vlan11

Feb 7 06:22:06: OSPF: Elect BDR 172.31.111.17

Feb 7 06:22:06: OSPF: Elect DR 172.31.111.234

Feb 7 06:22:06: DR: 172.31.111.234 (Id) BDR: 172.31.111.17 (Id)

Feb 7 06:22:11: OSPF: Retransmitting DBD to 172.31.111.234 on Vlan11

Feb 7 06:22:11: OSPF: Send DBD to 172.31.111.234 on Vlan11 seq 0x1A7D opt 0x42

flag 0x7 len 32

Feb 7 06:22:16: OSPF: Retransmitting DBD to 172.31.111.234 on Vlan11

Feb 7 06:22:16: OSPF: Send DBD to 172.31.111.234 on Vlan11 seq 0x1A7D opt 0x42

flag 0x7 len 32

Feb 7 06:22:21: OSPF: Retransmitting DBD to 172.31.111.234 on Vlan11

Feb 7 06:22:21: OSPF: Send DBD to 172.31.111.234 on Vlan11 seq 0x1A7D opt 0x42

flag 0x7 len 32

Feb 7 06:22:26: OSPF: Retransmitting DBD to 172.31.111.234 on Vlan11

Feb 7 06:22:26: OSPF: Send DBD to 172.31.111.234 on Vlan11 seq 0x1A7D opt 0x42

flag 0x7 len 32u al

If i go back to the original MSFC the OSPF works! I have cross checked both configs repeatedly. Any thoughs please?

[ruwhite]

It almost seems like you might have a unicast problem across the link--can you ping between them? Any parameter mismatches should be recorded in the debug output.

A question: Is it possible that the old mac address is in the other device's arp cache, and it's sending it's unicast traffic to the wrong place? So, multicast traffic is getting through, but not unicast?

Russ.W

[netops01]

Russ

I think you nailed it with the mac address! I went into the Nokia box and the arp cache showed the mac address of the normally active vlan interface (as you would expect). I now have to schedule down time to try clearing the cache. I will post results here when I get them. Thanks

Ian

[netops01]

Nope, the trouble still exists! Cleared arp cache and mls entries on all devices, still gets stuck in EXSTART/DR. Tried a different Vlan with just Cisco devices and again, stuck in EXSTART/DR! Next early morning attempt will be to swap the active Vlan from one MSFC to the other, save the config and then reboot! Oh well, we live and learn.......

Ian

[karief]

Hi..

Is it all the config been replicated by both msfc's ,I feel there is some misconfiguration's on both teh MSFC's.

The best thing is, as long as you are having two msfc's , why not change your approach..

1. Enable redundancy, High Availabilty

2. Enable priority between the two msfc's for all vlans

3. Enable OSPF config's

In this case I belevie you don't want to worry about any ospf stuffs.

regards

arief

[netops01]

Configs are identical. I even created a new vlan on each MSFC and the same behavour was seen. I need OSPF as I am not allowed to change the network at that level.

Ian

[ruwhite]

Can you ping 224.0.0.5 or 224.0.0.6 when this happens? I assume you can ping the unicast address.... Could you ping and use the extended option, and try a sweep range of sizes, from 100 bytes to 1500 bytes (or whatever the MTU is)?

:-)

Russ.W

[netops01]

No reply to multi or unicast pings. Looks like I will clear the config on the second MSFC and re-build from scratch. If the problem still exists then its a TAC matter.......

[netops01]

Anybody give me an idea of the cost of a TAC case, here in Saudi Arabia. We do not have any support contract for our hardware.

Thanks

Ian

[ruwhite]

Email me off line....

:-)

Russ.W

riw@cisco.com

[netops01]

Am starting rebuild now, will post results, if different than before!