Load balancing and overload control techniques are disclosed for use in a SIP-based network or other type of network comprising a plurality of servers. In a load balancing technique, a first server receives feedback information from at least first and second downstream servers associated with respective first and second paths between the first server and a target server, the feedback information comprising congestion measures for the respective downstream servers. The first server dynamically adjusts a message routing process based on the received feedback information to compensate for imbalance among the congestion measures of the downstream servers. In an overload control technique, the first server utilizes feedback information received from at least one downstream server to generate a blocking message for delivery to a user agent.

In accordance with embodiments disclosed herein, there are provided methods, systems, mechanisms, techniques, and apparatuses for traffic aggregation on multiple WAN backhauls and multiple distinct LAN networks; for traffic load balancing on multiple WAN backhauls and multiple distinct LAN networks; and for performing self-healing operations utilizing multiple WAN backhauls serving multiple distinct LAN networks. For example, in one embodiment, a first Local Area Network (LAN) access device is to establish a first LAN; a second LAN access device is to establish a second LAN; a first Wide Area Network (WAN) backhaul connection is to provide the first LAN access device with WAN connectivity; a second WAN backhaul connection to provide the second LAN access device with WAN connectivity; a management device is communicatively interfaced with each of the first LAN access device, the second LAN access device, the first WAN backhaul connection, and the second WAN backhaul connection; and the management device routes a first portion of traffic originating from the first LAN over the first WAN backhaul connection and routes a second portion of the traffic originating from the first LAN over the second WAN backhaul connection.

Techniques for identity and policy based routing are presented. A resource is initiated on a device with a resource identity and role assignments along with policies are obtained for the resource. A customized network is created for the resource using a device address for the device, the resource identity, the role assignments, and the policies.

The present disclosure provides for methods, network devices, and computer readable storage media for packet reordering. In one embodiment, a method includes receiving a first packet of a first flow at a network device and determining whether flow-identifying information extracted from the first packet matches an existing flow entry. The method also includes, in response to a determination that the flow-identifying information does not match any existing flow entries, generating a new transient flow entry that includes the flow-identifying information and packet-in state. The method also includes forwarding the first packet to a controller via a packet-in stream.

A link protection method in a software-defined networking (SDN), a corresponding switching device and network controller, where the method includes receiving, by a first switching device in the SDN, first information from a network controller in the SDN and link protection information, where the first information establishes a communication link between the first switching device and a destination device, and the link protection information instructs the first switching device to proactively perform primary-to-secondary link switching when a link is faulty, establishing, by the first switching device, the communication link with the destination device according to the first information, and determining, by the first switching device according to the link protection information, that the first switching device proactively performs the primary-to-secondary link switching.

A multi-domain network (5) has a Hierarchical Path Computation Element, H-PCE, architecture comprising a Parent Path Computation Element, P-PCE (31) and a Child Path Computation Element, C-PCE, (21-25) at each of the domains (11-15). A C-PCE (21-25) determines summarised domain topology information about the domain (11-15). The summarized domain topology information comprises at least one of: an indication of whether the domain (11-15) is capable of supporting Point-to-Multi-Point services; an indication of whether a node of the summarised domain topology is capable of supporting a branch point for Point-to-Multi-Point services. The summarised domain topology information is sent to a P-PCE (31) via a communication interface (35). The P-PCE (31) computes an end-to-end Point-to-Multi-Point tree using a topology based on summarised domain topology information (32) received from the Child Path Computation Elements. The P-PCE (31) requests each C-PCE (21-25) of a domain in the computed end-to-end Point-to-Multi-Point tree to compute an intra-domain path.

In a distributed control system including a central communication device, terminal communication devices to which target devices to be controlled are connected, and a network including multiple communication paths connecting the central communication device and terminal communication devices, each terminal communication device includes a calculation input/output performance storage unit storing input/output performance of a calculation unit for controlling the target devices to be controlled and a control input/output performance storage unit storing the performance of the input/output control units of the target devices to be controlled, and the central communication device collects input/output performance information previously stored in these storage units, determines communication paths of the terminal communication units and a packet division method in such a manner that an amount of a communication data of a network and response performance requested of the distributed control system are satisfied, on the basis of the collected input/output performance information and sets the determination results in the terminal communication devices.

A method for dynamic routing is provided. Status information of a multichannel video and data distribution service (MVDDS) channel from customer premises equipment (CPE) is received. The status information is evaluated to determine if data destined for the CPE over the MVDDS channel should be routed over a secondary channel. Data destined for the CPE is route over the secondary channel when the data is determined to be routed over the secondary channel.

The embodiments disclose a method and apparatus of selecting a path for transmission from paths between a first LER and a second LER in the MPLS network. The method may comprise: obtaining state information of the paths between the first LER and the second LER, the state information may indicate that a path is in up state or down state; obtaining stability information of at least one of the paths between the first LER and the second LER, the stability information may indicate that a path is stable or unstable; and selecting a path for transmission based on the state information and the stability information.

A method begins by a first computing device determining a routing plan to route a set of encoded data slices from the first computing device to a second computing device via a plurality of network paths of a communications network. The method continues with the second computing device receiving encoded data slices via one or more network paths. When the second computing device receives a decode threshold number of encoded data slices, the method continues with the second computing device sending a message to the communications network indicating receipt of the decode threshold number of encoded data slices. The method continues with a relay unit determining whether the relay unit is in possession of a not-yet delivered encoded data slice. When the relay unit is in possession of the not-yet delivered encoded data slice, the method continues with the relay unit ceasing forwarding of the not-yet delivered encoded data slice.