Table of contents
This OSPFv3 Introduction gives a high level overview of the protocol.
OSPFv3 is a link-state Interior Gateway Protocol (IGP) that supports IPv6 and Dual Stack. It's an open standard and is vendor neutral.
Let's take a moment to explain this definition in more detail:
- OSPF stands for Open Shortest Path First.
- The path(s) with the best metric will be placed into the routing table.
- There can be multiple "best paths". This can occur if multiple paths have the same metric and this metric value is better than all other alternate paths. If this occurs, the traffic will be balanced equally amongst the "best paths".
- v3 (Version 3)
- OSPFv3 runs IPv6 and Dual Stack only. IPv4 only is not supported. For IPv4 only, OSPFv2 must be used. OSPFv2 and v3 can be run on the same router at the same time in parallel. Their databases are kept completely separate...the two protocols do not communicate with each other.
- OSPFv3 uses IPv6's Link-Local addresses to advertise information across the network. Even if only IPv4 addresses are being advertised, devices must be enabled for IPv6 and have Link Local Addresses (LLAs) configured.
- Link State Protocol
- Since Link State Protocols send a lot of messages and each router creates its own map of the entire network from its perspective, they can consume a significant amount of network resources.
- IGP (Interior Gateway Protocol) - A routing protocol used within the same autonomous system (usually within the same organization).
- Open Standard - Open standards are available for all to read and implement.
- Vendor Neutral - It can be used by any vendor. Devices from different vendors can share the same OSPFv3 process.
Where to Use it
OSPFv3 is commonly used in large enterprise IPv6 networks at the access, distribution and core layers. It is also used in the Enterprise Data Center, WAN/MAN and branch offices.
Pros and Cons
OSPFv3 is very flexible and scalable. The trade-off, however, is that OSPFv3 is complex.
- Open Standard
- Vendor Neutral
- Very Scalable
- Can be significantly demanding on network resources (i.e. CPU, Bandwidth, etc)
Why was OSPF Created
OSPF was created to overcome the constraints of RIP (Routing Information Protocol).
- Faster Convergence – If a link fails, OSPF is faster at switching over to the next optimal route.
- Better Scalability – RIPng has a max hop count of 15. OSPF can accommodate many, many devices.
- Consideration of Network Conditions – RIP uses hop count as its metric. However, there are scenarios where the optimal path may be more hops. OSPF, by default, takes bandwidth and delay into an account.
- Load balancing – If multiple OSPF process best paths have the same metric, OSPF will do equal-cost load balancing.
How it Works
Before routers can begin to share OSPFv3 routing information they need to go through the following steps, in order:
- Form a Neighborship - After routers on the same link agree on the parameters for sending OSPFv3 messages, they become neighbors.
- Form an Adjacency - When neighbors have completely exchanged OSPFv3 database information they have become adjacent.
- Router links can be grouped into Areas to more efficiently share, or not share, database information. For example, an area may only require a single summary route to reach routers in another area.
- Choosing the Best Route(s) – Every router will use the information in their Link State Database (LSDB) to create a map of the entire network by running the CPU intensive Dijkstra algorithm. The Dijkstra algorithm calculates the Shortest Path First (SPF) tree for every destination in the routing table.
In this lesson we've:
- Defined OSPFv3
- Described where it's commonly used
- Listed the major pros and cons of the protocol
- Talked about why it was created
- Had a primer of how the protocol works
In the next lesson we'll cover OSPFv3 Authentication and Encryption.
RFC 5340: OSPF for IPv6
CCIE Routing and Switching, v5.0, Volume 1, Fifth Edition