Cisco Technology, Inc.

Ian Michael Charles Shand

Stewart Frederick Bryant

Clarence Filsfils

Stefano Benedetto Previdi

David Delano Ward

John Harper

George Swallow

Russell Ivan White

Kevin George Miles

Sami Boutros

Muthurajah Sivabalan

Alvaro E Retana

John Galen Scudder

Alexey Dimitrievich Zinin

Jean-Philippe Vasseur

Yi Yang

Anna Charny

Toerless Eckert

John H W Bettink

Jonathan C Barth

Bruce Stuart Davie

Gargi Nalawade

Olivier Bonaventure

Pierre Jean Rene Francois

Pierre Francois

Klaus Martini

Deepak Sreekantan

Mark T Schaefer

In one embodiment, micro-loops are avoided in ring topologies of packet switching devices by changing the order of propagation of link state information concerning failed communications between a particular packet switching device and a neighbor packet switching device. In one embodiment, the particular packet switching device communicates link state information of a high cost of the particular communications (e.g., in the direction from particular to neighbor packet switching devices) such that this link state information will propagate towards the particular packet switching device from at least from the furthest packet switching device in the ring topology that is currently configured to forward packets having a destination address of the neighbor packet switching device through the particular packet switching device.

Avoiding micro-loops in a ring topology of a network

Various techniques can be used to avoid loops during network convergence after a topology change such as a failure. For example, a method can involve detecting a failure that disrupts an existing forwarding path between a node and a destination node; calculating at least one updated forwarding path from the node to the destination node; identifying a maximum time for at least a portion of the network to reach a post-convergence state with respect to the failure; until the maximum time has elapsed at the node, explicitly specifying at least a portion of the updated forwarding path in packet headers of any packets being sent from the node to the destination node; and after the maximum time has elapsed at the node, non-explicitly specifying the portion of the updated forwarding path in packet headers of any additional packets being sent from the node to the destination node.

Loop avoidance during network convergence in switched networks

In an embodiment, an apparatus for forwarding data in a data communications network having as components nodes and links therebetween comprises a network interface configured to couple to a data communications network having as components nodes and links therebetween; first logic which when executed is operable to receive a repair address; second logic which when executed is operable to compute a repair path from a repairing node to the repair address around a failure component on a primary forwarding path between the apparatus and a repairing node; and third logic which when executed is operable to propagate the repair address along a label switched path (LSP), comprising the repair path, to the repairing node. The apparatus offers improved efficiency in computing a repair path in a network around a failed component because unnecessary computations are not performed with the disclosed apparatus and approach.

Forwarding data in a data communications network

In an embodiment, a method comprises determining a set of protected components that are associated with a notifying node; determining a single network repair address for the set of protected components, wherein the single network repair address is for use in response to unavailability of any of the protected components when transmitting network traffic to the notifying node; assigning the single network repair address to each of the protected components; wherein the notifying node is an internetworking device and wherein the method is performed by one or more processors.

Minimizing the number of not-via addresses

In one embodiment, a maintenance intermediate point (MIP) receives a packet traveling along a multi-protocol label switching (MPLS) label switched path (LSP) that extends from a first maintenance end point (MEP) to a second MEP. The receiving MIP decrements a time-to-live (TTL) value in a header of the packet. In response the TTL value in the header of the packet equaling a particular value, the receiving MIP examines an associated channel header (ACH) field in an operations, administration, and maintenance (OAM) message stored in a payload of the packet, and determines a particular OAM function to perform based on a code in the ACH field. The receiving MIP performs the particular OAM function.

Connection verification for MPLS label switched paths and pseudowires

In one embodiment, each node in a computer network determines a shortest looping ring back to the node through each of its neighbors. Each of these rings may then be marked in a particular direction, ensuring that any ring that shares a link with another ring is marked in such a way that the shared link is in the same direction in each of the rings that share the link. The links that are marked in the particular direction may be stored as part of a first topology. Conversely, the opposite direction on the links (e.g., bidirectional links or parallel unidirectional links) may be stored as a second topology that is link-disjoint from the first topology.

Creating balanced link-disjoint topologies in a computer network

In one embodiment, an apparatus for providing clock synchronization in a packet-based network, the network having as components nodes and links therebetween and having a network topology, is arranged to compute a forward clock synchronization packet path to a synchronization destination from the network topology according to a computation rule such that the return path for a clock synchronization packet from the synchronization destination is the same as the forward path.

Providing clock synchronization in a network

A method is described of constructing a transition route in a data communication network having as components nodes and links. Upon receipt of a transition notification identifying a first component a non-neighboring node constructs a transition route around the first component. In an embodiment, a node performs detecting the first component transition; issuing a transition notification identifying the first component and recognizable by nodes configured to construct a transition route around the first component; and upon expiry of a notification transition period, issuing a transition advertisement recognizable by all nodes on the network.

Constructing a transition route in a data communications network

In one embodiment, candidate routing backup paths are limited based on policy considerations. In one embodiment, a smaller plurality of next-hop candidates is established by policy filtering a plurality of next-hop candidates for backing up one or more routes over one or more links to a first packet switching device in a network, with the smaller plurality of next-hop candidates consisting of fewer next-hops than in the plurality of next-hop candidates. A shortest-path, loop free backup route is determined for each of the plurality of routes by considering each of the smaller plurality of next-hop candidates and not all of the plurality of next-hop candidates. This routing information corresponding to each of said determined shortest-path, loop free backup routes for use in backing up corresponding routes is installed in a routing data structure.

Limiting candidate routing backup paths based on policy considerations

In one embodiment, a packet switching device is configured to convert an Internet Protocol Version 6 (IPv6) destination address, of a received particular IPv6 packet, to a second, shorter destination address. This second destination address is then used to determine forwarding information for the received IPv6 packet, which is forwarded accordingly. In one embodiment, this second address is a 32-bit address, and in particular, an Internet Protocol Version 4 (IPv4) address. Thus, one embodiment can use the IPv4 forwarding infrastructure of a packet switching device for determining how to forward IPv6 packets. In a network according to one embodiment, packets are encapsulated in an IPv6 packet using an IPv6 destination address (that can be converted to an IPv4 address) of an egress edge packet switching device. Thus, core packet switching devices can forward IPv6 packets using IPv4 lookup operations.

Forwarding IPv6 packets based on shorter addresses derived from their IPv6 destination addresses

The embodiments described herein present methods and apparatus for dynamically assigning values in a network. A method includes receiving, at a designated value allocator from a network node, a first broadcast message using a link-state protocol. The method then identifies the first broadcast message as including a request to associate a parameter value to the network node and allocates an assigned parameter value to the network node based on the identification of the request included in the first broadcast message. A message that includes the assigned parameter value is then transmitted to the network node. Other embodiments are described.

System and method for dynamically assigning values in a network

An example method is provided and includes advertising information to network elements in a network environment. The information advertises a reachability characteristic of a first Internet protocol (IP) address associated with a first network element, and a second IP address associated with a second network element to which packets are to be sent instead of the first network element. The method also includes receiving the packets at the first network element, the packets being delivered from the second network element based on the information. In specific implementations, a preferred adjacency characteristic is identified in the information by a sub-TLV for routing the packets from the second network element to the first network element. If a failure occurs and a loop free alternate (LFA) protection mechanism is provisioned, the packets are sent via a label distribution protocol (LDP) label switched path (LSP) toward a designated interface of a network element.

System and method for providing a link-state path to a node in a network environment

In one embodiment, two neighboring nodes adjacent to each end-point node of a domain may be associated with a 'dual' (a logical structure). A first of the two neighboring nodes may be placed into a first family (and group) of the dual, while a second of the two neighboring nodes may be placed into an opposing second family (and group). Augmentations may then be performed, recursively merging and connecting the duals and/or groups, as well as connecting unplaced nodes to particular duals, families, and groups, according to forced augmentations and/or unforced augmentations where no forced augmentations exist. In the end, it may be determined whether the domain supports dual plane topologies based on whether a solution having one resultant dual with a single group in each opposing family is reached from the recursive augmentations.

Growing community of inventors

Cobham, United Kingdom

Ian Michael Charles Shand