Abstract: This application discloses example data sending methods and apparatuses, and an example FlexE switching system. In one example, when slice packets received include a SOP flag and an EOP flag of a same data packet, immediately data packet slices is restored in the slice packets to a FlexE data stream and the FlexE data stream is sent, or the data packet slices are restored in the slice packets to the FlexE data stream and the FlexE data stream is sent when a latency is greater than or equal to a first present duration. When the slice packets received include the SOP flag but do not include the EOP flag of a data packet, the data packet slices are restored in the slice packets and the FlexE data stream is sent when the latency reaches a second preset duration. The first preset duration is less than the second preset duration.

Abstract: A data transmission method includes obtaining dequeue information that indicates a queue which requests to output data in a communications device and a target data volume that is output from each queue at a time, and the communications device manages the target data volume based on a burst value, reading, based on the queue, a sub-packet descriptor (PD) that is obtained by segmenting the first PD, the sub-PD includes target description information indicating a target data packet, the first PD includes first description information indicating a first data packet set including the target data packet, the first data packet set and the sub-PD are stored in a packet cache including a dynamic random access memory (DRAM), the first PD is stored in a control cache including a static random access memory (SRAM), and determining, the target data packet based on the sub-PD, and sending the target data packet.

Abstract: An example exchange method includes: receiving, through a first interface, M bits; after receiving the M bits, receiving M bytes through the first interface, wherein each of the M bits indicates a status of a corresponding byte in the M bytes; encapsulating, based on the M bits corresponding to the M bytes, L bytes of the M bytes to obtain L encapsulated bytes, wherein M and L are integers greater than or equal to 1, and L is less than M; exchanging the L encapsulated bytes to a second interface; decapsulating the L encapsulated bytes to obtain L decapsulated bytes; and sending the L decapsulated bytes through the second interface.

Abstract: A flexible Ethernet (FlexE) frame forwarding method, including receiving a first frame through a FlexE client input channel, obtaining a first channel identifier used to indicate the FlexE client input channel and a first subchannel identifier carried in the first frame, where the first subchannel identifier is used to indicate a logical subchannel of the FlexE client input channel, searching a preset forwarding table based on the first channel identifier and the first subchannel identifier to obtain a second channel identifier and a second subchannel identifier, where the second channel identifier is used to indicate a FlexE client output channel, and the second subchannel identifier is used to indicate a logical subchannel of the FlexE client output channel, and forwarding the first frame based on the second channel identifier and the second subchannel identifier.

Abstract: This application discloses a data sending method and apparatus, and a FlexE switching system. When slice packets received by a sending module include a SOP flag and a EOP flag of a same data packet, it indicates that the sending module receives a complete data packet, and immediately sends data packet slices, or sends the data packet slices after a latency of first preset duration. In this way, a stream inside the data packet is not interrupted, a data sending latency is reduced, and transmission efficiency is improved. When the slice packets received by the sending module include a SOP flag of a data packet but do not include a EOP flag of the data packet, it indicates that the sending module has not received a complete data packet, and then sends data packet slices after a latency of second preset duration longer than the first preset duration.

Abstract: A data transmission method includes obtaining dequeue information that indicates a queue which requests to output data in a communications device and a target data volume that is output from each queue at a time, and the communications device manages the target data volume based on a burst value, reading, based on the queue, a sub-packet descriptor (PD) that is obtained by segmenting the first PD. the sub-PD includes target description information indicating a target data packet, the first PD includes first description information indicating a first data packet set including the target data packet, the first data packet set and the sub-PD are stored in a packet cache including a dynamic random access memory (DRAM), the first PD is stored in a control cache including a static random access memory (SRAM), and determining, the target data packet based on the sub-PD, and sending the target data packet.

Abstract: The invention disclose a flexible Ethernet frame forwarding method, including: receiving a first frame through a FlexE client input channel; obtaining a first channel identifier used to indicate the FlexE client input channel and a first subchannel identifier carried in the first frame, where the first subchannel identifier is used to indicate a logical subchannel of the FlexE client input channel; searching a preset forwarding table based on the first channel identifier and the first subchannel identifier to obtain a second channel identifier and a second subchannel identifier, where the second channel identifier is used to indicate a FlexE client output channel, and the second subchannel identifier is used to indicate a logical subchannel of the FlexE client output channel; and forwarding the first frame based on the second channel identifier and the second subchannel identifier.

Abstract: Embodiments of the present invention provide an exchange method, including: receiving M bytes through a first interface; encapsulating L bytes of the M bytes to obtain L encapsulated bytes, where M and L are integers greater than or equal to 1, and L is less than M; exchanging the L encapsulated bytes to a second interface; decapsulating the L encapsulated bytes to obtain L decapsulated bytes; and sending the L decapsulated bytes through the second interface.

Abstract: The present invention relates to the field of communication technologies and discloses a method and an apparatus for encoding a data address, so that attacks can be effectively prevented and resources and costs required to handle a bank conflict are reduced. In solutions provided by embodiments of the present invention, an exclusive-OR operation is performed on one or more bits of a received uncoded address by using multiple preset transform polynomials; and an encoded address is obtained according to a result of the exclusive-OR operation. The solutions provided by the embodiments of the present invention are applicable to designs that require a large-capacity DRAM, high performance and high reliability, and have an anti-attack demand.

Abstract: A data communication method and a data communication apparatus are provided. The method mainly includes: receiving, by an adapter card, information elements sent by a first switching network board, in which a format of the information element is an information element format capable of being identified by the first switching network board; and performing, by the adapter card, format conversion on the information elements according to an information element format capable of being identified by a second switching network board, and sending the converted information elements to the second switching network board. In this manner, different types of switching network boards can be interconnected and communicate with each other, so as to implement data transmission between network devices having different structures.

Abstract: The present invention relates to the field of communication technologies, and discloses a method and an apparatus for forwarding packets to solve the problems in the prior art, namely, if a long Internet Protocol (IP) address needs to be searched for at the time of forwarding a packet, the number of Ternary Content Addressable Memories (TCAMs) need to be increased, or an external Random Access Memory (RAM) needs to be accessed for more times, which leads to a high cost and low performance.

Abstract: A switching chassis includes more than one cascade unit and more than one switching unit, where: the cascade units have cascade interfaces to connect line processing chassis; the switching units have switching ports to connect the cascade interfaces; and any cascade interface of any cascade unit is connected to one switching port of any switching unit. A router cluster with the above switching chassis includes switching chassis and line processing chassis interconnected via optical fibers, where: any optical interface of any line processing chassis is connected to one cascade interface of any cascade unit; and any cascade interface of any cascade unit is connected to one switching port of any switching unit. With the present invention, the capacity of a router cluster can be expanded without the need to replace any component of the router cluster so that the expansion cost is lower.

Abstract: The present invention relates to the field of communication technologies and discloses a method and an apparatus for encoding a data address, so that attacks can be effectively prevented and resources and costs required to handle a bank conflict are reduced. In solutions provided by embodiments of the present invention, an exclusive-OR operation is performed on one or more bits of a received uncoded address by using multiple preset transform polynomials; and an encoded address is obtained according to a result of the exclusive-OR operation. The solutions provided by the embodiments of the present invention are applicable to designs that require a large-capacity DRAM, high performance and high reliability, and have an anti-attack demand.

Abstract: In the field of switching networks, a method, a device and a system for forwarding multicast packets are disclosed, which significantly reduce the requirement on the storage capacity of the switching network, thus implementing forwarding of a large number of multicast packets. The method includes: receiving a multicast packet, and a first forwarding identifier and a second forwarding identifier of the multicast packet; obtaining a destination line card corresponding to the multicast packet according to the first forwarding identifier; and forwarding the multicast packet and the second forwarding identifier to the destination line card. The method, device and system for forwarding multicast packets are applicable to situations where it is required to forward multicast packets in a single-level switching network or a multi-level switching network.

Abstract: The present invention discloses a data management method, device and data chip. The data management method includes: receiving write data of a write request; writing the write data according to a current data management mode, where when the data management mode is a first mode, the write data of the write request is stored in an on-chip cache and when the data management mode is a second mode, the write data of the write request is stored in the on-chip cache and an off-chip memory chip; and receiving a read request of the write data, searching for the write data in the on-chip cache according to the read request, and if the write data cannot be obtained from the on-chip cache, obtaining the write data from the off-chip memory chip, thereby reducing power consumption for data access to external memory chips.

Abstract: A tree-based node insertion method and a memory device are disclosed in embodiments of the present invention. The tree-based node insertion method includes: according to the key value of a node to be inserted, searching in a primary tree for the nearest node whose key value is smaller than and the most approximate to the key value of the node to be inserted; judging whether the current secondary tree is full, and if so, selecting a node from the current secondary tree as a split node, transferring all the nodes on the right of the split node in the current secondary tree to the new secondary tree, and if not, inserting the node to be inserted into the current secondary tree.

Abstract: A method for realizing backpressure of masses of ports and a device realizing the method are disclosed. The method for realizing backpressure of masses of ports includes: detecting whether user data transmitted to a channelized physical port reaches a backpressure threshold, generating an idle frame or a series of idle frames when the backpressure threshold is reached; combining the idle frame with the user data which needs to be transmitted to the channelized physical port reaching the backpressure threshold, and transmitting the combined data; discarding the idle frame before the combined data enters the channelized physical port. The idle frame is employed in the present invention to realize the backpressure. The idle frame occupies some of the transmission bandwidth and reduces the bandwidth of user data, while the idle frame does not enter the physical port. Therefore, the aim of backpressure is achieved. The bandwidth of idle frame can be pre-configured according to the requirements.

Abstract: A method and apparatus for implementing output queue-based flow control is provided. The method includes: implementing queue scheduling and flow control by using an output port-based cell queue and by counting the number of cells from different angles. In this system, the flow control and queue management are performed separately. The queue management is directly applied to the cell scheduling. The flow control does not directly depend on the cell statistical results in the queue management. Instead, it is implemented on the basis of the cell statistical results that are obtained according to the cell priority, output port and source chip number of the cells. Therefore, the provided method and apparatus may reduce and simplify the number of queues to be scheduled and implement fine and flexible back pressure control.

Abstract: The present invention provides a backpressure method, system, and intermediate stage switching node of a multistage switching network and an intermediate stage switching node. The method includes: (i) the intermediate stage switching node receives a first backpressure information; and (ii) the intermediate stage switching node sends at least part of the first backpressure information to an upper stage switching node, wherein there is no response sent by the intermediate switching node to at least part of the first backpressure information.

Abstract: A data communication method and a data communication apparatus are provided. The method mainly includes: receiving, by an adapter card, information elements sent by a first switching network board, in which a format of the information element is an information element format capable of being identified by the first switching network board; and performing, by the adapter card, format conversion on the information elements according to an information element format capable of being identified by a second switching network board, and sending the converted information elements to the second switching network board. In this manner, different types of switching network boards can be interconnected and communicate with each other, so as to implement data transmission between network devices having different structures.