Nokia BGP for Internet Routing flashcards
Nokia BGP for Internet Routing · Build recall with a focused review session.
32 cards. Read the term and the definition here. The deck at the top of this page is optional.
- BGP transport port
- BGP peers establish a TCP session on port 179. All BGP messages (OPEN, UPDATE, KEEPALIVE, NOTIFICATION) travel inside that TCP connection, which provides reliable, ordered delivery.
- BGP UPDATE message
- Carries withdrawn routes, path attributes, and NLRI for one attribute set. Multiple prefixes with identical attributes share one UPDATE; differing attributes require separate UPDATEs.
- ORIGIN attribute order
- Well-known mandatory attribute with values IGP (0), EGP (1), and INCOMPLETE (2). Lower values are preferred in best-path selection, so IGP beats EGP beats INCOMPLETE.
- iBGP split horizon
- A route learned from one iBGP peer is not re-advertised to another iBGP peer. This forces a full mesh or route reflection/confederation to distribute routes within an AS.
- BGP FSM states
- Idle, Connect, Active, OpenSent, OpenConfirm, Established. Active means the router is retrying the TCP connection, not that the session is operational.
- eBGP multihop requirement
- eBGP defaults to TTL 1, so peers must be directly connected. Multihop eBGP needs explicit multihop configuration, a route to the peer, and usually an update-source loopback.
- BGP router ID uniqueness
- The BGP identifier must be unique and nonzero within the AS. Nokia SR OS derives it from the configured router-id or the highest loopback address if unset.
- TCP MD5 authentication
- RFC 2385 signs TCP segments with a shared key per neighbor. Mismatched keys cause OPEN failures and NOTIFICATION code 6 subcode 5 or 7, closing the session.
- LOCAL_PREF default
- Well-known discretionary attribute, default 100, exchanged only among iBGP peers. Higher values are preferred; it is never sent to eBGP peers and is ignored when received from them.
- MED comparison scope
- MED is compared only between paths from the same neighboring AS by default. Enable always-compare-med to compare across ASes; MED is non-transitive and not propagated beyond the receiving AS.
- eBGP over iBGP rule
- When LOCAL_PREF, AS_PATH, ORIGIN, and MED tie, an eBGP-learned path is preferred over an iBGP-learned path. This is a standard step in the best-path algorithm.
- Multipath BGP requirement
- ECMP installation requires equal attributes up to the multipath step. Configure maximum-ipv4-routes and multipath (ebgp or ibgp) to allow multiple next hops for one prefix.
- LOCAL_PREF propagation rule
- LOCAL_PREF is iBGP-only. It is not included in updates to eBGP peers, and a value received from an eBGP peer is ignored. Default value is 100 on Nokia SR OS.
- NO_EXPORT community
- Well-known community 0xFFFFFF01 (NO_EXPORT). Routes carrying it must not be advertised outside the receiving AS or confederation boundary, limiting their propagation scope.
- BLACKHOLE community
- Community 65535:666 used for RTBH. A host route (/32 or /128) tagged with it is dropped by the ISP, typically triggered by an iBGP-injected route at the customer edge.
- SR OS policy evaluation
- Policy statements contain numbered entries evaluated in ascending order; the first matching entry wins. Each entry has from (match) and to (action) clauses plus a default-action.
- Route reflector reflection rules
- Routes from a client go to all other clients and non-clients. Routes from a non-client go only to clients and other non-clients, never back to the originator.
- ORIGINATOR_ID and CLUSTER_LIST
- ORIGINATOR_ID holds the router ID of the route originator; CLUSTER_LIST records cluster IDs traversed. Both prevent reflection loops, and CLUSTER_LIST is checked like AS_PATH.
- BGP confederation AS_PATH
- Confederations use sub-AS numbers internally and AS_PATH_CONFED segments. External peers see only the confederation identifier as the AS number, hiding internal structure.
- Default BGP timers
- Keepalive 60 seconds and hold time 180 seconds, a 3:1 ratio. Hold time negotiates to the lower of the two peers' values; hold time 0 disables keepalives.
- BGP graceful restart
- RFC 4724 capability that preserves forwarding during a control-plane restart. It requires negotiation and a restart time advertised in the OPEN message.
- AS_PATH prepend for inbound TE
- Prepending extra copies of your own AS lengthens the path and makes it less preferred for remote ASes. It is the standard outbound tool to influence inbound traffic.
- MED inbound traffic engineering
- MED biases inbound traffic when both paths come from the same neighboring AS. Across different ISPs MED is not comparable, so use prepending or ISP communities instead.
- ISP community-based TE
- Each ISP publishes its own community scheme. A customer tags routes with those communities, and the ISP's inbound policy interprets the tag to set local preference or prepend.
- RTBH trigger route
- A /32 or /128 host route tagged with community 65535:666, usually injected via iBGP from the customer edge. The ISP blackholes matching traffic to mitigate attacks.
- RPKI validation states
- A route is Valid if a ROA matches prefix and origin AS, Invalid if a ROA covers the prefix but the origin differs or length exceeds maxLength, and NotFound if no ROA covers it.
- Bogon prefix filtering
- Inbound filters drop unallocated and reserved space such as RFC 1918 ranges, 0.0.0.0/8, 127.0.0.0/8, and 224.0.0.0/4. This baseline prevents accidental or malicious leaks.
- Maximum-prefix protection
- Limits prefixes accepted from a neighbor, with a warning threshold and optional shutdown or idle-timeout. It protects against session-wide overload from a misbehaving peer.
- GTSM for BGP
- RFC 5082 Generalized TTL Security Mechanism checks the TTL of BGP packets. Both ends set TTL to 255 and verify the expected hop count, blocking spoofed off-path sessions.
- MP-BGP IPv6 encoding
- IPv6 NLRI travels in MP_REACH_NLRI and MP_UNREACH_NLRI with AFI 2 and SAFI 1 for unicast. The next hop is carried inside MP_REACH_NLRI, not the legacy NEXT_HOP attribute.
- SR OS per-family BGP
- Enable families per neighbor with 'family ipv4' and 'family ipv6'. Both can share one TCP session or use separate sessions, affecting convergence and policy granularity.
- SR OS BGP hunt command
- 'show router bgp routes <family> hunt' lists all candidate paths for a prefix and identifies the selected best path, useful for verifying policy and best-path outcomes.