In an embodiment, an out-of-order, reliable, end-to-end protocol is provided that can enable direct user-level data placement and atomic operations between nodes of a multi-node network. The protocol may be optimized for low-loss environments such as High Performance Computing (HPC) applications, and may enable loss detection and de-duplication of packets through the use of a robust window state manager at a target node. A multi-node network implementing the protocol may have increased system reliability, packet throughput, and increased tolerance for adaptively routed traffic, while still allowing atomic operations to be idempotently applied directly to a user memory location.

An approach, executed by a computer, for adjusting polling parameters. The approach includes one or more processors polling an endpoint application using an initial polling quantity and an initial polling frequency. The approach includes determining a first number of events not consumed in a queue of a listening application and a second number of events in a queue at the endpoint application at an end of a polling cycle. Furthermore, the approach includes adjusting at least one of the initial polling quantity and the initial polling frequency based, at least in part, on a comparison of the first number of events not consumed in the queue of the listening application and the second number of events in the queue at the endpoint application at the end of the polling cycle.

Embodiments provide a method for transmitting a service stream in a flexible Ethernet and an apparatus. The method includes: obtaining a to-be-transmitted service stream, where the service stream is to be transmitted by using a target virtual connection supported by a physical connection group between a transmit end and a receive end, the physical connection group includes multiple physical connections and supports at least one virtual connection, and the target virtual connection is any one of the at least one virtual connection; determining, from total bandwidth resources of the multiple physical connections and according to timeslot configuration tables used by the multiple physical connections, a timeslot bandwidth resource that belongs to the target virtual connection; and transmitting the service stream to the receive end by using the timeslot bandwidth resource that belongs to the target virtual connection.

A flow management method comprises receiving (210) of incoming downlink packets. The received downlink packets are classified (220) in sub-flows, based on information that is available in a header of respective such received downlink packet. The downlink packets of each sub-flow are queued (230) in a respective sub-flow queue. Downlink packets are extracted (250) from the sub-flow queues into a common outgoing flow. A sub-flow queue from which no downlink packets yet have been extracted is prioritized (240). The extraction comprises assigning of an indicator of last service occasion to each sub-flow queue when a predetermined amount of data has been extracted from said sub-flow queue. When no prioritized sub-flow queues are present, the sub-flow queue with the earliest last service is selected. When a prioritized sub-flow queue is present, the prioritized subflow queue is selected. The downlink packets of the common outgoing flow are sent (260).

A network node comprises an optical input, an optical output, a random-access queue and processing system. It receives a data packet, at the optical input and determines whether to process it as a guaranteed-service packet or as a statistically-multiplexed packet. A guaranteed-service packet is output within a predetermined maximum time of receipt, optionally within a data container comprising container control information. A statistically-multiplexed packet is queued. The node determines a set of statistically-multiplexed packets that would fit a gap between two guaranteed-service packets; selects one of the packets; and outputs it between the two guaranteed-service packets.

Technologies for dynamically transitioning network traffic host buffers of a network computing device include the software abstraction of one or more hardware queues of the network computing device based on a network flow type associated with network traffic received by the network computing device. The network computing device is configured to identify a queue transition event, completing pending transactions in one or more of the software abstracted queues, and transition the abstracted queues to handle the flow type associated with the queue transition event. Additionally, the network computing device is configured to realign the abstracted queues to be associated with one or more hardware components of the network computing device based on the second network traffic flow type, provide a ready indication to a client associated with the abstracted queues that indicates the abstracted queues are ready for polling, and process received network traffic associated with the second network traffic flow type in the abstracted queues. Other embodiments are described herein.

In an apparatus for receiving and forwarding data packets on a network, a network device includes a plurality of ports for coupling to the network and for transmitting packets to devices disposed in or coupled to the network. At least one processor configured to process packets received via the network processes packets by selectively forwarding processed packets to one or more of the ports. A plurality of queues are defined in a memory, each configured to store packets to be transmitted by ports in the plurality of ports. A queue manager is configured to selectively assign a subset of the plurality of queues to a subset of the plurality of ports.

The present invention relates to a wireless communication terminal and a wireless communication method for efficiently scheduling uplink multi-user transmission. To this end, provided are a base wireless communication terminal, including: a transceiver configured to transmit and receive a wireless signal; and a processor configured to control an operation of the transceiver, wherein the processor selects an access category for transmitting a trigger frame which solicits an uplink multi-user transmission, performs a backoff procedure for transmitting the trigger frame based on the selected access category, and transmits the trigger frame when a backoff counter of the backoff procedure expires and a wireless communication method using the same.

An approach, executed by a computer, for adjusting polling parameters. The approach includes one or more processors polling an endpoint application using an initial polling quantity and an initial polling frequency. The approach includes determining a first number of events not consumed in a queue of a listening application and a second number of events in a queue at the endpoint application at an end of a polling cycle. Furthermore, the approach includes adjusting at least one of the initial polling quantity and the initial polling frequency based, at least in part, on a comparison of the first number of events not consumed in the queue of the listening application and the second number of events in the queue at the endpoint application at the end of the polling cycle.

In an embodiment, an out-of-order, reliable, end-to-end protocol is provided that can enable direct user-level data placement and atomic operations between nodes of a multi-node network. The protocol may be optimized for low-loss environments such as High Performance Computing (HPC) applications, and may enable loss detection and de-duplication of packets through the use of a robust window state manager at a target node. A multi-node network implementing the protocol may have increased system reliability, packet throughput, and increased tolerance for adaptively routed traffic, while still allowing atomic operations to be idempotently applied directly to a user memory location.