Data transmission method, cluster system, and electronic device

Abstract

Embodiments of the present application disclose a data transmission method, a cluster system, and an electronic device. The method is applied to a first optical communication node, wherein the first optical communication node is provided with a first port, the first port is connected to a second port on a second optical communication node by means of a first optical fiber link, both the first optical communication node and the second optical communication node are intermediate nodes for communication between a sending end and a destination end, and the method comprises: providing a first standby port of the first port, the destination end being reachable by the first standby port by means of a second optical fiber link; receiving first information sent by the sending end and sending the first information to the second optical communication node by means of the first optical fiber link, the first information comprising an address of the destination end; when an abnormality is detected in the first optical fiber link, sending the first information to the first standby port by means of the first port; and by means of the first standby port, forwarding the first information to the destination end via the second optical fiber link. Use of the embodiments of the present application improves the reliability of data transmission.

Description

Data transmission methods, trunking systems and electronic devices

[0001] This application claims priority to Chinese Patent Application No. 202510361566.5, filed on March 25, 2025, with the Chinese National Intellectual Property Administration, entitled “Data Transmission Method, Cluster System and Electronic Device”, the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of computers, and in particular to data transmission methods, cluster systems, and electronic devices. Background Technology

[0003] Fiber optic interconnect technology plays a crucial role in large-scale data center networks. However, fiber optic port glitching severely restricts system reliability and stability. Existing statistics show that the probability of fiber optic port glitching events is significantly higher than expected. For example, a report based on a 2015 internal Google study indicated that in a network containing approximately 15,360 servers (e.g., 12 racks, 32 servers per rack, 40 ports per server), an average of three port up / down events occurred per hour. Although most fiber optic connector glitches recover spontaneously within 5 seconds, their negative impact on high-performance computing environments cannot be ignored.

[0004] Optical port intermittent interruptions are not solely caused by optical module quality issues. The root cause lies in uncontrollable factors at the physical level of the optical fiber, such as excessive fiber bending or poor connector contact. Therefore, simply improving the quality of the optical module cannot fundamentally solve the problem of optical port intermittent interruptions. Compared to cable connections, optical links inherently have a relatively lower level of reliability.

[0005] Existing solutions for intermittent optical port outages, such as relying on transport layer protocols for data retransmission or directly replacing fiber optics with cables, have limitations, especially in their application to high-performance bus protocols. This is because, to maximize transmission efficiency, bus protocols typically do not include redundant transport layer mechanisms. Traditionally, bus protocols have been primarily used within chassis (e.g., PCI-E buses), where reliability issues are relatively minor due to cable connections. However, with the continuous expansion of data center scale, high-performance bus protocols need to be extended to cross-chassis scenarios.

[0006] Therefore, how to develop an efficient optical port intermittent interruption solution for the characteristics of optical links without adding a transport layer, and improve the reliability of data transmission, is an urgent problem to be solved. Summary of the Invention

[0007] This application provides a data transmission method, a cluster system, and an electronic device to improve data transmission reliability.

[0008] In a first aspect, embodiments of this application provide a data transmission method applied to a first optical communication node. The first optical communication node has a first port, which is connected to a second port on a second optical communication node via a first optical fiber link. Both the first and second optical communication nodes are intermediate nodes for communication between a transmitter and a destination. The method includes: setting a first backup port for the first port; the first backup port being accessible to the destination via a second optical fiber link; the first backup port being either a port on the first optical communication node logically connected to the first port within the node, or a port on a third optical communication node connected to the first port via an external line; receiving first information sent by the transmitter and sending the first information to the second optical communication node via the first optical fiber link, the first information including the address of the destination; if an anomaly is detected in the first optical fiber link, sending the first information to the first backup port via the first port; and forwarding the first information to the destination via the second optical fiber link through the first backup port.

[0009] In this embodiment, the first optical communication node and the second optical communication node serve as intermediate nodes in the data transmission network, responsible for information forwarding. The first port is equipped with a first backup port, and the first port is connected to the first backup port. When the first port is operational, the first backup port can be idle. The first backup port can reach the destination via a second optical fiber link to handle first optical fiber link failures. When the first optical communication node receives first information containing the destination address from an upstream node, it queries the routing table to determine the next hop as the second optical communication node. Under normal circumstances, the first optical communication node forwards the first information to the second optical communication node via the first optical fiber link. However, if a failure is detected in the first optical fiber link between the first and second optical communication nodes, a fast switching mechanism is employed to avoid recalculating the route and retransmitting data, which would cause transmission delays. Specifically, after detecting a link (i.e., the first optical fiber link) failure, the first optical communication node no longer sends data through the faulty link but immediately forwards the first information to the first backup port. Subsequently, the first information is forwarded to the final destination via the first backup port. In this way, fault switching can be completed quickly at the link layer without the need for intervention from upstream nodes, thereby achieving rapid fault recovery and ensuring the reliability and continuity of data transmission.

[0010] In some embodiments, sending first information to a first backup port through a first port includes: receiving first response information sent by a second optical communication node through a second backup port via a third optical fiber link, the first response information including information about the last successfully received data packet of the N data packets in the first information; and sending data packets that the second port did not successfully receive in the first information to the first backup port based on the first response information.

[0011] In this embodiment, when the first optical communication node detects an anomaly in the first optical fiber link and switches to the first backup port to send the first information, the first optical communication node continuously monitors the response information from the second port. Once the first optical communication node receives the first response information sent by the second optical communication node through the second backup port, the first optical communication node can determine the last successfully received data packet of the second port. Based on this response information, the first optical communication node can identify which data packets in the first information were lost or damaged during previous transmission and were not successfully received by the second port. Subsequently, the first optical communication node can resend only these unreceived data packets to the first backup port, avoiding the repeated transmission of the same data packets to the destination, thus eliminating the need for data packet deduplication at the destination, greatly saving processing overhead at the destination, thereby significantly reducing network bandwidth usage and improving data transmission efficiency.

[0012] In some embodiments, forwarding the first information to the destination via the first backup port through the second optical fiber link includes: forwarding data packets in the first information that were not successfully received by the second port to the destination via the first backup port through the second optical fiber link.

[0013] In this embodiment, since the data packets in the first information have been successfully sent to the second port, the second port can accurately forward them to the final destination based on the destination address information in the data packets. Therefore, the first optical communication node does not need to resend all data packets in the first information to the first backup port. It can only resend the data packets that were not successfully received to the first backup port. Then, through the first backup port and the second optical fiber link, the data packets that were not successfully received by the second port in the first information are forwarded to the destination. This avoids repeatedly sending the same data packets to the destination, thus eliminating the need for data packet deduplication at the destination, greatly saving processing overhead at the end, thereby significantly reducing network bandwidth usage and improving data transmission efficiency.

[0014] In some embodiments, the second backup port can reach the first optical communication node via a third optical fiber link; the second backup port is a port on the second optical communication node that is logically connected to the second port within the node, or a port on the fourth optical communication node that is connected to the second port via a line outside the node.

[0015] In this embodiment, the second backup port is accessible to the first optical communication node via a third optical fiber link. The second backup port and the second port can be located on the same optical communication node, connected via the node's internal switching logic (i.e., through internal logic circuitry). When the first optical fiber link fails, the second optical communication node can switch the data stream from the second port to the second backup port via its internal switching logic. Alternatively, the second backup port and the second port can be located on different optical communication nodes, with the backup port situated on a separate physical device (the fourth optical communication node). The second backup port and the second port are connected via an external inter-node line. When the first optical fiber link fails, data cannot be directly transmitted from the second port to the first port; the data stream needs to be switched from the second optical communication node to the fourth optical communication node via an external line, allowing the second optical communication node to also transmit data to the first optical communication node via the second backup port.

[0016] In some embodiments, the first optical communication node is further provided with a first storage area, and the method further includes: when receiving first information sent by the transmitting end, storing the first information in the first storage area.

[0017] In this embodiment, to ensure the reliability of data transmission, the first optical communication node can temporarily store the first information in a first storage area after receiving it from the transmitter. If the first optical communication node detects a failure in the first optical fiber link during transmission of the first information to the second optical communication node, it can recover using the first information stored in the first storage area. Specifically, the first optical communication node reads the first information from the first storage area and sends it to the destination through a backup port, thereby avoiding data loss due to link failure, shortening recovery time, and improving the reliability of data transmission.

[0018] In some embodiments, when the first backup port is a port on the first optical communication node that is logically connected to the first port within the node, sending the first information to the first backup port through the first port includes: sending the first information to the first backup port through the first port and the internal switching logic of the node.

[0019] In this embodiment, when the backup port of the first port (i.e., the first backup port) is located within the same node, and the two ports are connected through the node's internal switching logic circuit, during data transmission, if the first optical communication node detects an anomaly in the first optical fiber link between the first port and the second port, i.e., it cannot send data to the second optical communication node through the first port, the first optical communication node can send first information to the first backup port through the first port and the node's internal switching logic. Then, it can continue to send the first information originally intended to be transmitted through the first optical fiber link through the first backup port, thereby achieving fault recovery and improving the reliability of data transmission.

[0020] In some embodiments, the first optical communication node is further provided with a third port, which is connected to the first port through an internal switching logic within the node; the third optical communication node is further provided with a fourth port, which is connected to the first spare port through an internal switching logic within the node; the third port and the fourth port are connected through an external line.

[0021] In this embodiment, the first backup port and the first port can be located on different optical communication nodes, i.e., the backup port is located on another independent physical device (the third optical communication node), and the first backup port and the first port can be indirectly connected through other ports. The first optical communication node can have a third port, and the first port and the third port are located on the same node, connected through internal node switching logic, allowing data exchange between them. The third optical communication node has a first backup port and a fourth port, connected through internal node switching logic, allowing data exchange between them. The third port and the fourth port are connected through an external inter-node line, allowing data exchange between them. The first backup port and the first port are connected through the third and fourth ports; when the first port fails, the data stream needs to be switched from the first optical communication node to the third optical communication node via an external line.

[0022] In some embodiments, sending first information to a first backup port through a first port includes: sending first information to a third port through the first port, and sending first information to a fourth port through the third port and a line outside the node; the fourth port is used to receive the first information and forward the first information to the first backup port.

[0023] In this embodiment, when the first port needs to send data to a backup port located on a different optical communication node, the data is first sent from the first port to another port of that node through the node's internal switching logic, then transmitted through an external line to the corresponding port of the node where the backup port is located, and finally reaches the backup port through the internal switching logic of the node where the backup port is located. This method utilizes the switching capabilities within the nodes and the external line connections between nodes to achieve cross-node port switching and data transmission, thereby enabling fault recovery and improving the reliability of data transmission.

[0024] Secondly, embodiments of this application provide a data transmission method applied to a trunking system. The trunking system includes at least a first optical communication node and a second optical communication node. The first optical communication node is provided with a first port, which is connected to a second port on the second optical communication node via a first optical fiber link. Both the first and second optical communication nodes are intermediate nodes for communication between a transmitter and a destination. The method includes: the first optical communication node setting a first backup port for the first port; the first backup port being accessible to the destination via a second optical fiber link; the first backup port being either a port on the first optical communication node logically connected to the first port within the node, or a port on a third optical communication node connected to the first port via an external line; the first optical communication node receiving first information sent by the transmitter and sending the first information to the second optical communication node via the first optical fiber link, the first information including the address of the destination; if an anomaly is detected in the first optical fiber link, the first optical communication node sending the first information to the first backup port via the first port; and the first optical communication node forwarding the first information to the destination via the second optical fiber link through the first backup port.

[0025] In this embodiment, the first optical communication node and the second optical communication node serve as intermediate nodes in the data transmission network, responsible for information forwarding. The first port is equipped with a first backup port, and the first port is connected to the first backup port. When the first port is operational, the first backup port can be idle. The first backup port can reach the destination via a second optical fiber link to handle first optical fiber link failures. When the first optical communication node receives first information containing the destination address from an upstream node, it queries the routing table to determine the next hop as the second optical communication node. Under normal circumstances, the first optical communication node forwards the first information to the second optical communication node via the first optical fiber link. However, if a failure is detected in the first optical fiber link between the first and second optical communication nodes, a fast switching mechanism is employed to avoid recalculating the route and retransmitting data, which would cause transmission delays. Specifically, after detecting a link (i.e., the first optical fiber link) failure, the first optical communication node no longer sends data through the faulty link but immediately forwards the first information to the first backup port. Subsequently, the first information is forwarded to the final destination via the first backup port. In this way, fault switching can be completed quickly at the link layer without the need for intervention from upstream nodes, thereby achieving rapid fault recovery and ensuring the reliability and continuity of data transmission.

[0026] In some embodiments, the first information includes N data packets, where N is an integer greater than 0. The method further includes: a second optical communication node receiving M data packets from the first information sent by the first optical communication node, where M is an integer greater than 0 and less than N; and a second port being used to forward the M data packets to the destination.

[0027] In this embodiment, the first information includes N data packets, and each data packet includes the address of the destination, for example, the destination address is set in the header of each data packet. Therefore, each data packet can reach the destination based on the destination address. During the process of the first optical communication node sending the first information to the second optical communication node through the first optical fiber link, if the first optical fiber link fails after successfully sending M data packets of the first information, the NM data packets of the first information will not be successfully transmitted to the second port of the second optical communication node. Consequently, the second optical communication node does not receive all N data packets, but only receives M of them. After receiving the M data packets of the first information, the second optical communication node can forward the received M data packets to the destination based on the destination address carried on the data packets. The strategy of immediately forwarding data packets after receiving some of them can minimize latency, improve efficiency, and provide partial service availability in the event of a link failure, thus improving the user experience.

[0028] In some embodiments, when the first optical fiber link malfunctions, the method further includes: the second optical communication node sending a first response information to the second backup port through the second port, the first response information including information of the last successfully received data packet among N data packets; the second optical communication node sending the first response information to the first optical communication node through the second backup port via the third optical fiber link.

[0029] In this embodiment, the second port may also have a second backup port. The second port and the second backup port are connected, and the second backup port can reach the first optical communication node via a third optical fiber link. Data transmission is possible between the second port and the second backup port. When the first optical fiber link fails, the second port on the second optical communication node can reply with a response message to the first port on the first optical communication node through the second backup port. Specifically, when the first optical fiber link fails and the second port cannot continue to send response messages to the first port through the link, the second port generates a response message (i.e., a first response message) after detecting the link failure, and sends it to the second backup port, which then sends it to the first optical communication node via the third optical fiber link. This response message includes at least the information of the last successfully received data packet by the second port, so as to inform the first optical communication node of the data packets that the second port has successfully received before the first optical fiber link fails, avoiding the repeated sending of the same data packets to the destination. This eliminates the need for data packet deduplication at the destination, greatly saving processing overhead at the end, thereby significantly reducing network bandwidth usage and improving data transmission efficiency.

[0030] In some embodiments, the second backup port can reach the first optical communication node via a third optical fiber link; the second backup port is a port on the second optical communication node that is logically connected to the second port within the node, or a port on the fourth optical communication node that is connected to the second port via a line outside the node.

[0031] In this embodiment, the second backup port is accessible to the first optical communication node via a third optical fiber link. The second backup port and the second port can be located on the same optical communication node, connected via the node's internal switching logic (i.e., through internal logic circuitry). When the first optical fiber link fails, the second optical communication node can switch the data stream from the second port to the second backup port via its internal switching logic. Alternatively, the second backup port and the second port can be located on different optical communication nodes, with the backup port situated on a separate physical device (the fourth optical communication node). The second backup port and the second port are connected via an external inter-node line. When the first optical fiber link fails, data cannot be directly transmitted from the second port to the first port; the data stream needs to be switched from the second optical communication node to the fourth optical communication node via an external line, allowing the second optical communication node to also transmit data to the first optical communication node via the second backup port.

[0032] Thirdly, embodiments of this application provide a cluster system, which includes at least a first optical communication node and a second optical communication node. The first optical communication node is provided with a first port, which is connected to a second port on the second optical communication node via a first optical fiber link. Both the first and second optical communication nodes are intermediate nodes for communication between a transmitter and a destination. The first optical communication node is used to: set a first backup port; the first backup port is accessible to the destination via a second optical fiber link; the first backup port is either a port on the first optical communication node that is logically connected to the first port within the node, or a port on a third optical communication node that is connected to the first port via an external line; receive first information sent by the transmitter and send the first information to the second optical communication node via the first optical fiber link, the first information including the address of the destination; if an abnormality is detected in the first optical fiber link, the first optical communication node sends the first information to the first backup port via the first port; and forward the first information to the destination via the second optical fiber link through the first backup port.

[0033] Fourthly, embodiments of this application provide an electronic device, including a first optical communication node, a first port on the first optical communication node, and a second port on a second optical communication node connected via a first optical fiber link; both the first and second optical communication nodes are intermediate nodes for communication between a transmitter and a destination; the first optical communication node is used to: set a first backup port on the first optical communication node; the first backup port is reachable from the destination via a second optical fiber link; the first backup port is a port on the first optical communication node logically connected to the first port within the node, or a port on a third optical communication node connected to the first port via an external line; receive first information sent by the transmitter, and send the first information to the second optical communication node via the first optical fiber link, the first information including the address of the destination; if an abnormality is detected in the first optical fiber link, send the first information to the first backup port via the first port; and forward the first information to the destination via the second optical fiber link through the first backup port.

[0034] Fifthly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on an electronic device, cause the electronic device to perform the method of any one of the first aspects described above, or to perform the method of any one of the second aspects described above. Attached Figure Description

[0035] Figure 1 is a schematic diagram of a cluster system architecture provided in an embodiment of this application.

[0036] Figure 2 is a schematic diagram of a chip system architecture provided in an embodiment of this application.

[0037] Figure 3 is a flowchart illustrating a data transmission method provided in an embodiment of this application.

[0038] Figure 4 is a schematic diagram of a first optical communication node sending first information to a second optical communication node according to an embodiment of this application.

[0039] Figure 5 is a schematic diagram of the first storage area of ​​a first optical communication node provided in an embodiment of this application.

[0040] Figure 6 is a schematic diagram of a second optical communication node forwarding received data packets according to an embodiment of this application.

[0041] Figure 7 is a schematic diagram of sending first information through a first backup port according to an embodiment of this application.

[0042] Figure 8 is a schematic diagram of another method of sending first information from a first port to a first backup port according to an embodiment of this application.

[0043] Figure 9 is a schematic diagram of sending a first response information according to an embodiment of this application.

[0044] Figure 10 is a schematic diagram of a method for transmitting first information through a first optical fiber link and a first backup port in accordance with an embodiment of this application.

[0045] Figure 11 is a schematic diagram of a port provided in an embodiment of this application.

[0046] Figure 12 is a schematic diagram of data transmission after a failure of the data link between 1-A and 2-A, according to an embodiment of this application.

[0047] Figure 13 is a schematic diagram of the connection between a first port group and a second port group provided in an embodiment of this application.

[0048] Figure 14 is a schematic diagram of another connection between the first port group and the second port group provided in an embodiment of this application. Detailed Implementation

[0049] The embodiments of this application will now be described with reference to the accompanying drawings.

[0050] The terms "first," "second," "third," and "fourth," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0051] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0052] The cluster system architecture involved in the embodiments of this application is described below:

[0053] This application provides a cluster system architecture. A cluster system is a system that connects multiple independent computer systems together to work collaboratively, appearing as a single system or service to the outside world. Cluster systems communicate through high-performance interconnect networks and use specific software to coordinate the work between nodes, thereby achieving higher computing power, reliability, and availability. Please refer to Figure 1, which is a schematic diagram of a cluster system architecture provided in this application embodiment. This cluster system architecture may include multiple racks, each rack containing, but not limited to, one or more servers and one or more network devices.

[0054] Each server may include multiple computing chips, such as a central processing unit (CPU), a graphics processing unit (GPU), a neural processing unit (NPU), and a deep learning processing unit (DPU). In addition, servers may contain other types of specialized chips, such as FPGAs (Field-Programmable Gate Arrays) or ASICs (Application-Specific Integrated Circuits) for accelerating specific tasks. Servers are primarily responsible for processing computational tasks. They receive tasks from cluster management software and utilize their CPU, memory, and other resources for computation. Servers may also be responsible for data storage; some servers may be dedicated to data storage, which could be part of a shared storage system or a distributed file system. Servers may also be responsible for providing network services; some servers may act as web servers, providing various network services such as web services, database services, and email services.

[0055] Each network device may include one or more switching chips (SWs). Network devices are responsible for data transmission and switching; for example, a network device can be a switch, forwarding data packets within a local area network (LAN) and sending packets to the correct destination based on their MAC addresses. Switches improve network efficiency. Different types of switches (e.g., Layer 1 switches, Layer 2 switches) have different functions and performance characteristics. For instance, a Layer 1 switch handles communication between servers within the same rack, while a Layer 2 switch handles communication between racks.

[0056] In summary, the servers and network devices in the rack work together to build and maintain an efficient, secure, and reliable network environment. Their functions complement each other to accomplish important tasks such as data transmission, switching, routing, security protection, and network management.

[0057] This application provides a chip system architecture. Referring to Figure 2, which is a schematic diagram of a chip system architecture provided in this application embodiment, this chip system can achieve inter-chip communication via optical fiber to improve the performance and efficiency of the cluster system. This enables the system to achieve high-bandwidth, low-latency data transmission, significantly improving overall performance and reducing power consumption. The chip system may include one or more switching chips and one or more computing chips. The switching chips can be located in the aforementioned network device, and the computing chips can be located in the aforementioned server. This architecture is suitable for applications requiring high-bandwidth, low-latency communication, such as high-performance computing (HPC), artificial intelligence (AI), and machine learning (ML).

[0058] Switching chips, such as SW0-SW5 in Figure 2, handle communication within the chip system and with external network devices. SW0-SW3 are primary switching chips, receiving data packets from the computing chip and forwarding them based on the destination MAC address. If the destination MAC address is not in the same network segment, the data packet is forwarded to a default gateway (e.g., a secondary switching chip), which then routes it according to the IP address and routing table information to reach its destination. SW4 and SW5 are secondary switching chips, receiving data packets from the primary switching chips and forwarding them to the correct destination based on the destination IP address and routing table information. Switching chips can be connected via optical fiber, which supports higher bandwidth and lower latency, significantly improving data transmission efficiency. Deploying these fiber-connected switching chips in network devices optimizes network traffic management and significantly improves overall network performance. The design of switching chips needs to consider the type, rate, and number of optical fiber interfaces, as well as the optical signal processing capabilities. A typical switching chip has one or more device ports. For example, an SW4 can have ports A, B, C, D, and E. Different ports can be connected to other chips via optical modules and optical fibers. It should be noted that different ports on the same switching chip can communicate with each other through internal switching logic.

[0059] Computing chips, such as GPU0-GPU7, are responsible for performing computational tasks. These can be general-purpose CPUs or application-specific accelerators, such as GPUs or NPUs. However, these computing chips can also communicate with switching chips and other computing chips via fiber optic cables. This fiber optic interconnect eliminates the bandwidth bottleneck of electronic interconnects, allowing for faster data exchange between servers. Deploying these fiber-connected computing chips in servers can significantly improve their computing power and meet the demands of various high-performance computing applications. Each computing chip has one or more device ports; for example, GPU0 may have ports A, B, C, D, and E. Different ports can connect to other chips via optical modules and fiber optic cables. It should be noted that different ports on the same computing chip can communicate through internal switching logic.

[0060] It should also be noted that the chip system can adopt a variety of different topologies. Figure 2 only shows one connection relationship. The chip system can also adopt other connection methods, such as star, ring or mesh structures, to meet different application requirements.

[0061] However, this architecture faces a serious challenge: fiber optical port glitching. The probability of glitching events is significantly higher than expected. For example, a report based on a 2015 internal Google study indicated that in a large data center network, multiple port online / offline events occur on average every hour. Although most glitches recover on their own within a short time, their negative impact on high-performance computing environments cannot be ignored. The root cause of glitching lies in uncontrollable factors at the physical layer of the optical fiber, such as excessive fiber bending and poor connector contact. Simply improving the quality of the optical module cannot fundamentally solve the problem. Existing solutions, such as relying on transport layer protocols for data retransmission or directly replacing optical fibers with cables, have limitations, especially in application to high-performance bus protocols (such as the UB-C protocol under development), because these protocols typically do not include redundant transport layer mechanisms. Traditionally, high-performance bus protocols are mainly used within the chassis, where reliability issues are relatively minor due to cable connections. However, as the cluster size increases, these protocols need to be extended to cross-chassis scenarios. Therefore, to solve the above problems, this application will describe in detail how to develop an efficient optical port intermittent interruption solution for the characteristics of optical links without adding a transport layer, so as to improve data transmission reliability. This will be described in detail later and will not be repeated here.

[0062] It is understood that the chip system architecture in Figure 2 is only one of the exemplary implementations provided in the embodiments of this application, and the chip system architecture in the embodiments of this application includes, but is not limited to, the above implementation methods.

[0063] Next, with reference to the accompanying drawings, we will introduce in detail the data transmission method provided in this application, and how to develop an efficient optical port interruption solution for the characteristics of optical links without adding a transmission layer, so as to improve the reliability of data transmission.

[0064] Please refer to Figure 3, which is a schematic flowchart of a data transmission method provided in an embodiment of this application. This data transmission method can be applied to a cluster system, which includes at least a first optical communication node and a second optical communication node. The optical communication node has optical communication capabilities and can be a chip or a device; no specific limitation is made in this application. The first optical communication node and the second optical communication node can be different devices or different chips within different devices. For example, the first optical communication node can be a server, and the second optical communication node can be a switch; or, the first optical communication node can be a computing chip in a server, and the second optical communication node can be a switching chip in a switch. The first optical communication node is provided with a first port, and the second optical communication node is provided with a second port. The first port refers to a physical interface on the first optical communication node used to connect to an optical fiber link; the second port refers to a physical interface on the second optical communication node, also used to connect to an optical fiber link, and is similar in type to the first port. The first port is connected to the second port through a first optical fiber link. The first optical fiber link includes an optical fiber cable connecting the first port and the second port. The optical fiber cable uses optical signals to transmit data and has advantages such as high bandwidth, low latency, and resistance to electromagnetic interference. The connection between the first port and the second port refers to the physical connection of the first optical fiber link between the first port and the second port, thereby establishing a data transmission channel between the two optical communication nodes. The first optical fiber link is not merely a fiber optic cable; it also includes a first optical module connected to the first port. This first optical module is responsible for converting electrical signals into optical signals for transmission in the optical fiber, and for converting received optical signals back into electrical signals. A detailed description of this data transmission method is as follows:

[0065] Step S101: The first optical communication node sets the first backup port of the first port.

[0066] Specifically, the first backup port can reach the destination via the second optical fiber link; the first backup port is either a port on the first optical communication node that is logically connected to the first port within the node, or a port on the third optical communication node that is connected to the first port via an external line.

[0067] It should be noted that step S101 can be triggered before sending the first information, or step S101 can be triggered during the sending of the first information, that is, after the first optical communication node receives the first information.

[0068] Step S102: The first optical communication node receives the first information sent by the transmitter and sends the first information to the second optical communication node through the first optical fiber link.

[0069] Specifically, the first information includes the address of the destination. During data communication, the data transmission path between the sender and destination of the first information typically includes multiple intermediate nodes; the first optical communication node and the second optical communication node are two such intermediate nodes. The first optical communication node can receive the first information from the upstream node, which includes address information indicating the destination. Based on this destination address information, the first optical communication node can further query a pre-stored routing table to determine that the next hop is the second optical communication node; that is, the second optical communication node is the next hop for the first information. If the first optical fiber link is functioning normally, the first optical communication node can forward the first information to the second optical communication node via the first optical fiber link.

[0070] For example, as shown in Figure 4, which is a schematic diagram of a first optical communication node sending first information to a second optical communication node according to an embodiment of this application, the first optical communication node can be equipped with port a and port b, and the second optical communication node can also be equipped with port a and port b. Port a of the first optical communication node can receive information sent by the sending end, and port b (i.e., the first port) of the first optical communication node can be connected to port a (i.e., the second port) of the second optical communication node. Port b of the second optical communication node can send information to the destination. It should be noted that there may be other optical communication nodes between the sending end and the first optical communication node, but these are not shown in Figure 4, and there may also be other optical communication nodes between the destination and the second optical communication node, but these are not shown in Figure 4. After the first optical communication node receives the first information sent by the sending end and determines that the next-hop optical communication node is the second optical communication node, if the first optical fiber link is functioning normally, the first optical communication node can send the first information to port a of the second optical communication node through its own port b.

[0071] In some embodiments, the first optical communication node is further provided with a first storage area, and the first information received by the first optical communication node from the transmitting end is stored in the first storage area.

[0072] Specifically, the first optical communication node can be equipped with a first storage area, which can be a power-loss volatile storage area. The first storage area can be a shared storage area within the first optical communication node, meaning it can be accessed by all ports on the first optical communication node; alternatively, it can be located within the first port of the first optical communication node, for use only by that port. To ensure data transmission reliability, the first optical communication node is configured to temporarily store the first information received from the transmitting end in the first storage area. If, during the transmission of the first information to the second optical communication node via the first optical fiber link, the first optical communication node detects a failure in the first optical fiber link, it can recover using the first information stored in the first storage area. Specifically, the first optical communication node reads the first information from the first storage area and sends it to the destination through a backup port, thereby avoiding data loss due to link failure, shortening recovery time, and improving data transmission reliability.

[0073] It should be emphasized that when the first optical communication node detects a failure in the first fiber optic link, it can read the temporarily stored first information from the first storage area and then forward the first information to the first backup port. This process is a link-level retransmission, and the sending end of the first information does not need to retransmit it.

[0074] For example, as shown in Figure 5, which is a schematic diagram of the first storage area of ​​a first optical communication node provided in an embodiment of this application, after receiving the first information sent by the transmitting end, the first optical communication node can store the first information in the first storage area, and can also send the first information to the second optical communication node through the first optical fiber link.

[0075] In some embodiments, the first optical communication node caches the first information in a local storage area (i.e., the first storage area) for a preset time period. If the first optical fiber link fails within the preset time period, the node can use the cached first information to perform a recovery operation. After the preset time period expires, the first information will be cleared from the storage area. The length of the preset time period can be determined based on the end-to-end data transmission delay.

[0076] In some embodiments, the first information includes N data packets, and each data packet includes the address of the destination, where N is an integer greater than 0. The second optical communication node receives M data packets from the first information sent by the first optical communication node, where M is an integer greater than 0 and less than N. The second port is used to forward the M data packets to the destination according to the address of the destination.

[0077] Specifically, the first information includes N data packets, and each data packet includes the address of the destination, for example, the destination address is set in the header of each data packet. Therefore, each data packet can reach the destination based on the destination address. During the process of the first optical communication node sending the first information to the second optical communication node through the first optical fiber link, if the first optical fiber link fails after successfully sending M data packets of the first information, the NM data packets of the first information will not be successfully transmitted to the second port of the second optical communication node. Consequently, the second optical communication node does not receive all N data packets, but only receives M of them. As shown in Figure 6, which is a schematic diagram of a second optical communication node forwarding received data packets according to an embodiment of this application, after the second optical communication node receives M data packets from the first information, it can forward the received M data packets to the destination based on the destination address carried on the data packets. The strategy of the second optical communication node immediately forwarding data packets after receiving some can minimize latency, improve efficiency, and provide partial service availability in the event of a link failure, thus improving the user experience.

[0078] Step S103: If an abnormality is detected in the first optical fiber link, the first optical communication module sends the first information to the first backup port through the first port.

[0079] Specifically, the first backup port can be connected to the second optical module. The first backup port is used to forward the first information to the destination based on the address of the destination via the second optical module. In this application, one port can be connected to one optical module. Currently, the optical fiber link (first optical fiber link) used for data transmission between the first and second optical communication nodes has failed and cannot function properly. This may be due to physical damage, signal attenuation, equipment failure, etc. The first backup port can be understood as a pre-configured port connected to the first port, used to take over the work of the first port when the first optical fiber link fails. When the first optical communication node detects a failure in the first optical fiber link, the first optical communication node no longer sends data through the failed first optical fiber link, but instead sends the first information to the first backup port through the first port.

[0080] Step S104: The first optical communication module forwards the first information to the destination via the second optical fiber link through the first backup port.

[0081] After the first backup port receives the first information, it can continue to transmit the first information that was originally intended to be transmitted through the first optical fiber link. In this application, when the main optical fiber link fails, the first optical communication node will call the backup port of the first port and continue to send data to the destination through the backup port, thereby realizing fault recovery and improving the reliability of data transmission.

[0082] For example, as shown in Figure 7, which is a schematic diagram of sending first information through a first backup port according to an embodiment of this application, in Figure 7, it is assumed that the first backup port of the first port (port b) on the first optical communication node is port c on the first optical communication node. The first port and the first backup port can be on the same optical communication node, and the first port and the first backup port can interact with each other through internal switching logic, such as the internal bus of the node. During data transmission, when the first port is working, the first backup port can be in an idle state. The first optical communication node and the second optical communication node can be intermediate nodes in the data transmission process. If the first optical fiber link between the first optical communication node and the second optical communication node fails, the sending end does not need to resend the first information, but instead forwards the first information to the first backup port through the first port, and then forwards the first information to the destination through the second optical fiber link via the first backup port. When the first optical communication node detects a failure in the first optical fiber link, the first optical communication node no longer sends data through the failed first optical fiber link, but instead sends the first information to the first backup port through the first port, and then continues to send the first information that was originally to be transmitted through the first optical fiber link through the second optical fiber link via the first backup port, thereby realizing fault recovery and improving the reliability of data transmission.

[0083] In some embodiments, the first port of the first optical communication node and the second port of the second optical communication node can send heartbeat messages to each other to detect whether a link failure has occurred.

[0084] Specifically, a heartbeat message is a periodically sent short message used to detect the availability of a communication link. The first port of the first optical communication node and the second port of the second optical communication node can send heartbeat messages to each other; that is, the first port of the first optical communication node sends a heartbeat message to the second port of the second optical communication node, and the second port of the second optical communication node also sends a heartbeat message to the first port of the first optical communication node—this is a bidirectional process. When the link is working normally, the first and second optical communication nodes periodically send heartbeat messages to each other. If each node receives a heartbeat from the other end on time, it indicates that the link is healthy. If the link fails, the nodes will not receive heartbeat messages from the other end. After a period of time (usually several heartbeat cycles), the nodes will determine that the link has failed.

[0085] In other embodiments, the first port of the first optical communication node can detect and determine whether the first optical fiber link has failed through various methods such as optical power monitoring, link layer protocol detection, bit error rate monitoring, and alarm information.

[0086] In some embodiments, the first backup port is located at the first optical communication node, and the first backup port is connected to the first port through the node's internal switching logic; or, the first backup port is located at the third optical communication node, and the first backup port is connected to the first port through an external line between the nodes.

[0087] Specifically, the first backup port and the first port can be located on the same optical communication node, meaning they are on the same physical device. The first backup port and the first port can be connected through the node's internal switching logic, i.e., via internal logic circuits such as a switching matrix, bus, or switching chip. When the first port fails, the internal switching logic will switch the data flow from the first port to the first backup port without external line intervention. Alternatively, the first backup port and the first port can be located on different optical communication nodes, i.e., the backup port is located on a separate physical device (the third optical communication node). The first backup port and the first port are connected through an external inter-node line, i.e., via external optical fiber or other types of lines. When the first port fails, the data flow needs to be switched from the first optical communication node to the third optical communication node via this external line. The first port has a first backup port, and the first port is connected to the first backup port. When the first port is active, the first backup port can be idle to handle link failures.

[0088] In some embodiments, when the first backup port is connected to the first port through the node's internal switching logic, the first optical communication node sends first information to the first backup port through the first port, including: the first optical communication node sends first information to the first backup port through the first port and the node's internal switching logic.

[0089] Specifically, when the first port and the first backup port are located within the same node, and the two ports are connected through the node's internal switching logic circuit, as shown in Figure 7, during data transmission, if an anomaly is detected in the first optical fiber link between the first port and the second port, the first optical communication node can read the first information from the first storage area through the first port (port b on the first optical communication node in Figure 7), and send the first information to the first backup port (port c on the first optical communication node in Figure 7) through the node's internal switching logic. The first backup port then continues to transmit the first information that was originally intended to be transmitted through the first optical fiber link, thereby achieving fault recovery and improving the reliability of data transmission.

[0090] In some embodiments, the first optical communication node is further provided with a third port, which is connected to the first port through internal switching logic within the node; the third optical communication node is further provided with a fourth port, which is connected to the first spare port through internal switching logic within the node; the third port and the fourth port are connected through external lines of the node.

[0091] Specifically, the first backup port and the first port can be located on different optical communication nodes, meaning the backup port is located on another independent physical device (the third optical communication node), and the first backup port and the first port are indirectly connected through other ports. The first optical communication node can have a third port, and the first port and the third port are located on the same node, connected through internal node switching logic, allowing data exchange between them. The third optical communication node has a first backup port and a fourth port, connected through internal node switching logic, allowing data exchange between them. The third port and the fourth port are connected through an external inter-node line, allowing data exchange between them. The first backup port is connected to the first port through the third and fourth ports; when the first port fails, the data flow needs to be switched from the first optical communication node to the third optical communication node via an external line. The first port has a first backup port, connected to the first port; when the first port is active, the first backup port can be idle to handle link failures.

[0092] In some embodiments, the first optical communication node sends first information to the first backup port through the first port, including: the first optical communication node sends first information to the third port through the first port, and sends first information to the fourth port through the third port and the node's external line; the fourth port is used to receive the first information and forward the first information to the first backup port.

[0093] Specifically, when the first port needs to send data to the first backup port located on a different optical communication node, the data is first sent from the first port to another port on that node through the node's internal switching logic. Then, it is transmitted via an external line to the corresponding port on the node where the first backup port is located, and finally reaches the first backup port again through the internal switching logic of that node. This method utilizes the switching capabilities within the nodes and the external line connections between nodes to achieve cross-node port switching and data transmission, thereby enabling fault recovery and improving the reliability of data transmission.

[0094] For example, as shown in Figure 8, which is a schematic diagram of another method for sending first information from a first port to a first backup port according to an embodiment of this application, the first port is assumed to be port b on the first optical communication node, the first backup port is port b on the third optical communication node, the second port is port a on the second optical communication node, the third port is port c on the first optical communication node, and the fourth port is port a on the third optical communication node. During data transmission, if an anomaly is detected in the first optical fiber link between the first port and the second port, the first optical communication node can read the first information from the first storage area through the first port and send the first information to the third port through the node's internal switching logic. Then, the third port sends the first information to the fourth port through the node's external connection line, and subsequently, the fourth port sends the first information to the first backup port through the node's internal switching logic. Furthermore, the first information originally intended to be transmitted through the first optical fiber link is continued to be sent through the first backup port, thereby achieving fault recovery and improving the reliability of data transmission.

[0095] In some embodiments, when the first optical fiber link malfunctions, the method further includes: the second optical communication node sending a first response information to the second backup port through the second port, the first response information including information of the last successfully received data packet among N data packets; the second optical communication node sending the first response information to the first optical communication node through the second backup port via the third optical fiber link.

[0096] Specifically, the second port can also be configured with a second backup port. The second port and the second backup port are connected, and the second backup port can reach the first optical communication node via a third optical fiber link. Data transmission is possible between the second port and the second backup port. When the first optical fiber link fails, the second port on the second optical communication node can reply with a response message to the first port on the first optical communication node through the second backup port. Specifically, when the first optical fiber link fails and the second port cannot continue to send response messages to the first port through this link, the second optical communication node will generate a response message (i.e., a first response message) after detecting the link failure, and send it to the second backup port through the second port, which in turn sends it to the first optical communication node via the third optical fiber link. This response message includes at least the information of the last successfully received data packet by the second port, so as to inform the first optical communication node of the data packets that the second port has successfully received before the first optical fiber link fails, avoiding the repeated transmission of the same data packets to the destination. This eliminates the need for data packet deduplication at the destination, greatly saving processing overhead at the end, thereby significantly reducing network bandwidth usage and improving data transmission efficiency.

[0097] In some embodiments, the second backup port can reach the first optical communication node via a third optical fiber link; the second backup port is located at the second optical communication node, and the second backup port and the second port are connected through the node's internal switching logic; or, the second backup port is located at the fourth optical communication node, and the second backup port and the second port are connected through an external line between the nodes.

[0098] Specifically, the second backup port can reach the first optical communication node via a third fiber optic link. The second backup port and the second port can be located on the same optical communication node, connected via the node's internal switching logic (i.e., through internal logic circuitry). When the first fiber optic link fails, the second optical communication node can switch the data stream from the second port to the second backup port using its internal switching logic. Alternatively, the second backup port and the second port can be located on different optical communication nodes, with the backup port situated on a separate physical device (the fourth optical communication node). The second backup port and the second port are connected via an external inter-node line. When the first fiber optic link fails, data cannot be directly transmitted from the second port to the first port; the data stream needs to be switched from the second optical communication node to the fourth optical communication node via an external line, allowing the second optical communication node to also transmit data to the first optical communication node via the second backup port.

[0099] In some embodiments, please refer to FIG9, which is a schematic diagram of sending a first response information according to an embodiment of the present application. The first optical communication node sends first information to a first backup port through a first port, including: the first optical communication node receiving the first response information sent by the second port through the second backup port. The first response information includes information about the last successfully received data packet of the second port out of N data packets; the first optical communication node sends the data packets that the second port did not successfully receive in the first information to the first backup port based on the first response information.

[0100] Specifically, when the first optical communication node detects an anomaly in the first fiber optic link and switches to the first backup port to send the first information, the first optical communication node continuously monitors the response information from the second port. Once the first optical communication node receives the first response information sent by the second optical communication node through the second backup port, the first optical communication node can determine the last successfully received data packet on the second port. Based on this response information, the first optical communication node can identify which data packets in the first information were lost or corrupted during previous transmission and were not successfully received by the second port. Subsequently, the first optical communication node can resend only these unreceived data packets to the first backup port, avoiding the repeated transmission of the same data packets to the destination. This eliminates the need for data packet deduplication at the destination, greatly saving processing overhead at the destination and significantly reducing network bandwidth usage, thereby improving data transmission efficiency.

[0101] In some embodiments, forwarding the first information to the destination via the first backup port through the second optical fiber link includes: forwarding data packets in the first information that were not successfully received by the second port to the destination via the first backup port through the second optical fiber link.

[0102] Specifically, since the data packets in the first message have been successfully sent to the second port, the second port can accurately forward them to the final destination based on the destination address information in the data packets. Therefore, the first optical communication node does not need to resend all the data packets in the first message to the first backup port. It can only resend the data packets that were not successfully received to the first backup port. Then, through the first backup port and the second optical fiber link, the data packets that were not successfully received by the second port in the first message are forwarded to the destination. This avoids repeatedly sending the same data packets to the destination, thus eliminating the need for data packet deduplication at the destination side. This greatly saves the processing overhead at the end side, thereby significantly reducing the network bandwidth usage and improving data transmission efficiency.

[0103] For example, as shown in Figure 10, which is a schematic diagram of a first information transmission via a first optical fiber link and a first backup port according to an embodiment of this application, assuming that port b on the first optical communication node is the first port, port a on the second optical communication node is the second port, and port b on the third optical communication node is the first backup port, if the first optical fiber link fails during the transmission of the first information, after the first port (port b) of the first optical communication node receives the first response information sent by the second port (port a) through the second backup port, the first port can determine the last successfully received data packet of the second port, for example, the last successfully received data packet is the Mth data packet. The first port can determine which data packets in the first information were not successfully received based on the first response information, for example, all data packets after the Mth data packet were not successfully received. Then, the first port can resend the unreceived data packets to the first backup port. Since the data packets in the first information that were successfully sent to the second port can be forwarded to the destination by the second port according to the destination address in the data packet, the first port does not need to resend all data packets in the first information to the first backup port, and therefore does not need to perform data packet deduplication on the destination side, greatly saving the processing overhead on the end side.

[0104] In some embodiments, a first port group (Group1) includes a first port and a first backup port, and a second port group (Group2) includes a second port and a second backup port. The first port and the first backup port can be on the same optical communication node (which can be a chip or a device) or on different optical communication nodes; the second port and the second backup port can be on the same optical communication node or on different optical communication nodes. The first port and the first backup port are mutually backups, and data forwarding between the two ports can be achieved through one or more links or internal switching logic; the second port and the second backup port are mutually backups, and data forwarding between the two ports can be achieved through one or more links or internal switching logic.

[0105] In some embodiments, the first backup port is connected to the second port. When the first data link fails, the first optical communication node can send the first information to the first backup port through the first port, and the first backup port can then transmit the first information to the destination via the second optical fiber link. Alternatively, the first backup port can also transmit the first information to the second port, and then transmit the first information to the destination through the second port.

[0106] In some embodiments, the two sets of ports communicate with each other, and during normal communication, the backup line of the backup port has no traffic.

[0107] For example, as shown in Figure 2, the first port group can be two ports on the GPU (GPU4 ports A / B), and the second port group can be an L1 switching chip that is bridging the connection (SW2 and SW3 are bridging each other via E). Alternatively, the first port group can be two ports on the L1 switching chip (SW0 ports F / G), and the second port group can be an L2 switching chip that is bridging the connection (SW4 and SW5 are bridging each other via E).

[0108] As shown in Figure 11, which is a schematic diagram of a port provided in an embodiment of this application, 1-A can be a first port, 2-A can be a second port, 2-B can be a first backup port, 2-B can also be a second backup port, and 1-B can be a destination port. When the data link from 1-A to 2-A fails, 1-A can resend the data to 1-B through 2-B.

[0109] Specifically, please refer to Figure 12. Figure 12 is a schematic diagram of data transmission after a data link failure between 1-A and 2-A according to an embodiment of this application. 1-A (which can correspond to GPU4 in Figure 2) sends data to 1-B (which can correspond to GPU5 in Figure 2) via 2-A (which can correspond to SW2 in Figure 2). Fault state: Data is sent from the sending end (TX) of 1-Aa, and the sent data is stored as a backup in the RetryBuffer (i.e., the first storage area). If 2-A receives data and verifies an error, it attempts to retransmit via a lower lane. Further, if the lower lane retransmission fails and all lanes are found to be unavailable (flashover state), it can wait for a preset time (TimeOut) to generate the first response information (LastACKMsg). This information is marked with a pointer to the last normally forwarded packet on this port, handling the situation where the receiving side needs to deduplicate after ACK loss. The local machine packages the LastACKMsg packet (address being the optical module-to-port chip address + port number) on this port. The local backup port forwards it directly. After reaching 2-B, 2-B forwards it according to the destination address. The packet travels through 2-A->2-B->1-Ab->1-Aa, reaching the faulty port. Upon receiving the LastACKMsg, the faulty port parses it and replays the data in the RetryBuffer from the point after the last forwarding pointer. The replayed packet is forwarded by this chip to the backup port according to the default route. After reaching 1-Ab, it returns to 2-A for normal forwarding or is forwarded normally according to the destination address.

[0110] In some embodiments, the first port group includes a first port and a first spare port, and the second port group includes a second port and a second spare port. The ports in the first port group are on the same optical communication node (same chip or same device), and the second optical communication group is on different optical communication nodes (different chips or different devices). For example, in Figure 2, the first port group is an L2 switching chip (SW0 and SW1, connected via E port) that is bridging each other, and the second port group is two ports of an L3 switching chip (AB ports of SW4).

[0111] In some embodiments, the first port group includes a first port and a first spare port, and the second port group includes a second port and a second spare port. The ports in the first port group are on the same optical communication node (the same chip or the same device), and the second optical communication group is on the same optical communication node (the same chip or the same device). For example, in Figure 2, the first port group and the second port group are two chips with multi-link connections (ports A and H of SW0, and ports A and H of GPU0).

[0112] In some embodiments, the first port group includes a first port and a first spare port, and the second port group includes a second port and a second spare port. The ports in the first port group are located on different optical communication nodes (different chips or different devices), and the second optical communication group is located on different optical communication nodes (different chips or different devices). For example, in Figure 2, the first port group has directly connected GPUs (GPU0 and GPU1 are directly connected via E port), and the second port group has bridging L1 chips (SW0 and SW1 are directly connected via E port). Alternatively, the first port group has directly connected GPUs (GPU4 and GPU5 are directly connected via E port), and the second port group also has directly connected GPUs (GPU6 and GPU7 are directly connected via E port and C port).

[0113] As shown in Figure 13, which is a schematic diagram of the connection between a first port group and a second port group provided in an embodiment of this application, the first port group includes chip L2-A, chip L2-B, and L2-C interconnected by a jumper wire, and the second port group includes chips L1-Aa, L1-Ab, and L1-Ac connected by internal switching logic.

[0114] As shown in Figure 14, which is another schematic diagram of the connection between the first port group and the second port group provided in the embodiment of this application, L2-A and L2-B can be connected through L3. That is, the interconnection of the two ports in the group can be achieved through a switching chip, etc.

[0115] In summary, in this embodiment, the first optical communication node and the second optical communication node act as intermediate nodes in the data transmission network, responsible for forwarding information. The first port is equipped with a first backup port, and the first port is connected to the first backup port. When the first port is operational, the first backup port can be idle to handle link failures. When the first optical communication node receives first information containing the destination address information from an upstream node, it queries the routing table to determine the next hop as the second optical communication node. Under normal circumstances, the first optical communication node forwards the first information to the second optical communication node via the first optical fiber link. However, if a failure is detected in the first optical fiber link between the first and second optical communication nodes, to avoid the sending end recalculating the route and retransmitting data, thus causing transmission delays, this solution employs a fast switching mechanism. Specifically, after detecting a link failure through the first port, the first optical communication node no longer sends data through the faulty link but immediately forwards the first information to the first backup port. Subsequently, the first information is forwarded to the final destination via the second optical fiber link from the first backup port. In this way, fault switching can be completed quickly locally without upstream node intervention, thereby achieving rapid fault recovery and ensuring the reliability and continuity of data transmission.

[0116] Furthermore, some existing technologies employ a dual-optical-module design for certain ports, where each port connects to two optical modules: one primary and one backup. During normal operation, only the primary optical module is active, while the backup module remains in standby mode. Data transmission is only switched to the backup module when the primary link fails. This approach presents the following problems: first, switching optical modules introduces additional latency, leading to a decrease in data transmission rate; second, the complex connection relationships between ports and optical modules increase the difficulty of management and maintenance.

[0117] This application differs in that each port on the optical communication node can be connected to only one optical module, eliminating the need to configure a spare optical module for each port. When the first link fails, data transmission is carried out through the spare port of the failed port without switching optical modules. This design simplifies the connection between ports and optical modules, reduces management complexity, and, more importantly, avoids the latency caused by optical module switching, thereby improving the overall data transmission rate. Furthermore, the elimination of the need for spare optical modules also reduces hardware costs.

[0118] This application provides a cluster system, which includes at least a first optical communication node and a second optical communication node. The first optical communication node is provided with a first port, which is connected to a second port on the second optical communication node via a first optical fiber link. Both the first and second optical communication nodes are intermediate nodes for communication between a transmitter and a destination. The first optical communication node is used to: set a first backup port; the first backup port is accessible to the destination via a second optical fiber link; the first backup port is either a port on the first optical communication node that is logically connected to the first port within the node, or a port on a third optical communication node that is connected to the first port via an external line; receive first information sent by the transmitter and send the first information to the second optical communication node via the first optical fiber link, the first information including the address of the destination; if an abnormality is detected in the first optical fiber link, the first optical communication node sends the first information to the first backup port via the first port; and forward the first information to the destination via the second optical fiber link through the first backup port.

[0119] It should be noted that for detailed descriptions of the first and second optical communication nodes, please refer to the descriptions in Figures 3-14 above, which will not be repeated here.

[0120] This application provides an electronic device including a first optical communication node. The first optical communication node has a first port, which is connected to a second port on a second optical communication node via a first optical fiber link. Both the first and second optical communication nodes are intermediate nodes for communication between a transmitter and a destination. The first optical communication node is configured to: set a first backup port; the first backup port is accessible to the destination via a second optical fiber link; the first backup port is either a port on the first optical communication node logically connected to the first port within the node, or a port on a third optical communication node connected to the first port via an external line; receive first information sent by the transmitter and send the first information to the second optical communication node via the first optical fiber link, the first information including the address of the destination; if an abnormality is detected in the first optical fiber link, send the first information to the first backup port via the first port; and forward the first information to the destination via the second optical fiber link through the first backup port.

[0121] It should be noted that for detailed descriptions of the first and second optical communication nodes, please refer to the descriptions in Figures 3-14 above, which will not be repeated here.

[0122] This application provides a computer-readable storage medium including instructions that, when executed on an electronic device, cause the electronic device to perform any of the above-described data transmission methods.

[0123] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0124] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0125] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical or other forms.

[0126] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0127] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0128] If the integrated units described above are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which can be a personal computer, server, or network device, specifically a processor in the computer device) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium may include various media capable of storing program code, such as a USB flash drive, portable hard drive, magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM).

[0129] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A data transmission method, characterized in that, It is applied to a first optical communication node, which is provided with a first port, and the first port is connected to a second port on a second optical communication node through a first optical fiber link; Both the first optical communication node and the second optical communication node are intermediate nodes for communication between the transmitting end and the destination end; the method includes: A first backup port is set for the first port; the first backup port can reach the destination through a second optical fiber link; the first backup port is a port on the first optical communication node that is logically connected to the first port within the node, or a port on the third optical communication node that is connected to the first port through a line outside the node. The system receives the first information sent by the sending end and sends the first information to the second optical communication node through the first optical fiber link. The first information includes the address of the destination end. If an anomaly is detected in the first fiber optic link, the first information is sent to the first backup port through the first port; The first information is forwarded to the destination via the second optical fiber link through the first backup port.

2. The method according to claim 1, characterized in that, Sending the first information to the first backup port through the first port includes: Receive the first response information sent by the second optical communication node through the second backup port via the third optical fiber link. The first response information includes information about the last successfully received data packet from the N data packets of the first information on the second port. Based on the first response information, send the data packets that the second port failed to receive in the first information to the first backup port.

3. The method according to claim 2, characterized in that, The step of forwarding the first information to the destination via the second optical fiber link through the first backup port includes: The data packets that were not successfully received by the second port in the first information are forwarded to the destination via the second optical fiber link through the first backup port.

4. The method according to claim 2 or 3, characterized in that, The second backup port can reach the first optical communication node through the third optical fiber link; the second backup port is a port on the second optical communication node that is logically connected to the second port within the node, or a port on the fourth optical communication node that is connected to the second port through a line outside the node.

5. The method according to any one of claims 1-4, characterized in that, The first optical communication node is further provided with a first storage area, and the method further includes: When the first information is received from the sending end, the first information is stored in the first storage area.

6. The method according to any one of claims 1-5, characterized in that, When the first backup port is a port on the first optical communication node that is logically connected to the first port within the node, sending the first information to the first backup port through the first port includes: The first information is sent to the first backup port through the internal switching logic of the first port and the node.

7. The method according to any one of claims 1-5, characterized in that, The first optical communication node is also provided with a third port, which is connected to the first port through internal switching logic within the node; The third optical communication node is also provided with a fourth port, and the first spare port and the fourth port are connected through the node's internal switching logic; The third port and the fourth port are connected by an external line.

8. The method according to claim 7, characterized in that, Sending the first information to the first backup port through the first port includes: The first information is sent from the first port to the third port, and from the third port to the fourth port via the line outside the node; the fourth port is used to receive the first information and forward it to the first backup port.

9. A data transmission method, characterized in that, The system is applied to a cluster system, which includes at least a first optical communication node and a second optical communication node. The first optical communication node is provided with a first port, and the first port is connected to a second port on the second optical communication node through a first optical fiber link. Both the first optical communication node and the second optical communication node are intermediate nodes for communication between the transmitting end and the destination end; the method includes: The first optical communication node is configured with a first backup port for the first port; the first backup port can reach the destination via a second optical fiber link; the first backup port is either a port on the first optical communication node that is logically connected to the first port within the node, or a port on the third optical communication node that is connected to the first port via a line outside the node. The first optical communication node receives the first information sent by the sending end, and sends the first information to the second optical communication node through the first optical fiber link. The first information includes the address of the destination end. If an anomaly is detected in the first optical fiber link, the first optical communication node sends the first information to the first backup port through the first port; The first optical communication node forwards the first information to the destination via the second optical fiber link through the first backup port.

10. The method according to claim 9, characterized in that, The first information includes N data packets, where N is an integer greater than 0. The method further includes: The second optical communication node receives M data packets from the first information sent by the first optical communication node, where M is an integer greater than 0 and less than N; the second port is used to forward the M data packets to the destination.

11. The method according to claim 10, characterized in that, When the first optical fiber link malfunctions, the method further includes: The second optical communication node sends a first response message to the second backup port through the second port. The first response message includes information about the last data packet successfully received by the second port out of the N data packets. The second optical communication node sends the first response information to the first optical communication node via the second backup port through the third optical fiber link.

12. The method according to claim 11, characterized in that, The second backup port can reach the first optical communication node through the third optical fiber link; the second backup port is a port on the second optical communication node that is logically connected to the second port within the node, or a port on the fourth optical communication node that is connected to the second port through a line outside the node.

13. A cluster system, characterized in that, The cluster system includes at least a first optical communication node and a second optical communication node. The first optical communication node is equipped with a first port, which is connected to a second port on the second optical communication node via a first optical fiber link. Both the first and second optical communication nodes are intermediate nodes for communication between the transmitting end and the destination end. The first optical communication node is used for: A first backup port is configured for the first optical communication node; the first backup port is accessible to the destination via a second optical fiber link; the first backup port is either a port on the first optical communication node that is logically connected to the first port within the node, or a port on a third optical communication node that is connected to the first port via an external line. The system receives the first information sent by the sending end and sends the first information to the second optical communication node through the first optical fiber link. The first information includes the address of the destination end. If an anomaly is detected in the first optical fiber link, the first optical communication node sends the first information to the first backup port through the first port; The first information is forwarded to the destination via the second optical fiber link through the first backup port.

14. An electronic device, characterized in that, The system includes a first optical communication node, which has a first port connected to a second port on a second optical communication node via a first optical fiber link. Both the first and second optical communication nodes serve as intermediate nodes for communication between a transmitter and a destination. The first optical communication node is used for: A first backup port is configured for the first optical communication node; the first backup port is accessible to the destination via a second optical fiber link; the first backup port is either a port on the first optical communication node that is logically connected to the first port within the node, or a port on a third optical communication node that is connected to the first port via an external line. The system receives the first information sent by the sending end and sends the first information to the second optical communication node through the first optical fiber link. The first information includes the address of the destination end. If an anomaly is detected in the first fiber optic link, the first information is sent to the first backup port through the first port; The first information is forwarded to the destination via the second optical fiber link through the first backup port.

15. A computer-readable storage medium comprising instructions, characterized in that, When the instructions are executed on an electronic device, the electronic device performs the method as described in any one of claims 1-8, or the method as described in any one of claims 9-12.

Images