Nokia BGP for Internet Routing key terms

A router that runs the Border Gateway Protocol, maintains peering sessions with other routers, and exchanges Network Layer Reachability Information describing paths to destination prefixes.

A remote router with which a local BGP speaker has established a configured session, identified by its IP address and autonomous system number.

A set of IP networks and routers under one administrative authority that presents a single, consistent routing policy to the Internet, identified by a unique AS number.

A BGP session between two routers in the same autonomous system. Learned routes are not re-advertised to other iBGP peers, so full mesh or route reflectors are required.

A BGP session between routers in different autonomous systems. It carries interdomain routes and normally prepends the local AS number to the AS_PATH.

Network Layer Reachability Information: the set of destination prefixes carried in a BGP UPDATE message, each paired with the path attributes that describe how to reach it.

The BGP message type that advertises new or withdrawn prefixes together with their path attributes. It is the primary vehicle for route exchange between peers.

A short periodic BGP message sent to confirm that a session is still operational. It is typically sent at one third of the configured hold time.

The finite state machine governing a BGP session, moving through Idle, Connect, Active, OpenSent, OpenConfirm, and Established states as the TCP session and capability negotiation proceed.

The negotiated maximum interval, in seconds, that a BGP speaker will wait for a KEEPALIVE or UPDATE message before declaring the peer down and resetting the session.

A 32-bit value, usually the highest loopback IPv4 address, that uniquely identifies a BGP speaker within its autonomous system and is carried in the OPEN message.

The first BGP message exchanged after TCP connection setup, carrying the version, local AS number, hold time, BGP identifier, and optional capability parameters.

The exchange of optional parameters in OPEN messages, allowing peers to agree on features such as multiprotocol address families, route refresh, and four-octet AS numbers.

An extension that lets BGP carry reachability for multiple address families, such as IPv4 and IPv6 unicast, VPN routes, and multicast, within the same session.

A named category of routing information, such as IPv4 unicast or IPv6 unicast, that is enabled independently on a BGP session and carries its own policy and route selection.

Address Family Identifier and Subsequent Address Family Identifier: the two-part code in MP-BGP that identifies the exact type of reachability, for example IPv6 unicast or IPv4 labeled-unicast.

A well-known mandatory BGP attribute listing the autonomous system numbers a route has traversed. Shorter paths are preferred, and the list also serves as a loop-prevention mechanism.

A policy technique that inserts extra copies of the local AS number into an advertised AS_PATH, making the path appear longer and therefore less preferred to remote routers.

A well-known mandatory attribute giving the IP address of the router that should be used to reach the advertised prefix. Its handling differs between iBGP and eBGP.

A configuration on an iBGP session that rewrites the NEXT_HOP of reflected or learned routes to the local router's address, ensuring downstream peers can resolve the next hop.

An optional non-transitive attribute used to suggest to a neighboring autonomous system which of several entry points it should prefer for traffic destined to a given prefix.

A well-known discretionary attribute used inside an autonomous system to rank routes; higher values are preferred, and it is not advertised to external peers.

The deterministic BGP process that compares candidate routes using attributes in a fixed order, such as weight, local preference, AS_PATH length, origin, MED, and peer address, to choose one route per prefix.

A vendor-specific, locally significant value used first in Nokia and Cisco best-path comparison; higher weight wins, and it is never advertised to any peer.

A well-known mandatory attribute indicating how a route entered BGP: IGP, EGP, or INCOMPLETE. IGP is preferred over EGP, which is preferred over INCOMPLETE.

An optional transitive attribute carrying 32-bit tags on a route, used to group prefixes so that routing policy can be applied consistently across many routers.

Predefined community values with standard meanings, such as NO_EXPORT and NO_ADVERTISE, that restrict where a route may be advertised without requiring custom policy.

An ordered set of match-and-action statements applied to routes on import or export, controlling which prefixes are accepted, modified, or advertised on a BGP session.

A named list of IP prefixes with optional length ranges used inside routing policy to match destination networks precisely, for example permitting only a default route.

A regular-expression-based policy match against the AS_PATH attribute, commonly used to reject routes containing the local AS number or to accept only customer and peer paths.

The mechanism by which a router rejects any received route whose AS_PATH already contains its own AS number, preventing routing loops between autonomous systems.

An iBGP router allowed to reflect routes learned from one iBGP peer to other iBGP clients, removing the need for a full iBGP mesh inside an autonomous system.

A value carried in reflected routes that identifies the route reflector cluster, allowing receivers to detect and discard loops caused by misconfigured reflection.

An optional non-transitive attribute added by a route reflector recording the router ID of the original advertiser, used to prevent reflected routes from looping back.

A method of dividing one autonomous system into internal sub-ASes that peer with each other using eBGP-like rules, reducing iBGP mesh requirements while presenting one AS externally.

A mechanism that penalizes prefixes whose routes flap repeatedly, suppressing unstable routes for a period so that churn does not propagate across the network.

Connecting one network to the Internet through two or more providers or links, improving resilience and allowing traffic engineering through BGP policy.

The deliberate use of BGP attributes and policy to influence which paths inbound and outbound traffic take, balancing load and meeting performance goals.

The Resource Public Key Infrastructure: a signed registry that binds IP prefixes to authorized origin AS numbers, enabling routers to validate the origin of BGP announcements.

The process of comparing a received route's prefix and origin AS against RPKI data, marking it valid, invalid, or not found, and applying policy accordingly.

A signed RPKI object stating which AS number is permitted to originate a specific prefix, optionally with a maximum prefix length.

An extension that cryptographically signs the AS_PATH so receivers can verify that the path was not altered in transit, complementing origin validation.

A per-peer safeguard that resets or logs the session when the number of received prefixes exceeds a configured threshold, protecting router memory from misbehaving peers.

A BGP capability allowing a speaker to request re-advertisement of a peer's routes without resetting the session, so inbound policy changes can be applied cleanly.

A capability that lets a restarting router preserve forwarding state while BGP reconverges, reducing traffic loss during control-plane restarts.

An address family that carries interior gateway protocol topology and link-state information through BGP, commonly used to feed path computation and analytics systems.

A configuration container in SR OS that holds shared BGP neighbor parameters, such as AS number, timers, and policies, inherited by the peers configured beneath it.

An SR OS command displaying each peer's state, received and sent prefix counts, uptime, and flap statistics, used as the first check when troubleshooting sessions.