The invention relates to a congestion processing method, apparatus, network device and storage medium. The method comprises: receiving a first data packet carrying a congestion mark through a target port; searching, in a preset greedy flow table, a target table entry that includes target flow information carried in the first data packet and an identifier of the target port, and the preset greedy flow table stores flow information of a data flow whose traffic is greater than a preset traffic threshold and an identifier of a port that receives the data flow corresponding to the flow information; processing, if the target table entry is found, the first data packet; removing, if the target table entry is not found, the congestion mark carried in the first data packet to obtain a second data packet, and processing the second data packet.

A packet transfer device includes a circuit configured to include a first queue to store a first packet classified into a high priority class and a second queue to store a second packet classified into a low priority class, a memory configured to store data configured to indicate possibilities of output for the first packet and the second packet for each time slot, a processor coupled to the memory and configured to control the output of the first packet and the second packet for each time slot according to the data stored in the memory, count a number of discards of the second packet within the second queue in a predetermined cycle, and change the data stored in the memory, when the number of discards is less than a first predetermined value, so as to reduce an output period of the second packet every the time slot.

Network device for transmitting packets having packet properties, including at least two input-output-buffers for queuing packets in the network device; a sojourn time calculator for calculating a sojourn related time for each head packet in the at least two input-output-buffers; a sojourn related time adaptor for, based on an adaptation function assigned to the corresponding input-output-buffer, adapting the sojourn related time into an adapted time for each head packet in the at least two input-output-buffers; and a scheduler for scheduling outgoing packets based on the adapted time.

Technologies are disclosed for providing a workload management service for first-in, first-out (FIFO) queues in a network-accessible message queuing service within a service provider network. When a host is overloaded, or about to become overloaded, the affected host sends an alarm to the workload management service. The workload management service determines a set of other hosts that have lower workloads based upon performance metrics. The workload management service selects a new host from the set of hosts for moving a FIFO queue from the overloaded cluster. The workload management service seals the FIFO queue fragment at the overloaded host for writing of new messages and new messages having the same message group identifier are written to a new FIFO queue fragment on the new host. Messages are not read from the new FIFO queue fragment until all messages are read from the sealed FIFO queue fragment on the overloaded cluster.

Technologies are disclosed for providing a workload management service for first-in, first-out (FIFO) queues in a network-accessible message queuing service within a service provider network. When a host is overloaded, or about to become overloaded, the affected host sends an alarm to the workload management service. The workload management service determines a set of other hosts that have lower workloads based upon performance metrics. The workload management service selects a new host from the set of hosts for moving a FIFO queue from the overloaded cluster. The workload management service seals the FIFO queue fragment at the overloaded host for writing of new messages and new messages having the same message group identifier are written to a new FIFO queue fragment on the new host. Messages are not read from the new FIFO queue fragment until all messages are read from the sealed FIFO queue fragment on the overloaded cluster.

Example implementations relate to hybrid arbitration of requests for access to a shared pool of resources. An example implementation includes receiving a set of requests for access to the shared pool of resources. The requests may each be from any number of traffic classes. A traffic class may be selected according to turn-based arbitration logic. Additionally, a request from each traffic class of a subset of received requests may be selected. A request selected by the age-based arbitration logic and of the selected traffic class may be granted access to the shared pool of resources.

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).

Network device for transmitting packets having packet properties, including at least two input-output-buffers for queuing packets in the network device; a sojourn time calculator for calculating a sojourn related time for each head packet in the at least two input-output-buffers; a sojourn related time adaptor for, based on an adaptation function assigned to the corresponding input-output-buffer, adapting the sojourn related time into an adapted time for each head packet in the at least two input-output-buffers; and a scheduler for scheduling outgoing packets based on the adapted time.

The dynamic proportioning of a maximum queue size of a data transport device queue based on throughput parameters may decrease routing latency of a data transport device. A maximum queue size parameter for a data queue may be calculated based on at least a plurality of throughput parameters during routing of data traffic from a data source device to a data recipient device. Subsequently, a maximum queue size of the data queue may be decreased according to the maximum queue size parameter to prevent enqueuing of incoming service frames into the data queue. The lack of enqueueing of the incoming service frames may cause the data source device to retransmit the one or more incoming service frames to the data transport device, instead of allowing the one or more incoming service frames to be enqueued and trapped in the data queue by additional incoming service frames.

Systems and methods are disclosed to communicate data between wireless access point with one or more low priority queues; one or more high priority queues; a quality of service (QoS) queue coupled to the low and high priority queues; a fairness counter; and a buffer coupled to the fairness counter and the QoS queue to map buffer mapping to the QoS queue when the fairness counter expires, the method including performing an association request for fairness scheduling; and injecting a low priority packet in a transmit opportunity of the high priority queue.