Abstract:
A network element acting as a forwarding plane within a software-defined network to reduce negative effects of slow-path packet processing is described. The network element, upon receiving a first packet, determines whether it matches any flow entry within a set of flow tables. When the first packet does not match any flow entry of the set of flow tables, the network element performs a miss procedure including inserting at least a portion of the first packet into a packet miss table of the network element, which is accessed by a controller. The controller may access the packet miss table using a pull mechanism. The miss procedure may also include accessing a configurable default rule table of the network element, which includes rules enabling limited forwarding for those packets that do not match any entry of the set of flow tables.
Abstract:
A method forward Ethernet frames at a node in a network supporting an implementation of shortest path bridging (SPB) protocol is disclosed. The method starts with a shortest path computation for the node (referred to as the computing node). The shortest path computation selects at least a shortest path to each destination node in the network, where a neighboring node on the shortest path to reach each node is recorded. Then it computes a downstream loop-free alternate (LFA) node for a destination node, where the LFA node is downstream of the computing node but not on the selected shortest path to the destination node from the computing node. Then when connectivity to the neighboring node on the computed shortest path is detected to be abnormal, the node forwards an Ethernet frame with a destination media access control (MAC) address corresponding to the destination node through the LFA node.
Abstract:
According to another embodiment of the invention, TWAMP path discovery is performed to determine a sequence of IP addresses of a forward direction TWAMP E2E path to be traversed by two different TWAMP test sessions between a Sender and a Reflector. Then, additional TWAMP test request packets are transmitted for the different TWAMP test sessions; and TWAMP test reply messages are received responsive to respective ones of the TWAMP test request packets. Responsive to the TWAMP test reply messages, a PM is determined that is specific to the different sequences of IP addresses of the forward direction TWAMP E2E paths traversed by the two different TWAMP test sessions.
Abstract:
A process is implemented in a network element of a packet data to route packets with embedded virtual local area network information and an IP header to far end addresses of virtual forwarding instances. The method receiving a frame including an embedded virtual network identifier (VNI). The VNI is looked up to identify a virtual forwarding instance (VFI), the VFI having a correlated an I-Component Service Identifier (I- SID) and an internet protocol (IP) to backbone media access control (B- MAC) table that is shared with a set of VFIs each having correlated I-SIDs. The VFI resolves an IP address of the frame to a B-MAC address using the shared IP to B-MAC resolution table and encapsulates the frame with the I-SID and B-MAC; and the frame is then sent to the remote B-MAC address.
Abstract:
A method forward Ethernet frames at a node in a network supporting an implementation of shortest path bridging (SPB) protocol is disclosed. The method starts with a shortest path computation for the node (referred to as the computing node). The shortest path computation selects at least a shortest path to each destination node in the network, where a neighboring node on the shortest path to reach each node is recorded. Then it computes a downstream loop-free alternate (LFA) node for a destination node, where the LFA node is downstream of the computing node but not on the selected shortest path to the destination node from the computing node. Then when connectivity to the neighboring node on the computed shortest path is detected to be abnormal, the node forwards an Ethernet frame with a destination media access control (MAC) address corresponding to the destination node through the LFA node.
Abstract:
A three stage folded Clos network is used for Ethernet routing with improved efficiency for computational complexity, network administration, multicast addressing and load redistribution upon failure. The network includes an array of root nodes coupled to an array of edge nodes. Forwarding states are computed and installed for spanning trees rooted on the root nodes. When an edge node is identified as having a failed connection to a root node a shortest path first (SPF) tree rooted on that edge node is constructed for each Backbone VLAN identifier (B VID) for the spanning trees rooted on that root node and use the failed connection. A filtering database in each node is populated for edge node pairs having a common service identifier and unicast and multicast data are forwarded between the edge node pairs according to the filtering database via the SPF trees using a hybrid multicast addressing.
Abstract:
A node in a communication network selects between equal cost shortest paths (ECSPs) using split tiebreakers. The node advertises multiple system identifiers (IDs) for that node, and each system ID is associated with a different set of traffic IDs that distinguish different virtual networks in the network. The node receives sets of system IDs for each of the other nodes in the network. After constructing a plurality of different ECSPs between two nodes the node selects one of the ECSPs for each traffic ID used between the two nodes. For a traffic ID, the node constructs path IDs for each of the ECSPs using one system ID of each node in the path that is associated with that traffic ID. Because of the configuration of the system IDs and the traffic IDs in the network traffic affected by a failure will be distributed across a plurality of surviving ECSPs.
Abstract:
A method in an Ethernet Bridge for improved load distribution in an Ethernet network that includes the Ethernet Bridge including determining a first set of one or more shortest paths between each Ethernet Bridge pair in the Ethernet network, selecting at least a first shortest path, calculating a link utilization value for each link of the Ethernet network, determining a second set of one or more shortest paths between each Ethernet Bridge pair in the Ethernet network, generating a path utilization value for each shortest path, selecting a second shortest path on the basis of said path utilization value, whereby the selection of the second shortest in light of path utilization minimizes the standard deviation of load distribution across the entire Ethernet network.
Abstract:
Un método en un puente Ethernet para la distribución de carga mejorada en una red Ethernet que incluye el puente Ethernet, comprende: determinar un primer conjunto de una o más rutas más cortas entre cada par de puentes Ethernet de la red Ethernet; seleccionar al menos una primera ruta más corta; calcular un valor de utilización de enlaces correspondiente a cada enlace de la red Ethernet; determinar un segundo conjunto de una o más rutas más cortas entre cada par de puentes Ethernet de la red Ethernet; generar un valor de utilización de rutas para cada ruta más corta; seleccionar una segunda ruta más corta sobre la base de dicho valor de utilización de rutas, por lo cual la selección de la segunda ruta más corta, a la luz de la utilización de rutas, minimiza la desviación estándar de distribución de carga a través de toda la red Ethernet.
Abstract:
A method implemented in a node of a multi-protocol label switching (MPLS) network for improved load distribution, including determining a first set of one or more shortest paths between each MPLS node pair, selecting at least a first shortest path by applying the common algorithm tie-breaking process, calculating a link utilization value for each link of the MPLS network, determining a second set of one or more shortest paths between each MPLS node pair, generating a path utilization value for each shortest path in the second set of shortest paths based on link utilization values corresponding to each shortest path, and selecting a second shortest path from the second set of shortest paths on the basis of said path utilization value, whereby the selection of the second subsets in light of path utilization minimizes the standard deviation of load distribution across the entire MPLS network.