Vyatta Network OS Documentation

Learn how to install, configure and operate the Vyatta NOS, which helps drive our virtual networking & physical platforms portfolio.

BGP overview

Border Gateway Protocol (BGP) is the principal inter-domain routing protocol used on the Internet. BGP version 4 is specified in RFC 4271, which obsoletes the original BGPv4 specification defined in RFC 1771.

The principal concept of BGP is that of the Autonomous System (AS). An AS is a routing domain that is under one administrative authority, and which implements its own routing policies. For example, one Internet Service Provider (ISP) would have its own AS, while another would have its own, different, AS. Many large enterprises also have their own AS, particularly if they are multi-homed (that is, connected to multiple ISPs). The BGP routing protocol is used to convey network reachability information from one AS to another.

Routers that are configured to run BGP between one another are known as BGP peers or BGP neighbors. BGP uses a TCP connection on the well-known port 179 to exchange routing information between peers. BGP peers that are configured within the same AS are referred to as internal BGP (iBGP) peers. A BGP peer that is configured in a different AS than another peer are referred to as external BGP (eBGP) peers.

There are two basic types of BGP route exchanges that occur between peers: route announcements and route withdrawals.

  • A route announcement tells a peer that it can reach a particular network via the announcing router, and includes attributes associated with that path.
  • A route withdrawal tells a peer that a previously announced route is no longer reachable via this peer.

All valid route announcements that are received on a BGP router are placed into the BGP table of the router. (These routes are typically referred to as BGP paths.) This means that, for a particular network prefix—for example, 10.0.0.0/8—the local BGP router might have recorded multiple available paths: one through any of its BGP peers. For each prefix, the BGP process uses a path selection algorithm to select the best available path from all those learned from its peers. Once the best path has been selected, that path becomes the candidate route from the BGP protocol for inserting into the active routing table.

Each BGP path includes several attributes that are used by the BGP path selection process to determine which path is the best available path. These attributes can also be used in user-defined routing policies applied to BGP; these can allow the router to perform additional actions on a matching path, such as determining whether to accept or reject a route announcement.

One of the most commonly used BGP path attributes is the AS path. The AS path lists each of the AS by which the prefix has been announced, and is used to prevent routing loops. The AS path is read from right to left, where the right-most AS is the AS that originated the network prefix (that is, it was the first AS to announce reachability for this prefix). This AS is known as the origin AS.

As a network prefix is advertised between AS, each AS prepends its own AS number to the AS path. For example, the AS path “4 3 2 1” indicates that AS 1 originated the network prefix. The prefix was advertised from AS 1 through AS 2, then from AS 2 through AS 3, and finally from AS 3 through AS 4.

Other BGP path attributes include origin, next hop, multi-exit discriminator (“med”), local preference (“local pref”), atomic aggregate, and aggregator. These attributes are described in more detail in another section of this document.