OSPF Best Practice

[MARTIN HUERTER]

Over the years we've been adding layer 3 devices to our OSPF backbone area 0 without too much thought put into it. Plug it in, turn on routing and wait for the OSPF neighbors to show up. It works fine and we've never had a problem with OSPF on our network, but after looking things over and evaluating our current situation, I thought I might overview our OSPF backbone to make sure we couldn't make improvements or if we are a house of cards just waiting for the wrong scenario for things to crash. A picture says a thousand words, especially in networking, so I have uploaded a generic diagram of our current OSPF backbone layout and device definition. I was hoping to get some feedback or suggestions if we need to change things, or use better practices than we've been using. 

Thanks!!

[MHM Cisco World]

follow

[Joseph W. Doherty]

It appears (?) you're running DR/BDR on p2p links.  If so, you might consider moving those to OSPF p2p.

[MARTIN HUERTER]

Joseph,

Thanks for your response. I guess I don't know what an OSPF p2p is, could you elaborate?

[Richard Burts]

p2p is point to point. On a multi access link (like Ethernet) where there are potentially multiple OSPF neighbors then OSPF goes through the process of electing DR and BDR. Then OSPF routers on that segment establish full adjacency with the DR and the BDR but not with other neighbors on that segment. If there are really going to be only 2 neighbors then there is no benefit in having DR and BDR. Configuring the interfaces as point to point tells OSPF to not go through the DR/BDR election on that interface.

[MARTIN HUERTER]

I knew p2p stood for point-to-point, I just didn't understand what "moving those to OSPF p2p" meant. 

[Richard Burts]

So now is it clear that moving those to OSPF p2p means using the optional interface parameter to change the OSPF interface type?

[Joseph W. Doherty]

Another Cisco OSPF option, you might consider, if supported on your platforms, is iSPF

[MHM Cisco World]

you config the IGP between the FW, VSS and Edge router, so the Edge router know all prefix add to FW and have path to it. 
Edge router also do NAT not the FW.
you use OSPF so you can config the Area 0 which include the Edge router and VSS and FW.
FW will be as ABR "if there is no static route config which it" between Area 0 and Area 1. 
FW will send summary route to VSS & Edge router, and hence minimize the OSPF DB.