RFC 1966
BGP Route Reflection An alternative to full mesh IBGP
RFC 1966: BGP Route Reflection 🦋
An alternative to full mesh IBGP
- Working group: Network Working Group
- Authors: T. Bates and R. Chandra, both from cisco Systems 💅
- Category: Experimental 🧪
- Date: June 1996 📅
- Obsoletes and Updates: none listed in the header
Status of This Memo 📜
Okay so this memo defines an Experimental Protocol for the Internet community 🧪. It is not an Internet standard of any kind. Discussion and suggestions for improvement are requested, so bestie, share your thoughts 💬. Distribution of this memo is unlimited 🌈.
Abstract ✨
The Border Gateway Protocol [1] is an inter-autonomous system routing protocol designed for TCP/IP internets 🌐. The tea is that BGP deployments are usually configured so that all BGP speakers within a single AS MUST be fully meshed, which means any external routing information MUST be re-distributed to all other routers in that AS. That is a serious scaling problem 😭. It has been well documented, and several alternatives have been proposed [2,3].
This document describes the use and design of a method known as "Route Reflection" that gets rid of the need for "full mesh" IBGP 💖.
1. Introduction ✨
Currently on the Internet, BGP deployments are configured so that all BGP speakers within a single AS MUST be fully meshed, and any external routing information MUST be re-distributed to all other routers in that AS. Honestly, this "full mesh" requirement clearly does not scale once there are a large number of IBGP speakers, which is super common in many of today's internet networks 🥲.
For n BGP speakers within an AS you MUST maintain n*(n-1)/2 unique IBGP sessions. With finite bandwidth and router CPU, this clearly does not scale 💀.
This scaling problem is well documented, and a number of proposals have been made to fix it [2,3]. This document is another alternative, and it's called "Route Reflection" 🦋. It changes the commonly understood concept of IBGP and adds two new optional transitive BGP attributes.
2. Design Criteria 🎀
Route Reflection was designed to satisfy the following criteria, and it's giving main character energy 👑.
- Simplicity: Any alternative MUST be both simple to configure as well as simple to understand.
- Easy Migration: It MUST be possible to migrate from a full mesh configuration without needing to change either topology or AS. Ngl, this is an unfortunate management overhead of the technique proposed in [3] 🙈.
- Compatibility: It MUST be possible for non compliant IBGP peers to continue to be part of the original AS or domain without any loss of BGP routing information 🫶.
These criteria came from the operational experiences of a very large, topology-rich network with many external connections 🌸.
3. Route Reflection 🦋
The basic idea of Route Reflection is very simple, so let's look at the example in Figure 1 below 💖.
+------ + +-------+
| | IBGP | |
| RTR-A |--------| RTR-B |
| | | |
+-------+ +-------+
\ /
IBGP \ ASX / IBGP
\ /
+-------+
| |
| RTR-C |
| |
+-------+
Figure 1: Full Mesh IBGPIn ASX there are three IBGP speakers (routers RTR-A, RTR-B and RTR-C). With the existing BGP model, if RTR-A receives an external route and it is selected as the best path, it MUST advertise that external route to both RTR-B and RTR-C. RTR-B and RTR-C, as IBGP speakers, will not re-advertise these IBGP learned routes to other IBGP speakers 🙅♀️.
Now, if this rule is relaxed and RTR-C is allowed to reflect IBGP learned routes, then it could re-advertise (or reflect) the IBGP routes learned from RTR-A to RTR-B, and vice versa. That would eliminate the need for the IBGP session between RTR-A and RTR-B, as shown in Figure 2 below ✨.
+------ + +-------+
| | | |
| RTR-A | | RTR-B |
| | | |
+-------+ +-------+
\ /
IBGP \ ASX / IBGP
\ /
+-------+
| |
| RTR-C |
| |
+-------+
Figure 2: Route Reflection IBGPThe Route Reflection scheme is based on this basic principle 👑.
4. Terminology and Concepts 🧚
We use the term "Route Reflector" (RR) to mean an IBGP speaker that participates in the reflection 🪞. The internal peers of an RR are divided into two groups:
- Client Peers
- Non-Client Peers
An RR reflects routes between these groups. An RR along with its client peers form a Cluster 🫧. The Non-Client peer MUST be fully meshed, but the Client peers need not be fully meshed. The Client peers SHOULD NOT peer with internal speakers outside of their cluster 🚫. Figure 3 depicts a simple example that outlines the basic RR components using the terminology above.
/ - - - - - - - - - - - - - -\
| Cluster |
+-------+ +-------+
| | | | | |
| RTR-A | | RTR-B |
| |Client | |Client | |
+-------+ +-------+
| \ / |
IBGP \ / IBGP
| \ / |
+-------+
| | | |
| RTR-C |
| | RR | |
+-------+
| / \ |
\ - - - - -/- - -\- - - - - - /
IBGP / \ IBGP
+-------+ +-------+
| RTR-D | IBGP | RTR-E |
| Non- |---------| Non- |
|Client | |Client |
+-------+ +-------+
Figure 3: RR Components5. Operation 🔥
When an RR receives a route, it selects the best path based on its path selection rule. After the best path is selected, it MUST do the following, depending on the type of peer it's getting the best path from:
- A route from a Non-Client peer: Reflect to all other Clients.
- A route from a Client peer: Reflect to all the Non-Client peers, and also to the Client peers other than the originator. (Hence the Client peers are not required to be fully meshed 🫶.)
- Route from an EBGP peer: Send to all the Client and Non-Client Peers.
An Autonomous System could have many RRs. An RR treats other RRs just like any other internal BGP speaker 🤝. An RR could be configured to have other RRs in a Client group or a Non-Client group.
In a simple configuration, the backbone could be divided into many clusters. Each RR would be configured with other RRs as Non-Client peers, so all the RRs will be fully meshed 🕸️. The Clients will be configured to maintain an IBGP session only with the RR in their cluster. Thanks to route reflection, all the IBGP speakers will receive reflected routing information 💌.
It is normal in an Autonomous System to have BGP speakers that do not understand the concept of Route Reflectors. Let's call them conventional BGP speakers 🦕. The Route Reflector scheme allows these conventional BGP speakers to co-exist. Conventional BGP speakers can be members of either a Non-Client group or a Client group. This allows for an easy and gradual migration from the current IBGP model to the Route Reflection model 🌈. One way to start is by configuring a single router as the designated RR and configuring other RRs and their clients as normal IBGP peers. Additional clusters can then be created gradually 🍓.
6. Redundant RRs 🛟
Usually a cluster of clients will have a single RR. In that case, the cluster is identified by the ROUTER_ID of the RR. Still, that's a single point of failure 😬. So, to make it possible to have multiple RRs in the same cluster, all RRs in the same cluster MUST be configured with a 4-byte CLUSTER_ID. That way an RR can tell apart the routes that came from other RRs in the same cluster 🧩.
7. Avoiding Routing Information Loops 🔁
As IBGP learned routes are reflected, it's possible for mis-configuration to form route re-distribution loops 🙃. The Route Reflection method defines the following attributes to detect and avoid routing information loops, and they're honestly iconic 💅.
ORIGINATOR_ID
ORIGINATOR_ID is a new optional, non-transitive BGP attribute of Type code 9. This attribute is 4 bytes long, and it's created by an RR. It carries the ROUTER_ID of the originator of the route in the local AS. A BGP speaker SHOULD NOT create an ORIGINATOR_ID attribute if one already exists. A route reflector MUST never send routing information back to the router specified in ORIGINATOR_ID 🚫🔁.
CLUSTER_LIST
CLUSTER_LIST is a new optional, non-transitive BGP attribute of Type code 10. It's a sequence of CLUSTER_ID values representing the reflection path that the route has passed 🗺️. It is encoded as follows:
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Attr. Flags |Attr. Type Code| Length | value ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+Where Length is the number of octets 🔢.
When an RR reflects a route from its Clients to a Non-Client peer, it MUST append the local CLUSTER_ID to the CLUSTER_LIST. If the CLUSTER_LIST is empty, it MUST create a new one 🆕. Using this attribute, an RR can identify if routing information has looped back to the same cluster due to mis-configuration. If the local CLUSTER_ID is found in the cluster-list, the advertisement will be ignored 🙅♀️💜.
8. Implementation and Configuration Considerations ✨
okay bestie, big config warning 🚨 you SHOULD be super careful that none of the BGP path attributes defined above can get modified through configuration when internal routing info is going between RRs and Clients and Non-Clients 🫶. if that happens it could result in looping of routes 😭🔁
some implementations let you change the NEXT_HOP attribute 💅 and yeah, there can be a real need for an RR to modify NEXT_HOP for EBGP learned routes it sends to its internal peers. BUT it MUST NOT be possible for an RR to set it on reflected IBGP routes 🙅♀️. that breaks the basic principle of Route Reflection and can result in potential black holeing of traffic 💀
An RR SHOULD NOT modify any AS-PATH attributes (i.e. LOCAL_PREF, MED, DPA) that could change consistent route selection 🧠. this could result in potential loops 🌀
the BGP protocol gives a Client no way to identify itself dynamically as a Client to an RR configured BGP speaker 🥺, so the simplest way to get this working is manual configuration 🎀
9. Security Considerations 🔒
Security issues are not discussed in this memo. 😤 so this doc does not get into security at all, don't assume it's covered here, periodt 💅
10. Acknowledgments 🫶
the authors would like to thank Dennis Ferguson, Enke Chen, John Scudder, Paul Traina and Tony Li for the many discussions that led to this work 💖 this idea was developed from an earlier discussion between Tony Li and Dimitri Haskin 🧚
11. References 📚
- [1] Rekhter, Y., and T. Li, "A Border Gateway Protocol 4 (BGP-4)", RFC 1771, March 1995.
- [2] Haskin, D., "A BGP/IDRP Route Server alternative to a full mesh routing", RFC 1863, October 1995.
- [3] Traina, P., "Limited Autonomous System Confederations for BGP", RFC 1965, June 1996.
12. Authors' Addresses 💌
Tony Bates
cisco Systems
170 West Tasman Drive
San Jose, CA 95134
Phone: +1 408 527 2470
EMail: tbates@cisco.com
Ravishanker Chandrasekeran
(Ravi Chandra)
cisco Systems
170 West Tasman Drive
San Jose, CA 95134
EMail: rchandra@cisco.com 🦋
📦 where this page comes from
- ietf: rfc1966 1996-06
- 💖 girlified by claude-haiku-5-5 on : a language model's rewrite of the original, under its terms