RFC 2081

RFC 2081: RIPng Protocol Applicability Statement 🎀 is by G. Malkin from Xylogics, dated January 1997, and it's Category: Informational 📘. It doesn't list any RFCs it obsoletes or updates. Okay, let's get into it! 💖

Status of this Memo ✨

Okay so this memo gives information to the Internet community 💅. It does not specify an Internet standard of any kind, so no cap, it's not a rulebook. Distribution of this memo is unlimited 🌈.

Abstract 🦋

As required by Routing Protocol Criteria (RFC 1264), this report defines the applicability of the RIPng protocol within the Internet 💜. It's a prerequisite to advancing RIPng on the standards track, so it's basically the homework before the big leagues 🔥.

1. Protocol Documents 📚

The RIPng protocol description is defined in RFC 2080. Go read that one too, bestie 🧚.

2. Introduction 🌸

This report is all about how RIPng can be useful in the new IPv6 Internet 🌐. Lowkey, the environments where RIPng is the IGP of choice are basically the same ones where RIP-2 (RFC 1723) shows up in the IPv4 Internet. And it's important to remember that RIPng is a simple extrapolation of RIP-2, so it has nothing conceptually new 🤓. That means the operational side of distance-vector routing protocols, RIP-2 especially, inside an autonomous system is already well understood 👑.

Now, RIPng is not meant to replace OSPFng in big autonomous systems 🙈. The limits on AS diameter and complexity that apply to RIP-2 also apply to RIPng. What it does do is let the smaller, simpler distance-vector protocol handle environments that need authentication or variable length subnet masks, but aren't big or complicated enough to need the larger, more complex link-state protocol 🫶.

The rest of this report walks through how each RIPng feature is useful within IPv6 ✨.

3. Applicability 💅

The big goal when building RIPng was to make the minimum necessary change to RIP-2 🎯. Honestly it's giving minimal glow-up: the IPv4 address got expanded into an IPv6 address, the IPv4 subnet mask got swapped for an IPv6 prefix length, the next-hop field got eliminated (but the functionality is still there ✨), and authentication got removed. The route tag field was kept. The maximum diameter of the network (the maximum metric value) is 15, and 16 still means infinity, aka unreachable 💀.

The basic RIP header is unchanged 🎀. The big change is that routing packets aren't arbitrarily limited in size anymore 😭. Routing updates are never forwarded, so the packet size is now set by the physical media and the sizes of the headers in front of the routing data (that's media MTU minus the combined header lengths) 🦋. The number of routes in a routing update is the routing data length divided by the size of a routing entry 🧮.

3.1 Prefix 💜

The address field of a routing entry is 128 bits long, expanded from the 32 bits RIP-2 had 🌈. That lets the RIP entry carry an IPv6 prefix 🔥.

3.2 Prefix Length 🍓

The 32-bit RIP-2 subnet mask field is replaced by an 8-bit prefix length field 🎀. It lets you specify how many bits of the prefix actually form the prefix ✨.

3.3 Next Hop 🧚

Okay this one is so clever 😍. Instead of just letting the recipient of an update set the next hop to whoever sent it, RIPng lets you specify the next hop directly. That's how you skip unnecessary hops through routers running multiple routing protocols 🦋. Check out this example topology:

         -----   -----         -----   -----
         |IR1|   |IR2|         |XR1|   |XR2|
         --+--   --+--         --+--   --+--
           |       |             |       |
         --+-------+-------------+-------+--
           |--------RIPng--------|

The Internal Routers (IR1 and IR2) are only running RIPng 🏠. The External Routers (XR1 and XR2) are both running BGP, for example 🌍. But only XR1 runs BGP and RIPng 🙃. Since XR2 isn't running RIPng, the IRs won't know it exists, so they'll never use it as a next hop, even if it's a better one than XR1 💔. Of course XR1 knows this, and it can use the Next Hop mechanism to point out that XR2 is the better next hop for some routes 👑. Main character energy 🫶.

3.4 Authentication 🔐

Authentication got added to RIP-2 because RIP-1 didn't have it, but it's been dropped from RIPng 😌. That's totally safe to do because IPv6, which carries the RIPng packets, has built-in security that IPv4 didn't have 🔒.

3.5 Packet Length 📦

If RIPng routing update packets are as big as possible, a complete update needs way fewer packets to be sent 🚀. This doesn't change how the distance-vector protocol works at all. It's just a performance boost ⭐.

3.6 Diameter and Complexity 🌀

The limit of 15 cost-1 hops comes from the distance-vector protocol itself, which depends on counting to infinity to resolve some routing loops 🔁. If infinity is too high, the time it takes to resolve (not to mention the number of routing updates that would get sent) becomes prohibitive 😩. If infinity is too small, the protocol becomes useless in a reasonably sized network 💀. Choosing 16 for infinity was made in the earliest RIP implementations, and experience has shown it to be a good compromise value 🍓.

RIPng will efficiently support networks of moderate complexity, meaning topologies without too many multi-hop loops 🕸️. It also efficiently supports topologies that change often, because routing table changes are made incrementally and don't need the computations that link-state protocols need to rebuild their maps 🧠✨.

4. Conclusion 👑

Because the basic protocol is unchanged, RIPng is just as correct a routing protocol as RIP-2 🎀. It fills the same niche for IPv6 that RIP-2 fills for IPv4 🦋. Iconic, periodt 💅.

5. Security Considerations 🔐

RIPng security is discussed in section 3.4 😤. Read it, it's important 🫶.

Author's Address 🖊️

Gary Scott Malkin
Xylogics/Bay Networks
53 Third Avenue
Burlington, MA 01803

Phone:  (617) 238-6237
EMail:  gmalkin@xylogics.com