OSPF routing issue with different network product

[alexals]

Hello there,

I hope someone can advise me on OSPF compatibility between Cisco, Radware Linkproof and Brocade router. Our core network has three network devices as I mentioned earlier. Our core switches are running in VSL mode as well as running OSPF routing between our two unit of Radware Linkproof and a Brocade router. Since we moving towards dynamic routing (OSPF) from static routing, I noticed that the OSPF routing keeps flapping on every 2 - 3 minutes in our Cisco 6509 (core switches).

001016: Sep 20 21:40:14.140 GMT+8: %OSPF-5-ADJCHG: Process 100, Nbr 172.31.255.4 on Vlan550 from LOADING to FULL
, Loading Done
001017: Sep 20 21:41:04.164 GMT+8: %OSPF-5-ADJCHG: Process 100, Nbr 172.31.255.4 on Vlan550 from LOADING to FULL
, Loading Done
001018: Sep 20 21:41:54.185 GMT+8: %OSPF-5-ADJCHG: Process 100, Nbr 172.31.255.4 on Vlan550 from LOADING to FULL
, Loading Done
001019: Sep 20 21:42:44.201 GMT+8: %OSPF-5-ADJCHG: Process 100, Nbr 172.31.255.4 on Vlan550 from LOADING to FULL
, Loading Done
001020: Sep 20 21:43:34.221 GMT+8: %OSPF-5-ADJCHG: Process 100, Nbr 172.31.255.4 on Vlan550 from LOADING to FULL
, Loading Done
001021: Sep 20 21:44:24.246 GMT+8: %OSPF-5-ADJCHG: Process 100, Nbr 172.31.255.4 on Vlan550 from LOADING to FULL
, Loading Done

This issue caused our network sluggish and unstable due to frequent routing flap. Under Cisco configuration section for OSPF:

router ospf 100
 router-id 172.31.255.255
 log-adjacency-changes
 nsf

.....................................

I noted 'nsf' command defined. I'm not sure whether this might caused the route flapping but I feel it should be removed since there's no other Cisco devices on the core network area.

How do I troubleshoot this route flapping issue?

Kindly advice.

Thank you in advance,

Alex

[Reza Sharifi]

Hi,

NSF should not cause flapping.  Its protocol to use between multiple Cisco devices for none stop forwarding.

You can remove it since you are dealing with other vendor's product anyway. Not familiar with other vendors equipment, but what device is 172.31.255.4 on Vlan550?

What interface on the Cisco side connected to this device?

Can you post the config?

https://www.cisco.com/en/US/docs/ios/12_2t/12_2t15/feature/guide/ftosnsfa.html

HTH

[alexals]

Hi there ! Thank you for your reply.

172.31.255.4 is actually our Radware Linkproof unit and we have two of them with the other unit using 172.31.255.3. VLAN 550 is where Cisco, Brocade and Radware Linkproof connected to it. The OSPF configuration on both Linkproof unit as well as Brocade router are general OSPF config. Here a snippet of OSPF config in Cisco switches

router ospf 100
 router-id 172.31.255.255
 log-adjacency-changes
 nsf
 redistribute static subnets route-map STATIC2OSPF
 network 10.10.10.253 0.0.0.0 area 0
 network 10.10.11.253 0.0.0.0 area 0
 network 10.10.12.251 0.0.0.0 area 0
 network 172.31.255.255 0.0.0.0 area 0

Regards,

Alex

[Mark Malone]

anything else appearing before the flaps in the logs ?

some things to check for causes
Unstable links within the network
Unstable neighbor(s) within the network

Duplicate router ID within the network

Lookk  at the TSHOOT section in this doc for OSPF , debug ip ospf monitor may help you pinpoint the issue

https://www.cisco.com/networkers/nw00/pres/2206_7-5.pdf

[alexals]

Hi,

None except the flaps. Here some snippets from our Cisco core-switches and Brocade

Cisco

Neighbor ID     Pri   State           Dead Time   Address         Interface
172.31.255.2      1   FULL/DROTHER    00:00:32    10.10.11.200  Vlan550
172.31.255.3      2   FULL/BDR        00:00:36    10.10.11.51   Vlan550
172.31.255.4      1   FULL/DROTHER    00:00:38    10.10.11.52   Vlan550

Brocade

Number of Neighbors is 4, in FULL state 3

Port   Address         Pri State      Neigh Address   Neigh ID        Ev Opt Cnt
1/1    172.16.32.1     1   FULL/BDR   172.16.32.2     172.31.254.3    5  66  0
v550   10.10.11.200  2   FULL/BDR   10.10.11.51   172.31.255.3    83 2   0
v550   10.10.11200  1   INIT/OTHER 10.10.11.52   172.31.255.4    1802   0
v550   10.10.11.200  1   FULL/DR    10.10.11.251  172.31.255.255  5  82  0

I guess my Radware Linkproof is not 'stable'....

Alex

[Julio E. Moisa]

Also you could execute a: debug ip ospf adj

You can verify the MTU configured on both interfaces and the network type under the interfaces, also check the timers. 

Also check this link:

https://community.brocade.com/t5/Ethernet-Switches-Routers/My-Brocade-to-Cisco-OSPF-connection-is-flapping/m-p/66334#M4387

[Mark Malone]

what do the interfaces look like , did you check each interface that participates in OSPF that there isn't issues with it or its attached SFPs just to be sure as something undtable in hardware may cause it too and as Julio has pointed out the the timers maybe increase them see if it stabilizes a bit

[Reza Sharifi]

Hi,

Is 172.31.255.255 is the IP address of the interface on the Cisco switch?

Can you post the interface config?

Also, all devices are in one vlan (550) now.  You  may want to put each connection in a seperate vlan and test again.

HTH

[alexals]

Hello all,

Thank you for your advices. I'm really appreciates it. I think the problem arose due to VRRP service running together with OSPF. Previous set up for our both Linkproof unit are using static routing with VRRP. When we moving towards to OSPF routing in our network. We did not disabled the VRRP (we thought it is okay to leave it active) and we let it run until we got into issues recently. I got feeling that the VRRP service might caused the routing flap.

Now I'm monitoring it whether it solves the problem.

Alex