Abstract: A congestion control method and apparatus, a device, and a storage medium, where the congestion control method includes sending first data packets to a receive end, where a quantity of the first data packets is the first value, receiving a plurality of second data packets corresponding to all or a portion of the first data packets, where the second packets include one or more third data packets and one or more fourth data packets, and adjusting, by a transmit end, a congestion window based on the second data packets to adjust a value of the congestion window to a second value.

Abstract: A data transmission method implemented by a network device, where the data transmission method includes receiving a first data packet sent by a transmit end, buffering the first data packet to a low-priority queue when the first data packet is sent by the transmit end during a first round-trip time (RTT) of a data transmission phase between the transmit end and a receive end, receiving a second data packet from the transmit end, buffering the second data packet to a high-priority queue when the second data packet is not sent by the transmit end during the first RTT, and forwarding the second data packet in the high-priority queue before the first data packet in the low-priority queue.

Abstract: A data transmission method implemented by a network device, where the data transmission method includes receiving a first data packet sent by a transmit end, buffering the first data packet to a low-priority queue when the first data packet is sent by the transmit end during a first round-trip time (RTT) of a data transmission phase between the transmit end and a receive end, receiving a second data packet from the transmit end, buffering the second data packet to a high-priority queue when the second data packet is not sent by the transmit end during the first RTT, and forwarding the second data packet in the high-priority queue before the first data packet in the low-priority queue.

Abstract: A data transmission method implemented by a network device, where the data transmission method includes receiving a first data packet sent by a transmit end, buffering the first data packet to a low-priority queue when the first data packet is sent by the transmit end during a first round-trip time (RTT) of a data transmission phase between the transmit end and a receive end, receiving a second data packet from the transmit end, buffering the second data packet to a high-priority queue when the second data packet is not sent by the transmit end during the first RTT, and forwarding the second data packet in the high-priority queue before the first data packet in the low-priority queue.

Abstract: A communication method includes obtaining, by a source remote direct memory access (RDMA) network interface card (RNIC), to-be-transmitted data sent by a source virtual RNIC (vRNIC), obtaining, by the source RNIC, identity indication information of a destination vRNIC and packet forwarding information, and encapsulating, by the source RNIC, the to-be-transmitted data to obtain a target packet, and sending the target packet to a destination RNIC, where the destination vRNIC is a vRNIC running on the destination RNIC.

Abstract: A data transmission method implemented by a network device, where the data transmission method includes receiving a first data packet sent by a transmit end, buffering the first data packet to a low-priority queue when the first data packet is sent by the transmit end during a first round-trip time (RTT) of a data transmission phase between the transmit end and a receive end, receiving a second data packet from the transmit end, buffering the second data packet to a high-priority queue when the second data packet is not sent by the transmit end during the first RTT, and forwarding the second data packet in the high-priority queue before the first data packet in the low-priority queue.

Abstract: This application discloses a data processing method and apparatus, and a switching device. The data processing method includes: obtaining a destination address of a data packet received by an input port; determining an available output port based on the destination address; determining a busy degree of the available output port, when there is no non-busy available output port in the available output port, determining a quantity of footprint queues on the available output port, and selecting an available output port with a largest quantity of footprint queues as a target output port; determining a busy degree of a queue on the target output port, and when there is no non-busy queue on the target output port, selecting a footprint queue on the target output port as a target output queue. In the foregoing manners, a network resource is properly used, and network blocking can be effectively alleviated.

Abstract: A communication method includes obtaining, by a source remote direct memory access (RDMA) network interface card (RNIC), to-be-transmitted data sent by a source virtual RNIC (vRNIC), obtaining, by the source RNIC, identity indication information of a destination vRNIC and packet forwarding information, and encapsulating, by the source RNIC, the to-be-transmitted data to obtain a target packet, and sending the target packet to a destination RNIC, where the destination vRNIC is a vRNIC running on the destination RNIC.

Abstract: A congestion control method and apparatus, a device, and a storage medium, where the congestion control method includes sending first data packets to a receive end, where a quantity of the first data packets is the first value, receiving a plurality of second data packets corresponding to all or a portion of the first data packets, where the second packets include one or more third data packets and one or more fourth data packets, and adjusting, by a transmit end, a congestion window based on the second data packets to adjust a value of the congestion window to a second value.

Abstract: A data transmission method implemented by a network device, where the data transmission method includes receiving a first data packet sent by a transmit end, buffering the first data packet to a low-priority queue when the first data packet is sent by the transmit end during a first round-trip time (RTT) of a data transmission phase between the transmit end and a receive end, receiving a second data packet from the transmit end, buffering the second data packet to a high-priority queue when the second data packet is not sent by the transmit end during the first RTT, and forwarding the second data packet in the high-priority queue before the first data packet in the low-priority queue.

Abstract: A task allocation method, a chip are disclosed. The method includes: determining the number of threads included in a to-be-processed task; determining, in a network-on-chip formed by a multi-core processor, a continuous area formed by routers-on-chip corresponding to multiple continuous idle processor cores whose number is equal to the number of the threads; when the area is a non-rectangular area, determining an extended area extended from the non-rectangular area; and when predicted traffic of each router-on-chip that is connected to a processor core in the extended area does not exceed a preset threshold, allocating the multiple threads of the to-be-processed task to the idle processor cores in the non-rectangular area. According to the task allocation method provided in the embodiments of the present invention, problems of large hardware overheads, a low network throughput, low system utilization are avoided.

Abstract: This application discloses a data processing method and apparatus, and a switching device. The data processing method includes: obtaining a destination address of a data packet received by an input port; determining an available output port based on the destination address; determining a busy degree of the available output port, when there is no non-busy available output port in the available output port, determining a quantity of footprint queues on the available output port, and selecting an available output port with a largest quantity of footprint queues as a target output port; determining a busy degree of a queue on the target output port, and when there is no non-busy queue on the target output port, selecting a footprint queue on the target output port as a target output queue. In the foregoing manners, a network resource is properly used, and network blocking can be effectively alleviated.

Abstract: A server connection method and system, which relates to the field of communications technologies, such that servers of a same specification are used to implement an optimal network, thereby reducing complexity of routing implementation which includes, providing ten servers, where each server includes five nodes, and the five nodes of each server are connected head-to-tail in series in a same connection manner, and connecting five nodes of any server in five of the ten servers to five nodes of each of the remaining five servers in a one-to-one correspondence manner, in order to form an optimal network of a Hoffman-Singleton graph.

Abstract: A task allocation method, a chip are disclosed. The method includes: determining the number of threads included in a to-be-processed task; determining, in a network-on-chip formed by a multi-core processor, a continuous area formed by routers-on-chip corresponding to multiple continuous idle processor cores whose number is equal to the number of the threads; when the area is a non-rectangular area, determining an extended area extended from the non-rectangular area; and when predicted traffic of each router-on-chip that is connected to a processor core in the extended area does not exceed a preset threshold, allocating the multiple threads of the to-be-processed task to the idle processor cores in the non-rectangular area. According to the task allocation method provided in the embodiments of the present invention, problems of large hardware overheads, a low network throughput, low system utilization are avoided.

Abstract: A task allocation method, a chip are disclosed. The method includes: determining a number of threads included in a to-be-processed task; determining, in a network-on-chip formed by a multi-core processor, a continuous area formed by routers-on-chip corresponding to multiple continuous idle processor cores whose number is equal to the number of the threads; if the area is a non-rectangular area, determining a rectangular area extended from the area; and if predicted traffic of each router-on-chip that is connected to a non-idle processor core and in the extended rectangular area does not exceed a preset threshold, allocating the multiple threads of the to-be-processed task to the idle processor cores in the area. According to the task allocation method provided in the embodiments of the present invention, problems of large hardware overheads, a low network throughput, low system utilization are avoided.

Abstract: A method for determining an intermediate routing node, including: determining at least one intermediate routing node that is used to transfer to-be-transmitted data between the two routing nodes when there is a fault in a communication path between two routing nodes, adding a channel dependency relationship between a to-be-verified intermediate routing node and the two routing nodes to a channel dependency graph of a routing network in which this fault has not occurred in order to enable the channel dependency graph to become an updated channel dependency graph, and determining the to-be-verified intermediate routing node as a final intermediate routing node that is used to transfer the to-be-transmitted data when the updated channel dependency graph does not have a dependency relationship loop. The determined intermediate routing node can be used to transfer to-be-transmitted data, which improves resource utilization of a routing network.

Abstract: A method and an apparatus for notifying a network abnormality are provided. An OpenFlow switch detects whether an abnormality occurs in an OpenFlow network. The OpenFlow switch sends a first asynchronous message for describing the abnormality occurring in the OpenFlow network to a controller when detecting an abnormality occurring in the OpenFlow network, so that the controller processes, according to the first asynchronous message, the abnormality occurring in the OpenFlow network. In this way, the abnormality in the network is notified in time, thereby improving efficiency in processing the abnormality in the OpenFlow network. A technical problem in the prior art that an abnormality occurring in an OpenFlow network and caused by an unexpected event cannot be notified in time, so that the abnormality occurring in the OpenFlow network cannot be processed in time and efficiency in processing the abnormality occurring in the OpenFlow network is relatively low is solved.

Abstract: A method for determining an intermediate routing node, including: determining at least one intermediate routing node that is used to transfer to-be-transmitted data between the two routing nodes when there is a fault in a communication path between two routing nodes, adding a channel dependency relationship between a to-be-verified intermediate routing node and the two routing nodes to a channel dependency graph of a routing network in which this fault has not occurred in order to enable the channel dependency graph to become an updated channel dependency graph, and determining the to-be-verified intermediate routing node as a final intermediate routing node that is used to transfer the to-be-transmitted data when the updated channel dependency graph does not have a dependency relationship loop. The determined intermediate routing node can be used to transfer to-be-transmitted data, which improves resource utilization of a routing network.

Abstract: A server connection method and system, which relates to the field of communications technologies, such that servers of a same specification are used to implement an optimal network, thereby reducing complexity of routing implementation which includes, providing ten servers, where each server includes five nodes, and the five nodes of each server are connected head-to-tail in series in a same connection manner, and connecting five nodes of any server in five of the ten servers to five nodes of each of the remaining five servers in a one-to-one correspondence manner, in order to form an optimal network of a Hoffman-Singleton graph.

Abstract: A task allocation method, a chip are disclosed. The method includes: determining a number of threads included in a to-be-processed task; determining, in a network-on-chip formed by a multi-core processor, a continuous area formed by routers-on-chip corresponding to multiple continuous idle processor cores whose number is equal to the number of the threads; if the area is a non-rectangular area, determining a rectangular area extended from the area; and if predicted traffic of each router-on-chip that is connected to a non-idle processor core and in the extended rectangular area does not exceed a preset threshold, allocating the multiple threads of the to-be-processed task to the idle processor cores in the area. According to the task allocation method provided in the embodiments of the present invention, problems of large hardware overheads, a low network throughput, low system utilization are avoided.