Border Gateway Protocol

Border Gateway Protocol (BGP) routing protocol in a kind of autonomous systems over TCP. BGP is the only protocol that is used to handle the size of the network like the Internet, is the only protocol that can deal well with having multiple connections to unrelated routing domains. BGP builds on the experience of EGP. The main function of the system is BGP and other BGP systems to exchange network reachability information. Network reachability information autonomous system (AS) includes information listed. This information is sufficient to construct a topography of AS connectivity from which routing loops, and can be implemented on the AS-level policy decisions
Interpretation
(BGP / BGP4: Border Gateway Protocol , Border Gateway Protocol)
BGP4 provides a sets new mechanism to support classless inter-domain routing. These include announcements support network prefix, BGP network canceled the concept of "class". BGP-4 also introduces mechanisms which allow aggregation of routes, including the set of AS paths. These changes provide support for the proposed super-net programs. BGP-4 uses a routing vector routing protocol, when configuring BGP, each autonomous system administrator to select at least one router as the "BGP spokesman" of the autonomous system.
1989, released the main external gateway protocol: border routing protocol (BGP), BGP-4 is a new version released in 1995.
BGP routing protocol used in the execution four kinds of packets: packet opens (open), the update packet (update), survival packet (keepalive), advertising packet (notification).
The structure and function of BGP
BGP is used to exchange routing information between the different autonomous systems (AS). When two AS needs to exchange routing information, each AS must specify a node running BGP, and the other to represent AS AS exchange routing information. This node can be a host. But usually a router to perform BGP. AS utilizing two BGP routers exchange information is referred to a border gateway (Border Gateway) or boundary router (Border Router) [1].
Because may connect with different AS, there may be multiple border routers running BGP within an AS. The same autonomous system (AS) two or more peer BGP is run between entities called IBGP (Internal / Interior BGP). Different home AS like BGP runs between entities called EBGP (External / Exterior BGP). Routers exchange information with other border on the AS border router is called AS (border / edge router). In the Internet Operating System (Cisco IOS), the distance of the route is 200 IBGP announcement, priority routing (IGP) is lower than the advertised EBGP and any interior gateway protocol. Other router implementations, the priority order is higher than EBGP IGP, but was higher than the IGP IBGP.
Gateway Routing Protocol BGP are external, may be implemented without inter-domain routing loop between autonomous systems. BGP routing protocol to communicate with the main Internet wide area network, such as different provinces, routing between different countries mostly rely on BGP. BGP can be divided into IBGP (Internal BGP) and EBGP (External BGP). The BGP neighbor relations (or communication peer / peer entity) is achieved by manual configuration, through TCP (port 179) session interactive data between peer entities. Transmitting 19 bytes of BGP router to remain viable periodic keep-alive messages to maintain the connection (default period is 30 seconds). In the routing protocol, only BGP uses TCP as the transport layer protocol.
IETF has developed a number of recommendations for BGP, are:
RFC 4271: BGP protocol version currently being used, called BGP4.
RFC 1654: The first specification BGP4 protocol.
RFC 1105, RFC 1163, RFC 1267 , RFC1771: BGP versions prior to BGP4.
Feature
BGP are external or inter-domain routing protocol. The main objective is to provide a guarantee for BGP routing information in the communication between routers in different AS. Neither purely BGP distance vector protocol, nor a pure link-state protocol, commonly referred to as path vector routing protocol. This is because BGP at the same time accessibility to publish a destination network, contains a list of IP packets to the destination network in the course must pass the AS. When path vector information is useful, because simply look at BGP routing updates AS number can effectively avoid loops. BGP is no restriction on the network topology, characterized by comprising:
(1) up to realize a communication of information between autonomous systems, communication networks. BGP is an external gateway protocol, allowing an AS to communicate with another AS. BGP routing information allowing an AS to advertise its internal network reachability information to other AS, the AS or by other networks reachable. Meanwhile, AS can understand this information from another AS. Similar to distance vector routing protocol, BGP for each destination network is provided next hop (next-hop) node.
(2) a plurality of coordination between BGP routers. If there are multiple routers in the BGP router is used with a series of other autonomous systems communicating peer routers, BGP can coordinator, routers make these consistent routing information within an autonomous system.
(3) BGP supports policy-based routing (policy-base routing). General distance vector routing protocol routing exact routes are advertised in the local. The implementation strategy BGP can be selected by the local administrator. BGP router can be configured as a network reachability between different policy domains and domain.
(4) a reliable transport. BGP routing information transmission using the reliable TCP protocol.
(5) path information. When reachability information for BGP destination network, the destination network designated to the processing of next hop information, advertisement further comprises the vector path (path vector), i.e. the need to go through this destination AS network list, enable the recipient to understand the path information destined for the destination network.
(6) incremental updates. BGP routing updates do not need all the packets transmitted in full routing information database, only a full exchange of information at startup. Subsequent routing updates advertise only change information network. This information network called incremental change (delta).
(7) BGP support untyped preparation (CIDR) and VLSM way. All are represented in the network advertised network prefix plus the subnet mask of the way.
(8) routing aggregation. BGP allows the sender to gather together the routing information, represented by a plurality of entries associated destination network to save network bandwidth.
(9) BGP also allow the recipient to identify and authenticate packets, to verify the identity of the sender.