Method and device for keeping butt joint of optical module channels

Abstract

The invention provides a method and a device for keeping butt joint of optical module channels. The method comprises the following steps: setting a mapping port of each local optical module channel to be in an MAC layer forced working state; sending a local speed reduction capability negotiation message; receiving an opposite terminal speed reduction capability negotiation message from an opposite terminal; when it is detected that a receiving end signal of a mapping port of any local optical module channel is abnormal, generating a fault channel notification message carrying a local fault optical module channel identifier, and sending the fault channel notification message through any local optical module channel; and according to the channel combination corresponding to each local support speed reduction rate and the channel combination corresponding to each opposite end support speed reduction rate, selecting an optical module channel combination which is commonly supported and has the maximum support speed reduction rate.

Description

Technical Field

[0001] The present application relates to communication technology, and more specifically, to a method and apparatus for maintaining interconnection between optical module channels. Background Art

[0002] In the current AI computing network scenarios represented by artificial intelligence and big data, the interconnection architecture between servers and network switches is usually designed with high bandwidth and low latency as the main goals. Especially in large-scale AI training clusters, servers are connected to network leaf switches through high-speed Ethernet ports (such as 100G, 200G, 400G, etc.), forming the so-called "leaf-spine" network structure. In actual deployment, due to various considerations such as cost, power consumption and space, most servers adopt a single-port access (Single-Homed) method, that is, the network adapter card (NIC, commonly known as network card) of each server is only connected to one port of a leaf switch. Although this single access design simplifies the network structure and improves resource utilization, it also brings the risk of single point of failure (SPOF).

[0003] In this single-access model, if a server network card or its connected optical module fails, the server will be unable to access the AI ​​training cluster resources over the network, forcing the AI ​​training tasks running on that server to be interrupted. Compared to ordinary services, AI training tasks typically have long cycles, large data volumes, and high resource consumption. Once training is interrupted, not only will the initial investment in computing power and time be lost, but it may also lead to uncertainty in training results, reduced model accuracy, and even the need to restart the training process from scratch, resulting in significant business and economic costs.

[0004] When a device using optical module lane keep-alive detects a single lane failure, it doesn't shut down the port directly. Instead, it dynamically adjusts the port's operating speed, allowing the remaining functioning lanes to continue transmitting data at a reduced speed. However, optical module lane keep-alive requires support from both the switch and server on both sides. However, the server and switch come from different manufacturers, use different optical modules, and have different software implementation mechanisms. This lacks a negotiation mechanism that allows the server and switch to implement speed reduction using the same optical module lane keep-alive mechanism. Summary of the Invention

[0005] The purpose of this application is to provide a method and apparatus for maintaining the connection of optical module channels, so as to enable communication equipment to reduce the speed using the same optical module channels.

[0006] To achieve the above-mentioned purpose, the present application provides a method for maintaining the docking of optical module channels, the method comprising setting the mapping port of each local optical module channel to a forced working state of the MAC layer; sending a first speed reduction capability negotiation message; wherein the first speed reduction capability negotiation message carries local speed reduction capability negotiation information including: local supported rate, number of local channels, local single-channel rate, multiple locally supported speed reduction rates, and channel combination corresponding to each locally supported speed reduction rate; receiving a second speed reduction capability negotiation message from the opposite end; wherein the first speed reduction capability negotiation message carries the speed reduction capability negotiation information of the opposite end including: opposite end supported rate, number of opposite end channels, opposite end single-channel rate, multiple supported speed reduction rates of the opposite end, and channel combination corresponding to each opposite end supported speed reduction rate; when an abnormal signal is detected at the receiving end of the mapping port of any local optical module channel, a first fault channel notification message is generated and sent through any local optical module channel; wherein the first fault channel notification message carries the local fault optical module channel identifier; according to the channel combination corresponding to each locally supported speed reduction rate and the channel combination corresponding to each opposite end supported speed reduction rate, an optical module channel with the maximum supported speed reduction rate is selected.

[0007] To achieve the above-mentioned purpose, the present application also provides a device for maintaining the connection of optical module channels, which can be applied to a switching and forwarding device of a network device. The device includes a processor and a storage medium storing machine-executable instructions; the processor performs the following operations by running the machine-executable instructions: setting the mapping port of each local optical module channel to a MAC layer forced working state; sending a first speed reduction capability negotiation message; wherein the first speed reduction capability negotiation message carries local speed reduction capability negotiation information including: local supported rate, local channel number, local single-channel rate, multiple locally supported speed reduction rates, and a channel combination corresponding to each locally supported speed reduction rate; receiving a second speed reduction capability negotiation message from the opposite end; wherein the first speed reduction capability negotiation message carries the opposite end's speed reduction capability negotiation information including: opposite end supported rate, opposite end channel number, opposite end single-channel rate, multiple supported speed reduction rates of the opposite end, and a channel combination corresponding to each opposite end supported speed reduction rate; when detecting that the receiving end signal of the mapping port of any local optical module channel is abnormal, generating a first fault channel notification message and sending it through any local optical module channel; wherein the first fault channel notification message carries a local fault optical module channel identifier; according to each local supported speed reduction rate corresponding channel combination and each opposite end supported speed reduction rate corresponding channel combination, selecting an optical module channel with the maximum supported speed reduction rate that is commonly supported.

[0008] The beneficial effect of the present application is that a device on either side of a network device can detect a failure in the transmitting end of the optical module channel of the other end, notify the other end to support speed reduction according to the negotiated speed reduction capability negotiation information, and use the same optical module channel to support speed reduction. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 A flowchart of an embodiment of a method for achieving optical module channel docking provided by this application;

[0010] Figure 2 A schematic diagram of a single connection of network devices in the artificial intelligence training parameter network;

[0011] Figure 3 A schematic diagram of an embodiment of the present application for achieving optical module channel connection maintenance in a faulty channel;

[0012] Figure 4 A schematic diagram of an embodiment of the present application providing a method for restoring a faulty channel to maintain interconnection of optical module channels;

[0013] Figure 5 This is a schematic diagram of an embodiment of a device for achieving optical module channel docking provided by the present application. DETAILED DESCRIPTION

[0014] The present invention will be described in detail with reference to a plurality of examples shown in the accompanying drawings. In the following detailed description, a number of specific details are provided to provide a comprehensive understanding of the present invention. Known methods, steps, components, and circuits are not described in detail in the examples to avoid obscuring the understanding of the examples.

[0015] Among the terms used, the term "including" means including but not limited to; the term "containing" means including but not limited to; the terms "above," "within," and "below" are inclusive; the terms "greater than" and "less than" are exclusive. The term "based on" means based on at least a portion.

[0016] Figure 1 FIG2 is a flow chart of an embodiment of a method for achieving optical module channel docking provided by the present application, which includes:

[0017] Step 101, setting the mapping port of each local optical module channel to a MAC layer forced working state;

[0018] Step 101: Send a first deceleration capability negotiation message; wherein the first deceleration capability negotiation message carries local deceleration capability negotiation information including: local supported rate, number of local channels, local single channel rate, multiple locally supported deceleration rates, and channel combinations corresponding to each locally supported deceleration rate;

[0019] Step 102: Receive a second speed reduction capability negotiation message from the peer end; wherein the first speed reduction capability negotiation message carries the peer end's speed reduction capability negotiation information including: the peer end's supported rate, the number of peer end channels, the peer end's single channel rate, the peer end's multiple supported speed reduction rates, and the channel combination corresponding to each peer end's supported speed reduction rate;

[0020] Step 103: When a receiving end signal of a mapped port of any local optical module channel is detected to be abnormal, a first fault channel notification message is generated and sent through any local optical module channel; wherein the first fault channel notification message carries a local fault optical module channel identifier;

[0021] Step 104 : Select a channel combination with a maximum supported downspeeding rate based on each channel combination corresponding to the local supported downspeeding rate and each channel combination corresponding to the peer supported downspeeding rate.

[0022] Figure 1 The beneficial effect of the embodiment is that a device on either side of the network device can detect a failure at the transmitting end of the optical module channel at the opposite end, and notify the opposite end to support speed reduction using the same optical module channel according to the negotiated speed reduction capability negotiation information.

[0023] Figure 2 Schematic diagram of a single connection of network devices in artificial intelligence training parameter networking.

[0024] Figure 2 In the figure, one side of device 21 and device 22 is a switch device and the other side is a server. In the artificial intelligence training parameter networking, the optical module between the switch device and the server is connected to realize a single connection.

[0025] Devices 21 and 22 establish a one-to-one mapping relationship between the transmitting end (Tx) and receiving end (RX) of each port of the MAC (or PHY physical layer) chip and a physical transceiver channel (lane) of the optical module.

[0026] In the device 21, the mapping relationship between each port of the MAC module 211b and each channel of the optical module 212 is as follows:

[0027] Port 1 of MAC module 211b ( Figure 2 The transmitting end Tx1 and the receiving end Rx1 of the optical module 212 are mapped to the channel lane1 of the optical module 212; wherein the transmitting direction is: serializer Ser2→laser (LD), and the receiving direction is: photodetector (PD)→deserializer DES1.

[0028] Port2 of MAC module 211b ( Figure 2 The transmitting end Tx2 and the receiving end Rx1 of the optical module 212 are mapped to the channel lane 2; wherein the transmitting direction is: serializer Ser2→laser (LD), and the receiving direction is: photodetector (PD)→deserializer DES2.

[0029] MAC module 211b switch chip port 3 ( Figure 2The mapping between the transmitting end Tx3 and the receiving end Rx3 and the channel lane3 of the optical module 212 is not shown; wherein, the transmitting direction is: serializer Ser3→laser (LD), and the receiving direction is: photodetector (PD)→deserializer DES3.

[0030] MAC module 211b switch chip port 4 ( Figure 2 (not shown) The transmitting end Tx4 and the receiving end Rx4 are mapped to the channel lane4 of the optical module 212; wherein the transmitting direction is: serializer Ser4→laser (LD), and the receiving direction is: photodetector (PD)→deserializer DES4.

[0031] In this embodiment, the mapping relationship between the four ports of the MAC module 221b of the device 22 and each lane of the optical module 222 can refer to the mapping relationship in the above device 21. However, the above mapping relationship is only for example and does not limit the specific implementation.

[0032] Port 11 of MAC module 221b ( Figure 2 The transmitting end Tx11 and the receiving end Rx11 of the optical module 222 are mapped to the channel lane1 of the optical module 222; wherein the transmitting direction is: serializer Ser12→laser (LD), and the receiving direction is: photodetector (PD)→deserializer DES11.

[0033] Port 12 of MAC module 221b ( Figure 2 The transmitting end Tx12 and the receiving end Rx12 of the optical module 222 are mapped to the channel lane2 of the optical module 222; wherein the transmitting direction is: serializer Ser12→laser (LD), and the receiving direction is: photodetector (PD)→deserializer DES12.

[0034] MAC module 221b switch chip port 13 ( Figure 2 The mapping between the transmitting end Tx13 and the receiving end Rx13 and the channel lane3 of the optical module 222 is not shown; wherein, the transmitting direction is: serializer Ser3→laser (LD), and the receiving direction is: photodetector (PD)→deserializer DES3.

[0035] MAC module 221b switch chip port 14 ( Figure 2 (not shown) The transmitting end Tx14 and the receiving end Rx14 are mapped to the channel lane4 of the optical module 222; wherein the transmitting direction is: serializer Ser14→laser (LD), and the receiving direction is: photodetector (PD)→deserializer DES14.

[0036] In the device 21 , the CPU 211 a of the switch chip 211 sets the MAC layers of port 1 , port 2 , port 3 , and port 4 of the MAC module 211 b to a forced working (up) state.

[0037] In the device 22 , the CPU 221 a of the switch chip 221 sets the MAC layers of the ports port11 , port12 , port13 , and port14 mapped to the channels lane1 - lane4 to a forced working (up) state.

[0038] The CPU 211a of the switching chip of the device 21 generates a speed reduction capability negotiation message carrying speed reduction capability negotiation information, which includes at least supported rates, number of channels, single channel rate, multiple supported speed reduction rates, and channel combinations corresponding to each supported speed reduction rate.

[0039] In this application, the speed reduction capability negotiation message can be implemented by an LLDP (Link Layer Discovery Protocol) message, which carries the above-mentioned speed reduction capability negotiation information through multiple custom TLV fields.

[0040] For example, in the speed reduction capability negotiation message sent by device 21, the speed reduction capability negotiation information includes: {(400G, 4, 100G); 300G / 200G / 100G}; {array[1]=(2,3,4); array[2]=(1,3,4); array[3]=(1,2,4)}; {array[1]=(3,4); array[2]=(3,4); array[3]=(2,4)}; {array[1]=(4); array[2]=(4); array[3]=(4)}

[0041] {(400G,4,100G);300G / 200G / 100G}(400G,4,100G) means: 400G supported rate, 4 lanes, 100G single lane rate, and three supported reduced speed rates of 300G / 200G / 100G.

[0042] array[1]=(2,3,4) indicates that lane 1 fails and lanes 2, 3, and 4 support a reduced speed of 300G.

[0043] array[2]=(1,3,4) indicates that lane 2 fails and supports the 300G support rate reduction supported by lanes 1, 3, and 4;

[0044] array[3]=(1,2,4) indicates that lane 3 fails and supports the 300G support rate reduction supported by lanes 1, 2, and 4;

[0045] array[1]=(3,4) indicates that lane 1 fails and supports the 300G support rate reduction supported by lanes 1, 2, and 4;

[0046] array[2]=(3,4) indicates that lane 2 fails and supports the 300G support rate reduction supported by lanes 1, 2, and 4;

[0047] array[3]=(1,2) indicates that lane 3 fails and supports the 200G support rate reduction supported by lanes 1 and 2};

[0048] array[4]=(1,2) indicates that lane 4 fails and the supported speed of 200G supported by lanes 1 and 2 is reduced;

[0049] array[1]=(4); indicates that lane 4 fails and supports the 100G support rate reduction supported by lane 4;

[0050] array[2]=(4); indicates that lane 4 fails and supports the 100G support rate reduction supported by lane 4;

[0051] array[3]=(1) indicates that lane 4 fails and the supported speed of lane 1 is reduced to 1200G.

[0052] array[4]=(1) indicates that channel lane 4 fails, and the supported speed reduction rate of 100G supported by channel lane 4 is supported.

[0053] In the speed reduction capability negotiation message sent by device 22, the speed reduction capability negotiation information includes: {(400G, 4, 100G); 300G / 200G / 100G}; {array[1] = (2, 3, 4); array[2] = (1, 3, 4)}; {array[1] = (3, 4); array[2] = (3, 4); array[3] = (2, 4)}; {array[1] = (4); array[2] = (4); array[3] = (4)}

[0054] In the speed reduction capability negotiation message sent by the device 22, the speed reduction capability negotiation information and the speed reduction capability negotiation information in the speed reduction capability negotiation message sent by the device 22 indicate the same speed reduction capability, and no further details are given.

[0055] Devices 21 and 22 communicate via lanes lane 1 to lane 4 that support a rate of 400.

[0056] Figure 3 This is a schematic diagram of an embodiment of notifying a faulty optical module channel provided by the present application.

[0057] The CPU 211 a of the device 21 monitors the signals of RX1 - RX4 of the channels lane 1 - lane 4 respectively, and the CPU 211 a of the device 21 monitors the signals of RX11 - RX14 of the channels lane 1 - lane 4 respectively.

[0058] Devices 21 and 22 cannot detect abnormalities in the TX signals of their respective lanes. However, by detecting abnormalities in the RX signals of their respective lanes, they can detect abnormalities in the TX signals of the same lanes on the other end.

[0059] Figure 3 In the example, the laser LD of channel lane 3 of device 22 fails, and the CPU 211a of device 21 detects that the RX signal of local channel lane 3 is abnormal, and sends a fault channel notification (lane-fault notification, LFN) message 31 carrying the identifier of the fault channel lane 3.

[0060] The CPU 211 a of the device 21 selects a mapped port of one channel among the channels lane 1 - lane 4 to send a fault channel notification message 31 .

[0061] Since the MAC layer of port1, port2, port3, and port4 of device 21 is in a forced working (up) state and the MAC layer of port12, port22, port23, and port24 of device 22 is in a forced working (up) state; the link status of the physical layer of these ports does not affect the logical consideration that these ports are in the working state UP (up / up) on the MAC layer, thereby continuing forwarding and avoiding refreshing of the associated forwarding table entries.

[0062] In one example, CPU 211a of device 21 selects transmit port Tx3 of mapped port 3 of lane 3 in lanes 1-4 and sends a faulty channel notification message 31. Faulty channel notification message 31 is converted into an optical signal by optical module 212 and sent to device 22. The PD of optical module 222 of device 22 converts the received optical signal into an electrical signal and sends it to receive port RX13 of mapped port 13 of lane 3 of device 22 via deserializer DES13.

[0063] If CPU 211a of device 21 selects the transmitting end Port1 of the mapped port of lane 1 in lanes 1-4 to send a faulty channel notification message 31, faulty channel notification message 31 is converted into an optical signal by optical module 212 and then transmitted. The PD of lane 1 of optical module 222 of device 22 converts the received optical signal into an electrical signal and transmits it to the receiving end RX11 of the mapped port Port11 of lane 1 via deserializer DES11 of lane 1.

[0064] After sending the fault channel notification message 31 , the CPU 211 a of the device 21 starts the port speed reduction; after receiving the fault channel notification message 31 , the CPU 221 a of the device 22 starts the port speed reduction.

[0065] Based on the speed reduction capability negotiation information of the devices themselves and the received speed reduction capability negotiation information, devices 21 and 22 select the channel combination with the maximum speed reduction rate supported by both parties after the failure of channel lane 3: the combination of lane 1 and lane 2 that supports a speed reduction rate of 200G.

[0066] In the speed reduction capability negotiation information sent by CPU211a of device 21, the channel combinations supporting the speed reduction rate corresponding to the faulty channel lane3 include: array[3]=(1,2,4) represents the combination of lane1, lane2, and lane4 supporting a speed reduction rate of 300G, array[3]=(2,4) represents the combination of channels lane1 and lane2 supporting a speed reduction rate of 200G, and array[3]=(4) represents the channel lane4 supporting a speed reduction rate of 100G.

[0067] In the speed reduction capability negotiation information sent by CPU221a of device 22, the channel combinations supporting the speed reduction rate corresponding to the faulty channel lane3 include: array[3]=(2,4) represents the combination of 200G channels lane1 and lane2, and array[3]=(4) represents the 100G channel lane4.

[0068] The CPU 211 a of the device 21 and the CPU 221 a of the device 22 select a Lane combination of the maximum supported deceleration rate that is supported by both parties: array[3]=(2,4).

[0069] In another example, if the CPU211a of device 21 and the CPU221a of device 22 fail in channel lane 3, the channel combination corresponding to the faulty channel lane 3 that supports the speed reduction rate cannot obtain an intersection (capability mismatch), which means that the automatic speed reduction of devices 21 and 22 cannot maintain the successful docking of the optical module channel. Then the CPU211a of device 21 and the CPU221a of device 22 will respectively close the ports port3 and port 13 mapped to the faulty channel lane 3, no longer maintain the forced working state of the MAC layer, and refresh the forwarding table items associated with the mapping ports of the faulty channel lane 3.

[0070] Figure 4 This is a schematic diagram of an embodiment of the present application providing a method for restoring a faulty channel to maintain the connection of optical module channels.

[0071] The CPU 211 a of the device 21 detects that the RX signal of the local channel lane 3 is normal, and sends a lane-recover notification (LRN) message 41 carrying an identifier of the restored channel lane 3 .

[0072] The CPU 211 a of the device 21 selects a mapped port of lane 4 from among the channels lane 1 - lane 4 to send a channel recovery notification message 41 .

[0073] Optical module 212 of device 21 converts channel recovery notification message 41 of channel lane 4 into an optical signal and sends it to device 22. PD of channel lane 4 of optical module 222 of device 22 converts the received optical signal into an electrical signal and sends it to receiving end RX14 of mapped port port14 via deserializer DES14 of channel lane 4.

[0074] The CPU 221a of the device 22 receives the channel recovery notification message 41, detects the receiving end RX of the mapped port of each local channel, and determines that the receiving signal of the receiving end RX of the mapped port of local channels lane 1, lane 2, lane 3, and lane 4 is normal; generates a channel recovery confirmation (lane-recover notification ACK, LRN ACK) message 42, and selects one of the four channels lane 1 to lane 4, for example, channel lane 3 to send.

[0075] After sending the channel recovery confirmation message 42, the CPU 221a of the device 22 starts the port speed reduction recovery according to the initially negotiated supported rate of 400G, number of channels of 4, and single channel rate of 100G.

[0076] The CPU 211a of the device 21 receives the channel recovery confirmation message 42 through the corresponding channel lane 3, and starts the port speed reduction recovery according to the initially negotiated supported rate of 400G, number of channels 4, and single channel rate of 100G.

[0077] Figure 5 This is a schematic diagram of an embodiment of a device for achieving optical module channel connection maintenance provided by the present application. The device 50 can be applied to a switching and forwarding device of a network device. The device includes a processor and a storage medium storing machine-executable instructions; the processor performs the following operations by running the machine-executable instructions: setting the mapping port of each local optical module channel to a MAC layer forced working state; sending a first speed reduction capability negotiation message; wherein the first speed reduction capability negotiation message carries local speed reduction capability negotiation information including: local supported rate, local channel number, local single-channel rate, multiple locally supported speed reduction rates, and a channel combination corresponding to each locally supported speed reduction rate; receiving a second speed reduction capability negotiation message from the opposite end; wherein the first speed reduction capability negotiation message carries the opposite end's speed reduction capability negotiation information including: opposite end supported rate, opposite end channel number, opposite end single-channel rate, multiple supported speed reduction rates of the opposite end, and a channel combination corresponding to each opposite end supported speed reduction rate; when detecting that the receiving end signal of the mapping port of any local optical module channel is abnormal, generating a first fault channel notification message and sending it through any local optical module channel; wherein the first fault channel notification message carries a local fault optical module channel identifier; selecting a channel combination with a mutually supported maximum supported speed reduction rate based on each channel combination corresponding to each locally supported speed reduction rate and each channel combination corresponding to the opposite end supported speed reduction rate.

[0078] The processor performs the following operations by running machine executable instructions: when the receiving end signal of the mapping port of the faulty optical module channel is detected to be normal, a first channel recovery notification message is generated and sent through any local optical module channel; wherein, the first channel recovery notification message carries the recovered local optical module channel; when the first channel recovery confirmation message is received through any local optical module channel, the local supported rate is restored.

[0079] The processor performs the following operations by running machine executable instructions: receiving a second fault channel notification message through any local optical module channel; wherein the second fault channel notification message carries the channel identifier of the faulty optical module on the opposite end; and selecting a channel combination with a maximum supported speed reduction rate based on each local supported speed reduction rate corresponding channel combination and each opposite end supported speed reduction rate corresponding channel combination.

[0080] The processor performs the following operations by running machine executable instructions: receiving a second channel recovery notification message; wherein the message carries the optical module channel recovered by the opposite end; detecting whether the receiving signal of the receiving end RX of the mapping port of each local optical module channel is normal; generating a second channel recovery confirmation message and sending it through any local optical module channel.

[0081] The processor executes the following operations by running machine executable instructions: determining that no intersection can be obtained according to each channel combination corresponding to the local supported speed reduction rate and each channel combination corresponding to the opposite end supported speed reduction rate; and closing the mapping port of the local faulty optical module channel or the opposite end faulty optical module channel.

[0082] In this application, a machine-readable storage medium may be any electronic, magnetic, optical, or other physical storage device for storing or containing information (e.g., executable instructions, data, etc.). For example, any machine-readable storage medium herein may be any type of random access memory (RAM), volatile memory, non-volatile memory, flash memory, storage drive (e.g., hard drive), solid-state drive, any type of storage optical disc (e.g., optical disc, DVD, etc.), and similar devices, or a combination thereof. In addition, any machine-readable storage medium herein may be a non-transitory machine-readable storage medium.

[0083] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A method for achieving optical module channel docking, characterized in that: The method comprises Set the mapping port of each local optical module channel to the MAC layer forced working state; Send a first deceleration capability negotiation message; wherein the first deceleration capability negotiation message carries local deceleration capability negotiation information including: local supported rate, number of local channels, local single channel rate, multiple locally supported deceleration rates, and channel combinations corresponding to each locally supported deceleration rate; Receive a second speed reduction capability negotiation message from the peer end; wherein the first speed reduction capability negotiation message carries the speed reduction capability negotiation information of the peer end, including: the peer end supported rate, the number of channels of the peer end, the rate of a single channel of the peer end, multiple supported speed reduction rates of the peer end, and the channel combination corresponding to each of the speed reduction rates supported by the peer end; When detecting that a receiving end signal of a mapping port of any of the local optical module channels is abnormal, generating a first fault channel notification message and sending it through any of the local optical module channels; wherein the first fault channel notification message carries a local fault optical module channel identifier; According to each channel combination corresponding to the local supported speed reduction rate and each channel combination corresponding to the opposite end supported speed reduction rate, an optical module channel with a maximum supported speed reduction rate is selected.

2. The method according to claim 1, characterized in that The method further comprises, When it is detected that the receiving end signal of the mapping port of the faulty optical module channel is normal, a first channel recovery notification message is generated and sent through any of the local optical module channels; wherein the first channel recovery notification message carries the recovered local optical module channel; When a first channel recovery confirmation message is received through any of the local optical module channels, the local supported rate is restored.

3. The method according to claim 1, characterized in that The method further comprises, Receiving a second fault channel notification message through any of the local optical module channels; wherein the second fault channel notification message carries an identifier of the faulty optical module channel at the opposite end; According to each channel combination corresponding to the local supported speed reduction rate and each channel combination corresponding to the opposite end supported speed reduction rate, an optical module channel with a maximum supported speed reduction rate is selected.

4. The method according to claim 3, characterized in that The method further comprises, Receive a second channel recovery notification message, wherein the second channel recovery notification message carries the optical module channel restored by the other end; Detecting that a received signal at a receiving end RX of a mapping port of each local optical module channel is normal; A second channel recovery confirmation message is generated and sent through any of the local optical module channels.

5. The method according to claim 1 or 3, characterized in that The method further comprises, Determining that no intersection can be obtained according to each channel combination corresponding to the local supported speed reduction rate and each channel combination corresponding to the peer end supported speed reduction rate; The mapping port of the local faulty optical module channel or the opposite end faulty optical module channel is closed.

6. A device for achieving optical module channel docking, characterized in that: The device can be applied to a switching and forwarding device of a network device, and includes a processor and a storage medium storing machine-executable instructions; the processor can perform the following operations by running the machine-executable instructions: Set the mapping port of each local optical module channel to the MAC layer forced working state; Send a first deceleration capability negotiation message; wherein the first deceleration capability negotiation message carries local deceleration capability negotiation information including: local supported rate, number of local channels, local single channel rate, multiple locally supported deceleration rates, and channel combinations corresponding to each locally supported deceleration rate; Receive a second speed reduction capability negotiation message from the peer end; wherein the first speed reduction capability negotiation message carries the speed reduction capability negotiation information of the peer end, including: the peer end supported rate, the number of channels of the peer end, the rate of a single channel of the peer end, multiple supported speed reduction rates of the peer end, and the channel combination corresponding to each of the speed reduction rates supported by the peer end; When detecting that a receiving end signal of a mapping port of any of the local optical module channels is abnormal, generating a first fault channel notification message and sending it through any of the local optical module channels; wherein the first fault channel notification message carries a local fault optical module channel identifier; According to each channel combination corresponding to the local supported speed reduction rate and each channel combination corresponding to the opposite end supported speed reduction rate, an optical module channel with a maximum supported speed reduction rate is selected.

7. The device according to claim 6, characterized in that The processor may execute the following operations by executing the machine executable instructions: When it is detected that the receiving end signal of the mapping port of the faulty optical module channel is normal, a first channel recovery notification message is generated and sent through any of the local optical module channels; wherein the first channel recovery notification message carries the recovered local optical module channel; When a first channel recovery confirmation message is received through any of the local optical module channels, the local supported rate is restored.

8. The device according to claim 6, characterized in that The processor may perform the following operations by running the machine executable instructions: receiving a second fault channel notification message through any of the local optical module channels; wherein the second fault channel notification message carries an identifier of a faulty optical module channel at the opposite end; According to each channel combination corresponding to the local supported speed reduction rate and each channel combination corresponding to the opposite end supported speed reduction rate, an optical module channel with a maximum supported speed reduction rate is selected.

9. The device according to claim 8, characterized in that The processor may execute the following operations by executing the machine executable instructions: Receive a second channel recovery notification message, wherein the second channel recovery notification message carries the optical module channel restored by the other end; Detecting that a received signal at a receiving end RX of a mapping port of each local optical module channel is normal; A second channel recovery confirmation message is generated and sent through any of the local optical module channels.

10. The device according to claim 6 or 8, characterized in that The processor may execute the following operations by executing the machine executable instructions: Determining that no intersection can be obtained according to each channel combination corresponding to the local supported speed reduction rate and each channel combination corresponding to the peer end supported speed reduction rate; The mapping port of the local faulty optical module channel or the opposite end faulty optical module channel is closed.

Images