Cisco Support Community
Showing results for 
Search instead for 
Did you mean: 
New Member

When to use implement OSPF

Quite concisely:

I'd like to know when it's time to implement an IGP, in my specific inquiry OSPF? There's an overwhelming amount of literature on the subject, but none that I've found that specifically spells out when it makes sense to implement an IGP. 

I understand many of you may have a super technical reply, but I'm looking for a condensed response ... i.e. short, descriptive paragraph, so a fellow (still a little green like myself) can really understand the logic. I have plenty of material on LSA types, architecture, route summarization, etc. 


Once again, just looking for clear explanation when you know it's time to begin using OSPF. More fundamentally, I don't entirely understand what is accomplished with OSPF, besides the generic explanation of rapid convergence ... which I see as more efficient way to exchange routes than (for example) having a myriad of static routes across your Metro Ethernet hub and spoke network. Any feedback is appreciated, especially from the senior folks who can give a powerful, condensed explanation. 



Everyone's tags (1)
Hall of Fame Super Silver

You have come pretty close to

You have come pretty close to stating the fundamental issue already but let me try to explain it a bit differently. You are correct that the fundamental choice is to establish routing logic using static routes or by using a dynamic routing protocol. So let us look at the implications of that choice.


On the positive side for static routes are that they are simple, they are manually configured so you are in complete control of how traffic will flow in your network. And static routes do not consume any of the bandwidth on your network links. On the negative side for static routes is that they are manually configured so it is manual effort every time you need to configure one and manual effort every time you need to change one. The other (and really important) negative is that static routes do not react to changes in your network. If a link in your network goes down the static routes still tries to send traffic over it (and drops the traffic because it can not go through) until you recognize the problem and reconfigure the static route.


On the positive side for a dynamic protocol like OSPF is that they will automatically select the best route to a destination when more than one path is possible, and that they react to changes in the network. If a link goes down and there is another way to get to the destination then the dynamic routing protocol will send traffic over the alternate link until it is repaired then the routing protocol will resume sending traffic over the preferred link. No manual effort here :) On the negative side for dynamic routing protocols is that they do consume some bandwidth for sending routing protocol traffic. Also that the routing protocol makes decisions about how to send traffic and sometimes that might not be the way that you would want the traffic to go.


So the question you ask about choosing static routes or choosing a dynamic routing protocol comes down to how you view your network. If it is reasonable small and is pretty static (very few changes occur and links rarely go down) and especially if usually there is only one way to get to destinations then static routes are an excellent choice. As your network gets larger and a bit more complex, (and if you do not want to worry about some link that goes down at midnight) then the balance shifts and the dynamic routing protocol is the better choice.


Perhaps another way to look at this is that it comes down to how you want choices about how traffic flows in the network to be made. If you want to be responsible for all of the decisions about how traffic should flow and about how to react to changes in the network then choose static routes. If you want the choices to be made by something that is automated and reduces manual effort then choose the dynamic routing protocol.





CreatePlease to create content