Nokia BGP Fundamentals for Services flashcards

39 cards. Read the term and the definition here. The deck at the top of this page is optional.

BGP transport port

BGP runs over TCP port 179, providing reliable ordered delivery. BGP itself has no retransmission or sequencing and relies on TCP plus KEEPALIVE messages for liveness.

BGP message types

Four message types: OPEN (session parameters and capabilities), UPDATE (withdrawn routes, path attributes, NLRI), KEEPALIVE (19-byte header only), and NOTIFICATION (error code and subcode, then close).

Well-known mandatory attributes

ORIGIN, AS_PATH, and NEXT_HOP must appear in every UPDATE carrying reachability. Their absence triggers a NOTIFICATION with an Update Message Error, closing the session.

AS_PATH loop detection

A BGP speaker rejects any route whose AS_PATH already contains its own AS number. AS_SEQUENCE segments are ordered hops; AS_SET segments count as one hop for length.

iBGP split-horizon rule

A route learned from one iBGP peer is not re-advertised to another iBGP peer. This forces full mesh or route reflectors/confederations to distribute iBGP routes within an AS.

ORIGIN attribute values

ORIGIN is well-known mandatory: 0 = IGP (i), 1 = EGP (e), 2 = INCOMPLETE (?). Lower values are preferred during best-path selection, so IGP-originated routes beat redistributed ones.

BGP state machine

States: Idle, Connect, Active, OpenSent, OpenConfirm, Established. Stuck in Active usually means TCP connect failure; stuck in OpenSent or OpenConfirm indicates OPEN or capability mismatch.

eBGP multihop and TTL

eBGP defaults to TTL 1, requiring directly connected peers. Multihop eBGP needs ebgp-multihop and a route to the peer address; TTL security (GTSM) is often added.

BGP hold time negotiation

Each OPEN carries a hold time; the effective value is the lower of the two. KEEPALIVE is typically one-third of hold time. Hold time 0 disables keepalives entirely.

BGP router ID

A 32-bit BGP identifier, usually the highest loopback IPv4 address. Duplicate router IDs break route-reflector loop detection and ORIGINATOR_ID checks, so uniqueness matters.

BGP authentication options

TCP MD5 (RFC 2385) and TCP-AO (RFC 5925) authenticate BGP sessions at the TCP layer. Mismatched keys cause connection resets or Bad Peer AS notifications rather than BGP-level errors.

iBGP versus eBGP next-hop

eBGP-learned routes carry the eBGP peer address as NEXT_HOP. iBGP preserves that original next-hop, so the IGP must resolve it unless next-hop-self or a route reflector is used.

LOCAL_PREF scope

LOCAL_PREF is exchanged only between iBGP peers and never sent to eBGP neighbors. It expresses outbound preference within the local AS and is compared before AS_PATH length.

MED scope and comparison

MED is sent between eBGP peers as an inbound hint. By default it is compared only among paths from the same neighbor AS and is not propagated across AS boundaries.

Weight attribute

Weight is a Cisco-local value, default 0, never carried in any BGP message. It is the first best-path criterion and outranks LOCAL_PREF, so it is used for local route preference only.

Best-path order summary

Weight, LOCAL_PREF, locally originated, shortest AS_PATH, lowest ORIGIN, lowest MED, eBGP over iBGP, lowest IGP metric to next-hop, oldest eBGP path, lowest router ID, lowest neighbor address.

BGP multipath ECMP

bgp maximum-paths installs multiple equal paths when the first N best-path criteria tie. The IGP-cost criterion is skipped for multipath candidates, enabling ECMP across parallel links.

AS-path prepending

Prepending repeats the local AS number in AS_PATH to lengthen the path seen by neighbors. It influences inbound traffic selection because shorter AS_PATH is preferred earlier than MED.

network versus redistribute

A network statement originates a prefix only if a matching route exists in the routing table. Redistribute injects routes from other protocols. Both can be gated by route-maps.

Prefix-list matching

A prefix-list entry matches prefix and mask length; ge and le specify the allowed range. For example, permit 10.0.0.0/8 ge 16 le 24 matches subnets of 10/8 with masks 16 through 24.

AS-path regex anchors

In AS-path access-lists, ^$ matches locally originated routes, ^65001_ matches paths starting with 65001, and _65001_ matches 65001 anywhere in the path.

Well-known communities

NO_EXPORT (0xFFFFFF01) prevents advertisement outside the AS or confederation. NO_ADVERTISE (0xFFFFFF02) prevents advertisement to any peer. NO_PEER (0xFFFFFF04) blocks advertisement to peers.

RPKI origin validation states

A route is Valid if a ROA covers its prefix and origin AS, Invalid if a ROA covers the prefix but not the origin AS, and NotFound if no ROA exists. Invalid routes are usually dropped or deprioritized.

Route reflector reflection rules

An RR reflects client routes to other clients and non-clients, and non-client routes to clients only. Non-client to non-client is not reflected, preserving standard iBGP split-horizon behavior.

ORIGINATOR_ID and CLUSTER_LIST

The RR adds ORIGINATOR_ID (the originating router ID) and CLUSTER_LIST (cluster IDs traversed). A router seeing its own router ID or cluster ID in these attributes discards the route.

Confederation AS_PATH handling

Within a confederation, member-AS numbers are carried in AS_CONFED_SEQUENCE and not prepended to the public AS_PATH. They are stripped when the route leaves the confederation.

Peer groups and templates

Peer groups apply shared outbound policy to many neighbors, cutting configuration and UPDATE generation. Peer templates separate session parameters from policy and are inherited by neighbors.

MP_REACH and MP_UNREACH

MP-BGP carries non-IPv4-unicast families in MP_REACH_NLRI (attribute 14) and MP_UNREACH_NLRI (attribute 15). The AFI/SAFI pair identifies the address family and subsequent address family.

AFI/SAFI pairs

Common pairs: IPv4 unicast (1/1), IPv6 unicast (2/1), VPNv4 (1/128), VPNv6 (2/128), EVPN (25/70), labeled unicast (1/4). Both peers must negotiate the family in OPEN.

Route Distinguisher and Target

The RD (8 bytes) makes identical prefixes unique in VPNv4. The Route Target extended community controls VRF import and export, determining which VPN sites can reach each other.

EVPN route types

EVPN route types include Type 1 Ethernet Auto-Discovery, Type 2 MAC/IP Advertisement, Type 3 Inclusive Multicast, Type 4 Ethernet Segment, and Type 5 IP Prefix for inter-subnet routing.

MRAI timer

MinRouteAdvertisementInterval dampens UPDATE churn. Cisco defaults are 30 s for eBGP and 0 s for iBGP (5 s in some implementations), trading convergence speed for stability.

BGP graceful restart

Graceful restart (RFC 4724) uses a capability and End-of-RIB marker. The receiving peer keeps stale routes while the restarting peer rebuilds, provided forwarding state is preserved.

BGP additional paths

Add-path (RFC 7911) advertises multiple paths per prefix using path identifiers. It improves convergence and multipath availability without waiting for a best-path change.

BGP next-hop tracking

When the IGP route to a BGP next-hop disappears, the BGP route becomes invalid. BFD accelerates this detection to sub-second, triggering faster reconvergence or PIC repair.

Max-prefix protection

A per-neighbor maximum-prefix limit with a warning threshold and shutdown or restart action protects the RIB from a misbehaving peer flooding routes. It is standard on SP edge sessions.

BGP NOTIFICATION codes

Codes: 1 Message Header Error, 2 Open Message Error, 3 Update Message Error, 4 Hold Timer Expired, 5 Finite State Machine Error, 6 Cease with subcodes such as Administrative Shutdown or Reset.

SP edge inbound filtering

SP edge routers should filter customer announcements with prefix-lists tied to registered prefixes, plus max-prefix limits and RPKI validation, to block leaks and hijacks at the boundary.

BGP session troubleshooting order

Check session state first, then received and advertised prefix counts, then the specific prefix, then policy on both sides. Verify return path exists because asymmetric routing is normal.