What is BGP Path Selection?

Understanding BGP Path Selection

BGP path selection is the deterministic, ordered process the Border Gateway Protocol uses to choose a single best route to a destination prefix when multiple paths are available. BGP evaluates a list of attributes in strict sequence, stopping at the first criterion that breaks the tie, and installs the winning route in the routing table while advertising it to neighbors.

The selection follows a defined order of attributes. On Cisco routers the algorithm prefers, in turn: highest Weight (Cisco-proprietary, local to the router), then highest Local Preference (consistent within an autonomous system), prefer locally originated routes, then shortest AS_PATH, lowest Origin type, lowest Multi-Exit Discriminator (MED), eBGP over iBGP, lowest IGP metric to the next hop, and finally tie-breakers such as oldest route and lowest router ID. Network operators manipulate these attributes to engineer traffic along preferred links.

This matters for security and reliability because BGP governs how traffic flows across the global internet between autonomous systems. The path-selection process determines which route is trusted and used, so misconfiguration or malicious manipulation of attributes can divert traffic. Route hijacking and route leaks exploit this by advertising more attractive paths, drawing traffic to an attacker for interception or causing outages. Understanding the selection order is essential to designing resilient routing and defenses like prefix filtering and RPKI origin validation.

For example, a multihomed enterprise connects to two ISPs and wants outbound traffic to prefer ISP A. The engineer sets a higher Local Preference on routes learned from ISP A, which BGP evaluates before AS_PATH length, so even if ISP B advertises a shorter AS path, the router still selects ISP A as the best path. Should ISP A fail, BGP re-runs the selection process and automatically shifts traffic to ISP B, maintaining connectivity.