JNCIS-SP key terms

A router with interfaces in two or more OSPF areas. It keeps a separate link-state database for each area and injects type 3 summary LSAs so other areas can reach its connected networks.

OSPF area 0.0.0.0. Every other area must attach to it directly or through a virtual link, and inter-area traffic is normally forwarded across it.

An OSPF area that blocks type 5 external LSAs. Routers inside reach destinations outside the area through a default route advertised by the area border router.

An OSPF stub area that still allows external routes, carried inside it as type 7 LSAs. The area border router translates those into type 5 LSAs for the rest of the domain.

An OSPF link-state advertisement generated by an area border router that advertises a prefix from one area into another. It describes reachability rather than topology.

The metric OSPF assigns to a link, by default the reference bandwidth divided by interface bandwidth. Lower costs are preferred, and the sum of costs along a path gives the route metric.

On a multiaccess OSPF segment, the router elected to form adjacencies with all other routers and to originate the network LSA, which reduces the number of adjacencies required.

A check that OSPF packets come from a trusted neighbor, using a shared plaintext, MD5 or SHA key configured on the interface or for the whole area.

The ISO address that identifies an IS-IS router: an area identifier, a system ID, and a one-byte selector that is zero for routing purposes.

A router that forms adjacencies in the Level 2 backbone and carries inter-area prefixes. A Level 1/2 router also participates in a single Level 1 area.

The router elected on a broadcast IS-IS link to originate the pseudonode LSP and send periodic CSNPs. Election is preemptive, so a higher priority router takes over.

A link-state PDU originated by the designated intermediate system that represents a broadcast segment as a virtual node, so the segment appears as a single node in the SPF tree.

An IS-IS metric encoding using 24-bit values, which supports very large interface costs and is required for traffic engineering and for telling links apart in large networks.

A flag in a Level 1 LSP set by a Level 1/2 router to tell its area that a path toward other areas exists, so Level 1 routers can point a default route at it.

Advertising selected Level 2 or inter-area prefixes into a Level 1 area so that Level 1 routers can reach those destinations directly instead of through a default route.

A BGP session between routers inside the same autonomous system. Routes learned from one internal peer are not passed to another internal peer unless route reflectors or a full mesh are used.

The ordered list of autonomous systems a BGP route has crossed. It is used to detect loops and as a tie-breaker, where a shorter list is preferred.

The BGP attribute naming the IP address traffic should be forwarded to for a route. For external peers it is usually the peer address; internal peers often need it resolved through the IGP.

A well-known discretionary attribute used inside an autonomous system. The route with the highest value wins, and the value is not carried to external peers.

An optional non-transitive attribute that suggests which entry point into an autonomous system is preferred. Lower values win, and it is compared only among routes from the same neighbor AS.

An internal BGP router permitted to reflect routes learned from one client to other clients and peers, removing the need for every internal router to peer with every other.

A large autonomous system split into internal sub-ASes that run external-style sessions with each other while still appearing as one autonomous system to the outside world.

A tag carried in the BGP community attribute and matched in routing policy to group routes and apply actions, such as the well-known no-export and no-advertise values.

The ordered comparison BGP applies to candidate routes: weight, local preference, locally originated, AS path length, origin, MED, external over internal, then IGP cost to next hop and router ID.

A protocol that exchanges label bindings between MPLS routers so that a label switched path follows the IGP shortest path toward each destination prefix.

A set of MPLS labels pushed onto a packet and processed from the top down. Stacking lets a service label, such as a VPN label, be carried over transport labels.

The router one hop before the egress removes the transport label so the egress receives an unlabeled or single-label packet, saving it one lookup.

A signalling protocol that reserves bandwidth and sets up explicitly routed label switched paths with constraints, supporting traffic engineering and fast reroute.

A separate routing and forwarding table on a provider edge router that holds the routes and interfaces of one VPN customer, isolated from other customers' tables.

An eight-byte value prepended to a customer prefix to make it unique inside BGP, so overlapping addresses from different VPNs can coexist in the same table.

A BGP extended community attached to VPN routes that determines which VRFs import those routes and which VRFs export them.

The multiprotocol BGP address family that carries VPN routes, consisting of an eight-byte route distinguisher followed by an IPv4 prefix, plus a label and extended communities.

A link that carries several VLANs by inserting a four-byte tag holding a VLAN identifier into each frame, letting one port serve many broadcast domains.

The VLAN whose frames cross an 802.1Q trunk without a tag. Both ends must agree, because a mismatch merges two VLANs and can create a loop.

The switch elected as the top of the spanning tree using the lowest bridge priority and then the lowest MAC address; all path costs are measured from it.

Several physical links bundled so that they act as one logical link, adding bandwidth and redundancy. LACP is used to negotiate which links belong to the bundle.

A routing policy term that lists match conditions and the actions to take, such as accept, reject or set an attribute, and is evaluated in order with other terms.

A lightweight hello protocol that detects failures between two routers within milliseconds and can trigger fast routing or MPLS recovery.

A router keeps its control-plane state synchronized with a standby engine so that a switchover does not reset routing protocol sessions or adjacencies.

A capability that lets a router restart its control plane while still forwarding traffic, with helper neighbors holding routes and labels until sessions are re-established.