Border Gateway Protocol (BGP) is designed to manage the routing of packets across the Internet. This is done through exchange of routing information between certain routers. This protocol ensures correct routing for optimal reachability, while directing packets between systems. The latter are essentially anonymous networks (AS) that are managed by service providers or single-based enterprises. The traffic is routed within single networks is known as internal BGP. However, BGP also connects one autonomous system to other autonomous systems; known as external BGP.
BGP ensures network stability, which results in routers adapting to effectively send packets through other reconnections. This, of course, is if one Internet path goes down and is unable to handle the routing of packets. Similarly, BGP bases its decisions on existing paths, along with network guidelines and/or policies. These rules are configured and governed by network administrators. For each BGP router, a standard routing table is maintained to transmit packets in transit. These tables are also used to correlate with separate routing tables, which are known as routing information bases (RIB).
RIB is essentially a data table, which is stored on servers within the BGP routers. RIB not only contains route information, but also manages updates to tables as changes happen. In a nutshell, the RIB consists of vital information stemming from connected external peers – along with internal peers. BGP protocols are also based on TCP/IP and client-based servers to effectively transmit routing information. Client-servers play a pivotal role in this process, since they initiate BGP sessions by sending requests to specific servers.
BGP protocol is an intricate and detailed process that has many multi-faceted layers. However, the core aspect revolves around sending updated router table information in the event something changes. This too only pertains to affected information as BGP does not have built-in discovery mechanisms. This means all peer connections must be set up manually – with programmed addresses at both points.
BGP also makes path decisions based on its characteristics and hop counts. Similarly, it is based on current reachable levels, however, can also facilitate multiple paths if present. With the latter, BGP is then used to communicate desired organizational paths in and out of their networks. This usually pertains to major hosting facilities, but also relates to P2P agreements and route advertisement behaviors. As a form of machine-learning, BGP even defines arbitrary tags – which are known as communities. These tags are utilized to control certain agreements and path behaviors among peers or shared routers.
BGP-4 current versions are designed to support classless inter-domain routing (CIDR). Similarly, these versions also support IPv6, which were certified back in 2005. CIDR and IPv6 enable the continued use of IPv4. CIDR is also able to facilitate more addresses within networks, which are far superior to current IP address assignments and schemes.
USD 16 / Mo
USD 92 / Mo