Cisco Technology, Inc.

Daniel C Tappan

James Neil Guichard

Eric Michel Levy-Abegnoli

Francois Le Faucheur

Eric C Rosen

W Scott Wainner

Richard W Bradford

Luc Louis Revardel

Parag Kumar Jain

Wojciech Dec

Earl Hardin Booth, Iii

Ole Troan

Christopher Metz

W Mark Townsley

Mohammed Azhar Sayeed

Danny Prairie

Bertrand Duvivier

Jean-Philippe Vasseur

Bruce Stuart Davie

Sunil K Srivastava

Robert Hanzl

Jeremy Lawrence

Simon Spraggs

A method, apparatus and computer program product for routing data within a packet-switched network using a PW wherein the PW is terminated directly on the layer-3 routing device such that certain services and applications can be utilized is presented. The method, apparatus and computer program product receives an encapsulated layer-2 Protocol Data Unit (PDU) from a pseudowire emulating a service. The encapsulation is removed from the encapsulated layer-2 PDU and a layer-2 circuit associated with the pseudowire is terminated. The circuit is treated as an interface and the PDU is forwarded based on upper layer protocol information within the PDU.

Psuedowire termination directly on a router

A method for supporting a GMPLS hierarchy through multiple routing instances in a network having multiple regions is disclosed. Each region has associated therewith a separate routing instance. The method includes initiating creating a link between two regions at an edge device of a first region and selecting a path between the two regions. A message is forwarded from the edge device to a network device coupled to a second region and a response is received from the network device at the edge device. A new LSP linking the two regions is inserted into a routing database for a routing instance associated with the second region.

Method for supporting a GMPLS hierarchy through multiple routing instances

IPv6 traffic may be carried through an MPLS IPv4 network without the use of IPv6-over-IPv4 tunneling. This provides great savings in overhead, signaling, and state information storage and also allows for routing through the MPLS IPv4 network to adjust in response to changes in network state. In one embodiment, an edge node of an MPLS IPv4 network resolves a destination IPv6 network of a received IPv6 packet to an MPLS label switched path. The resolution exploits received inter-domain routing information. This information identifies the IPv4 address of an egress node that is usable as a gateway to the destination network. Within the inter-domain routing information, the IPv4 address may be encoded in IPv6 format.

IPv6 over MPLS IPv4 core

Pseudowire termination directly on a router

A method and apparatus providing highly scalable server load balancing are disclosed. Data packets from a client are routed through one or more routers to a server load balancer, which is selected from among a plurality of server load balancers in a network. In response to receiving a request packet, a particular server site to process the client request is selected. A first path to a second router associated with the particular server site, and a second path to a server load-balancing device associated with the second router, are determined. A mapping of flow identifying information, associated with the packet, to a first label value that identifies the first path and to a second label value that identifies the second path, is created. The first label value and the second label value are stored in the packet. All subsequent packets associated with the client request are forwarded to the server load-balancing device based on looking up the first label value and second label value in the mapping. As a result, a network is scalable to process and load-balance numerous client requests, which are efficiently routed to the site, server load-balancer, and server that are handling the request.

Method and apparatus providing highly scalable server load balancing

A system and method are provided for separately distributing edge-device labels and routing information across routing areas of a computer network. Because the edge-device labels are distributed separately from network routing information, the process of distributing the edge-device labels does not preclude conventional edge-device address summarizations. Illustratively, a novel 'label mapping' LSA is employed for distributing the edge-device labels across routing areas. The label-mapping LSA may be embodied as an area-scope OSPF opaque LSA (type 10) or an IS-IS LSP containing TLVs of area scope. Advantageously, the present invention is generally applicable whenever label values are allocated to edge devices in a multi-area computer network and data is 'tunneled' through the network from one edge device to another.

Hierarchical label distribution for inter-area summarization of edge-device addresses

A first node generates and transmits a notification message including routing policy attributes such as network address information and a corresponding gateway identifier. The gateway identifier identifies a gateway in a physical network through which future generated data messages shall be forwarded to at least one host computer (e.g., any computer having an associated network address) as indicated by the network address information. A second node receiving the notification message utilizes the routing policy attributes to dynamically update its database identifying how to forward data packets. In this way, nodes (e.g., CE routers) of a network can be dynamically configured to support routing of messages based on the network address information and gateway identifier disseminated along with the notification message.

Methods and apparatus to distribute policy information

IPv6 traffic may be carried through an MPLS IPv4 network without the use of IPv6-over-IPv4 tunneling. An IPv6 packet is sent through the MPLS IPv4 network through a label switched path (LSP). The IPv6 packet is encapsulated with a label stack associated with the LSP. A second level label is used in the label stack (in addition to the label associated with the LSP). This second level label provides important benefits.

Two label stack for transport of network layer protocols over label switched networks

Systems and methods for link discovery and verification technique that minimize the need for line termination resources that generate and interpret packets. Of two nodes verifying a link to one another, only one node need have any line termination capability. The node lacking line termination capability simply loops back packets generated by the other node thus verifying the link. Thus, an optical cross-connect can verify links to a wide variety of node types by employing a single line termination unit capable of terminating any suitable packet type. Alternatively, a router can verify connectivity to an optical cross-connect even when the optical cross-connect lacks any line termination capability at all. This saves greatly on implementation costs for optical networks.

Link discovery and verification procedure using loopback

Systems and methods for link discovery and verification that minimize the need for line termination resources to generate and interpret packets. To verify a link between two nodes, a first of the nodes toggles light or another signal output by one of its ports while the other node attempts to detect the toggling through its own ports. A link is therefore verified between the port that is toggling and the port that detects the toggling. This link verification technique does not require packet termination capability at either node. Furthermore, this link verification technique is very simple and may be executed very quickly even at nodes having a very large number of ports to which links maybe established.

Link discovery and verification using loss of light

A router has a first interface to receive a packet from an external autonomous system and a second interface to transmit the packet as an outgoing packet to a border router. A processing engine places a first tag on the outgoing packet in accordance with a standard tag switching protocol. A shared field in the outgoing packet has at least one bit to indicate a use of the shared field, the at least one bit set by the processing engine to indicate the shared field carries a second tag, the second tag indicating a route from the border router to a destination of the packet.

Hierarchical label switching across multiple OSPF areas

In a communications-networking autonomous system consisting of an OSPF domain, autonomous-system border routers (I-ASBR and E-ASBR) cause exchange of hierarchical forwarding labels whose hierarchies are based on OSPF areas. A border router transmits into the domain an OSPF LSA Update message containing an AS-External LSA whose External Route Tag field other routers interpret as specifying a label to be used for forwarding. When that LSA is flooded into the OSPF domain, area border routers respond by flooding new LSAs created from the received one by replacing the label contained in the External Route Tag field with labels that specify their forwarding tables' locations containing information for forwarding to the originating autonomous system border router. In so doing, they enable packets destined for an extra-domain location to be forwarded through the autonomous system without requiring non-border routers to allocate labels to the exterior location or to border routers outside their areas. This enables a networking backbone to conserve its label space even though it carries a large variety of external traffic.

An interface ( , or ) in a service-provider network's transit label-switching router (P ) employs resource-management messages to inform neighbor routers of the bandwidths that it can allocate to various routes that it supports. To allocate its available bandwidth, it employs a weight value set for the route by an ingress router (PE ) in a system to which the label-switching router belongs. The transit router treats the weight as a relative bandwidth: when the sum of the bandwidths requested for various routes exceeds the bandwidth available, the router sets the bandwidths for at least some routes in accordance with the ratio of a given route's weight to the sum of the weights assigned to all routes among which it divides available bandwidth in this manner. That is, it assigns at least some bandwidth to all such routes, regardless of how small the resultant bandwidth may be. In this way, the network can operate with virtually no packet loss and yet remain available to all of its customers.

Growing community of inventors

Boxborough, MA, United States of America

Daniel C Tappan