Nokia VPRN key terms

A separate routing table instance inside one router, with its own interfaces, routes, and protocol sessions. It lets overlapping customer address space coexist on shared hardware, which is the basis of VPRN service separation.

Nokia's name for an IP-VPN service built on BGP/MPLS VPN principles. The provider edge holds per-customer VRFs, exchanges VPN routes with MP-BGP, and forwards customer traffic across an MPLS or SR core.

An 8-byte prefix prepended to a customer IPv4 prefix so the resulting VPN-IPv4 route is unique across all tenants. It identifies the originating VRF but carries no forwarding meaning by itself.

A BGP extended community attached to VPN routes that controls import and export between VRFs. A VRF imports routes carrying its import targets, so matching targets creates the actual membership of a VPN.

Multiprotocol BGP, which carries address families beyond plain IPv4 unicast, including VPN-IPv4 and IPv6. In a VPRN it distributes customer routes between provider edge routers with next-hop and label information.

The MP-BGP address family whose entries combine a route distinguisher with an IPv4 prefix. Provider edge routers exchange these entries so customer routes stay distinguishable even when prefixes overlap.

A policy that rewrites the next hop of routes advertised to a peer so they point at the advertising router. It forces traffic through the local provider edge and is needed for correct VPN label allocation.

The process by which a provider edge assigns an MPLS label to each VPN route it advertises, so returning traffic can be mapped back to the correct VRF. The label is advertised with the route in MP-BGP.

An MPLS behaviour where the second-to-last router removes the outer transport label before forwarding to the egress. The egress router then sees only the service label, reducing its lookup work.

A protocol that distributes labels for IGP-learned prefixes so routers can build label switched paths hop by hop. It provides the transport tunnels that carry VPRN traffic across an MPLS core.

A source-routing architecture where the headend encodes a path as a list of segment identifiers, removing the need for per-hop label signalling. Nokia SR OS supports SR-MPLS and SRv6 as transport for VPN services.

A label or IPv6 address representing a network instruction such as a node, adjacency, or service. A stack of SIDs in the packet header steers traffic without per-hop state.

Topology-Independent Loop-Free Alternate, a fast reroute method that precomputes a backup path using segment routing to guarantee loop freedom. It protects against link and node failures with sub-50-millisecond convergence.

A lightweight hello protocol that detects path failures in milliseconds independently of routing protocols. It is commonly bound to IGP or BGP sessions and to VPRN peerings to trigger fast reconvergence.

A link-state interior gateway protocol that floods link information and runs shortest-path-first computation. In Nokia networks it provides the underlay reachability for loopbacks that BGP and MPLS depend on.

A link-state interior gateway protocol operating directly at layer two, widely used in service provider cores. It scales well and carries traffic engineering extensions for MPLS and segment routing.

Importing routes learned by one protocol into another, such as connected or OSPF routes into BGP. It requires policy to prevent loops and to control which prefixes enter the VPRN routing table.

A set of match-and-action rules applied to routes on import or export. Policies select, tag, modify, or reject routes and are central to controlling VPRN route exchange and redistribution.

A layer-two broadcast domain identified by an 802.1Q tag on trunk links. Service providers use VLANs to separate customer traffic on shared access ports and to map traffic into services.

Encapsulating a customer VLAN tag inside a provider VLAN tag so many customers can share one provider port. It preserves customer tagging while giving the provider a per-customer outer identifier.

The Nokia term for the point where a service attaches to an access port, such as a port plus VLAN combination. It defines the encapsulation and the ingress and egress behaviour for customer traffic.

A Nokia object that defines a tunnel from one service router to another, carrying service traffic over MPLS, GRE, or another transport. Services bind to SDPs to reach remote provider edge nodes.

A multipoint layer-two service that makes remote sites appear on one bridged LAN. The provider learns MAC addresses per site and forwards frames only toward the segment where the destination was learned.

The process of recording a frame's source MAC and its ingress point so later frames to that address can be forwarded selectively. Entries age out after inactivity to adapt to moves.

A loop-prevention rule that stops a router or bridge from advertising a route or forwarding flooded traffic back out the interface it arrived on. In VPLS it prevents frames looping across pseudowires.

A layer-two protocol that blocks redundant links to create a loop-free topology. Modern variants converge quickly, and Nokia access nodes also support MSTP and fast ring alternatives.

Bundling several physical links into one logical link using LACP or static configuration. It increases bandwidth and provides link-level redundancy while presenting a single interface to services.

A LAG whose member links terminate on two different chassis that coordinate to act as one. It gives access devices redundant connections without relying on spanning tree to block a link.

Nokia's controller and orchestration suite for IP and optical networks. It provides service provisioning, path computation, assurance, and closed-loop automation across multivendor domains.

A data modelling language that defines the structure, types, and constraints of device configuration and state. NSP and NETCONF rely on YANG models to configure and read network devices consistently.

A protocol for configuring network devices using structured XML data over a secure session. It supports candidate configurations, validation, and transactional commits, making changes safer than CLI scripting.

Google Network Management Interface, a gRPC-based protocol for reading, writing, and subscribing to device state using paths. It is widely used for streaming telemetry and automated configuration.

Continuously streaming operational data from devices using structured models rather than polling. It gives near-real-time visibility into counters, interface state, and protocol sessions for assurance and automation.

Automation that observes network state, compares it with intent, and takes corrective action without manual steps. Assurance data triggers remediation such as rerouting or reconfiguring a service.

An approach where operators declare the desired outcome, such as reachability or latency, and the system computes and enforces the configuration. It shifts effort from device commands to service-level objectives.

The Nokia SR OS command interface built on YANG models, using a hierarchical, candidate-and-commit style. It aligns CLI operations with model-driven configuration and supports compare and rollback.

A filter that rate-limits traffic destined to the router's own CPU, protecting routing protocols and management from floods. It classifies packets by protocol and applies separate rate limits per class.

An ordered set of match criteria and permit or deny actions applied to packets on an interface or service. It filters traffic by fields such as source, destination, protocol, and port.

A spoofing defence that checks whether a packet's source address is reachable through the interface it arrived on. Strict mode drops packets that fail the check; loose mode only verifies reachability somewhere.

A suite of protocols that encrypts and authenticates IP packets between peers. It protects traffic over untrusted links and is often used for overlay VPNs or to secure management connectivity.

An IEEE 802.1AE standard that encrypts and authenticates Ethernet frames hop by hop on a link. It protects provider links and data center fabrics against eavesdropping and tampering at layer two.

A filter restricting which source addresses and protocols may reach a device's management interfaces. It limits exposure of SSH, SNMP, and NETCONF to approved operator networks.

A router that connects a data center fabric to the WAN or to other fabrics, hosting Layer 2 and Layer 3 boundaries. It typically terminates VXLAN or EVPN and maps tenant networks into VPN services.

A BGP control-plane solution for delivering Ethernet services, using route types to advertise MAC and IP reachability. It replaces flood-and-learn with controlled signalling and supports multihoming and proxy ARP.

A tunnelling scheme that encapsulates Layer 2 frames in UDP with a 24-bit network identifier, allowing large-scale tenant separation. It is usually paired with EVPN for control-plane learning.

The device that encapsulates and decapsulates VXLAN traffic, mapping frames to and from the overlay tunnel. Each endpoint has an IP address used as the tunnel source and destination.

The same IP and MAC address configured as the default gateway on every leaf switch, so hosts keep one gateway when they move. It removes the need for first-hop protocol changes during migration.