cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
732
Views
15
Helpful
4
Replies

Interior Routing Protocol

hs08
Spotlight
Spotlight

Hi..

OSPF is part of Interior Routing Protocol and the term of Interior Routing Protocol is to communicate within a single autonomous systems. So i want to ask which OSPF config part indicated the autonomous system / number? Are all OSPF router should use same process ID to indicate the router in same autonomous system?

1 Accepted Solution

Accepted Solutions

Giuseppe Larosa
Hall of Fame
Hall of Fame

Hello @hs08 ,

there is no autonomous system concept in OSPFv2 and the process id is a local identifier that is never sent out in OSPF Hello packets so the routers can have a different process-id they need to agree on the area number on the area type and on hello and dead timers and authentication to build adjacency.

In OSPFv3, there is a field called instance-id that is used to implement multiple address families support (IPv6 unicast, IPv4 unicast) in that case the routers have to agree on this parameter but again it is not an autonomous system.

Hope to help

Giuseppe

 

View solution in original post

4 Replies 4

Giuseppe Larosa
Hall of Fame
Hall of Fame

Hello @hs08 ,

there is no autonomous system concept in OSPFv2 and the process id is a local identifier that is never sent out in OSPF Hello packets so the routers can have a different process-id they need to agree on the area number on the area type and on hello and dead timers and authentication to build adjacency.

In OSPFv3, there is a field called instance-id that is used to implement multiple address families support (IPv6 unicast, IPv4 unicast) in that case the routers have to agree on this parameter but again it is not an autonomous system.

Hope to help

Giuseppe

 

"there is no autonomous system concept in OSPFv2 . . "

Well, yes and no.

What's an OSPF ASBR?  ; )

(to OP)

However, unlike BGP, there's no explicit ASN, and as @Giuseppe Larosa correctly describes, OSPF process number ID does not need to match between OSPF neighbors.

BTW, keep in mind AS, in general, has a broader meaning than being tied into a routing protocol, for example see: https://www.techopedia.com/definition/11063/autonomous-system-as

Going by the prior definition, I might interconnect ASs using OSPF or I might have different IGPs, active, within the same AS.  BGP, though, as a routing protocol, by design lends itself to working within and between ASs due to having eBGP and iBGP.

In theory, I could build an IPv4 Internet just using OSPF, but it would need to be even much more structured than our existing IPv4 BGP based Internet.  (BTW, just to touch upon "structure", consider why our existing IPv4 Internet [generally] doesn't allow network prefixes larger than /24 between ASs or that BGP's ASN is "only" a 16-bit number.)

The link posted by Joseph gives a very good definition of autonomous system: "An autonomous system (AS) is a network or a collection of networks that are all managed and supervised by a single entity or organization." This concept is a basic foundation of all of the routing protocols. BGP, as an Exterior Routing Protocol, explicitly identifies AS membership as part of its configuration. Interior Gateway Protocols use the concept but do not necessarily identify AS membership in their configuration. Perhaps it might help to think about it in this way: if you are running OSPF in your network and configuring a layer 3 device in your network that connects to a layer 3 device in someone else's network would you choose to run OSPF on that interface? If you would do that you are sharing all of your routing information with them and they are sharing all of their routing information with you, and they are really not separate from you.

HTH

Rick

"If you would do that you are sharing all of your routing information with them and they are sharing all of their routing information with you, and they are really not separate from you."

Ain't necessarily so.

Possibly @Richard Burts has something in mind like:

 

OSPF "AS" 1 <ASBR> OSPF "AS" 2

 

And if so, he would be correct.

But if you have something like:

 

OSPF "AS" 1 <ASBR> OSPF Exchange <ASBR> OSPF "AS" 2

 

Each ASBR only has access to its own OSPF topology and the OSPF topology of the "Exchange", which could even be just a single p2p link.  Routes exchanged between the two ASs are strictly controlled on each ASBR.

OSPF "AS" 1 cannot "see" or "know" OSPF "AS" 2 routes unless the ASBR connected to OSPF "AS" 2 redistributes them into the OSPF "Exchange".  Converse is true for the other direction.  Each ASBR also, of course, has full control over what routes it "injects" into its own local OSPF topology.

Basically, what's just described is much like what happens on two eBGP peer routers, but without using BGP.

BTW, I'm not saying OSPF is "better" than BGP, but I am saying, that perhaps an OSPF ASBR (Autonomous System Border Router) is well named.

Review Cisco Networking for a $25 gift card