Abstract: A network device identifies an Open Shortest Path First (OSPF) link between the network device and a layer 2 network as one of a point-to-multipoint over broadcast interface or a point-to-multipoint over non-broadcast multi access (NBMA) interface, and performs database synchronization and neighbor discovery and maintenance using one of a broadcast model or a NBMA model. The network device also generates a link-state advertisement for the network device, where the link-state advertisement includes a separate link description for each point-to-point link within the layer 2 network; and sends the link-state advertisement to each fully adjacent neighbor in the layer 2 network.

Abstract: Detecting if a label-switched path (LSP) is functioning properly. To test that packets that belong to a particular Forwarding Equivalence Class (FEC) actually end their MPLS LSP on an label switching router (LSR) that is an egress for that FEC, a request message carrying information about the FEC whose LSP is being verified may be used. The request message may be forwarded like any other packet belonging to that FEC. A basic connectivity test as well as a fault isolation test are supported. In a basic connectivity test mode, the packet should reach the end of the LSP, at which point it is sent to the control plane of the egress LSR. The LSR then verifies that it is indeed an egress for the FEC. In a fault isolation test mode, the packet is sent to the control plane of each transit LSR, which performs various checks that it is indeed a transit LSR for the LSP. The transit LSR may also return further information that helps check the control plane against the data plane, i.e.

Abstract: Virtual Private Networks (VPNs) are supported in which customers may use popular internet gateway protocol (IGPs) without the need to convert such IGPs, running on customer devices to a single protocol, such as the border gateway protocol (BGP). Scaling problems, which might otherwise occur when multiple instances of an IGP flood link state information, are avoided by using a flooding topology which is smaller than a forwarding topology. The flooding topology may be a fully connected sub-set of the forwarding topology.

Abstract: A router receives a control plane message for constructing a first LSP to a destination within a network that conforms to a first type of LSP. The control plane message includes a label for the first LSP and an identifier that identifies a first type of data traffic. The router receives a second control plane message for constructing a second LSP that conforms to the first type of LSP. The second control plane message includes a label for the second LSP and an identifier that identifies a second type of data traffic. The router installs forwarding state in accordance with policies that associate the first and second types of data traffic with different LSPs of a second type that each traverse different paths through the network, and forwards packets via the interface in accordance with the installed forwarding state.

Abstract: Virtual Private Networks (VPNs) are supported in which customers may use popular internet gateway protocol (IGPs) without the need to convert such IGPs, running on customer devices to a single protocol, such as the border gateway protocol (BGP). Scaling problems, which might otherwise occur when multiple instances of an IGP flood link state information, are avoided by using a flooding topology which is smaller than a forwarding topology. The flooding topology may be a fully connected sub-set of the forwarding topology.

Abstract: Techniques are described for performing non-revertive failover with network devices. A network device including a control unit and interface cards receives routing information protocol (RIP) updates each having a metric value. The control unit signals bidirectional forwarding detection (BFD) sessions based on the metric values of each of the RIP updates with, for example, a media gateway. The control unit also selectively installs a RIP route based on the metric values. The media gateway monitors the BFD sessions, and upon failure of an active BFD session, indicates the network device to perform non-revertive failover by sending a revised plurality of RIP updates. The network device performs non-revertive failover according to the revised plurality of RIP updates. Because of the flexibility of BFD, the network device need not revert back to a previous RIP route, therefore curtailing excessive failover.

Abstract: Methods and apparatus for allowing routers in an autonomous system to implement LDP and RSVP at the same time. RSVP can be used in the network core with LDP being used in network regions surrounding the core. LDP LSPs are tunneled through the RSVP network core using RSVP LSPs and label stacking techniques. During route selection LDP LSPs which use an RSVP LSP tunnel are preferred over alternative LDP LSPs having an equal cost associated with them to create a preference for traffic engineered routes.

Abstract: A connection between network nodes in a communication network is backed up. A failover label-switched path (LSP) is created starting at a first network node of a connection and ending 5 at the second node of the connection, while bypassing the protected connection. In the event of connection failure, data is transmitted through the failover LSP. A network operator can selectively protect different types of data by using filters that bind one or more types of traffic received over one or more interfaces to different failover LSPs.

Abstract: Outages in a label-switched path (LSP) are detected quickly using adjacency information, such as IGP “hello neighbor” queries. Protocols, such as IGP can be instructed to check certain adjacencies more frequently. For example, a node in an LSP can determine a next downstream node in the LSP as an adjacency, check that adjacency much more frequently, and in the event that the adjacency is lost (due to a down link, a down interface, or a down node), switch any LSPs that have the adjacency to a failover path or route. Since these checks need not be performed using hardware signaling, outages of shared media interfaces, such as Ethernet and Gigabit Ethernet, can be checked.

Abstract: A connection between network nodes in a communication network is backed up. A failover label-switched path (LSP) is created starting at a first network node of a connection and ending at the second node of the connection, while bypassing the protected connection. In the event of connection failure, data is transmitted through the failover LSP. A network operator can selectively protect different types of data by using filters that bind one or more types of traffic received over one or more interfaces to different failover LSPs.

Abstract: Methods and apparatus for allowing routers in an autonomous system to implement LDP and RSVP at the same time. RSVP can be used in the network core with LDP being used in network regions surrounding the core. LDP LSPs are tunneled through the RSVP network core using RSVP LSPs and label stacking techniques. During route selection LDP LSPs which use an RSVP LSP tunnel are preferred over alternative LDP LSPs having an equal cost associated with them to create a preference for traffic engineered routes.

Abstract: The label distribution protocol (LDP) is extended to set up a point to multi-point (P2MP) label switched path (LSP) across a computer network from a source network device to one or more destination network devices. LDP is extended to create a P2MP label map message containing a label and a P2MP forwarding equivalence class (FEC) element having a root node address and an identifier. The P2MP FEC element may, for example, associate an address of the root node of the P2MP LSP with an opaque identifier. The P2MP FEC element uniquely identifies the P2MP LSP. The P2MP FEC element may be advertised with a label in a P2MP label map message. A source network device or the destination network devices may initiate setup and teardown of the P2MP LSP. The P2MP label map messages may be propagated from the destination network devices to the source network device.

Abstract: Detecting if a label-switched path (LSP) is functioning properly. To test that packets that belong to a particular Forwarding Equivalence Class (FEC) actually end their MPLS LSP on an label switching router (LSR) that is an egress for that FEC, a request message carrying information about the FEC whose LSP is being verified may be used. The request message may be forwarded like any other packet belonging to that FEC. A basic connectivity test as well as a fault isolation test are supported. In a basic connectivity test mode, the packet should reach the end of the LSP, at which point it is sent to the control plane of the egress LSR. The LSR then verifies that it is indeed an egress for the FEC. In a fault isolation test mode, the packet is sent to the control plane of each transit LSR, which performs various checks that it is indeed a transit LSR for the LSP. The transit LSR may also return further information that helps check the control plane against the data plane, i.e.

Abstract: Detecting if a label-switched path (LSP) is functioning properly. To test that packets that belong to a particular Forwarding Equivalence Class (FEC) actually end their MPLS LSP on an label switching router (LSR) that is an egress for that FEC, a request message carrying information about the FEC whose LSP is being verified may be used. The request message may be forwarded like any other packet belonging to that FEC. A basic connectivity test as well as a fault isolation test are supported. In a basic connectivity test mode, the packet should reach the end of the LSP, at which point it is sent to the control plane of the egress LSR. The LSR then verifies that it is indeed an egress for the FEC. In a fault isolation test mode, the packet is sent to the control plane of each transit LSR, which performs various checks that it is indeed a transit LSR for the LSP. The transit LSR may also return further information that helps check the control plane against the data plane, i.e.

Abstract: Testing the liveliness of a data plane of a label switched path (LSP) using a two stage approach. The first stage may use a general echo request operation that may be implemented using hardware. Therefore, the first stage does not heavily burden the control plane of the LSR. If a suspect LSP passes the first stage of the diagnostic operation, nothing more needs to be done. If, however, the suspect LSP fails the first stage, the diagnostic operation proceeds to a second stage. The second stage of the diagnostic operation sends probing massages through the suspect LSP, but uses the control plane to deliver the acknowledging messages. If the suspect LSP fails the second stage of the diagnostic operation, the ingress LSR can infer that the LSP is down, and begin recovery actions. The probing messages may include padding so that MTU limits can be tested.

Abstract: Techniques are described for distribution of label switched packets, such as multiprotocol label switched (MPLS) packets, across multiple physical data paths. The techniques may, for example, be used to load balance the label switched packets across an aggregated link having two or more logically associated physical interconnects. A network device, for example, includes an interface card to receive packets associated with a common label switched path (LSP), and a control unit to distribute the packets across multiple paths. The network device may include label data that maps ranges of labels, such as MPLS labels, to types of payloads carried by the label switched packets. In accordance with the label data, the control unit extracts flow information from the label switched packets and distributes the label switched packets across the paths based on the flow information.

Abstract: Methods and apparatus for implementing bi-directional logical signal interfaces (LSIs) in communications systems which use uni-directional label switched paths (LSPs), e.g., MPLS networks, are described. To implement an LSI, two uni-directional LSPs between the same end points, e.g., routers, and extending in opposite directions, are associated together. The association of LSPs may be done by setting LSI configuration information in the routers at both ends of an LSI. Each router at the end of an LSI serves as an egress router for one of the LSPs associated with the LSI and an ingress router for the other LSP associated with the LSI. To enable an egress router to determine which, if any, LSI a packet or message corresponds to, a real as opposed to a null label is used when sending packets over an LSI LSP to an LSI LSP egress router.