Congestion tolerance method, network device, storage medium and computer program product

Abstract

The present application discloses a congestion fault tolerance method, a network device, a storage medium and a computer program product, and relates to the computer field. The method comprises: when the second device sends congestion, receiving a PFC back pressure frame sent by the second device, determining the PFC state of the second device according to the PFC back pressure frame, and not triggering the ECN congestion mark of the first device when the PFC state of the second device is abnormal. The PFC state of the second device can be determined according to the PFC back pressure frame sent by the second device. When the PFC state of the second device is abnormal, it indicates that there is a risk of PFC deadlock in large-scale networking, and the PFC back pressure frame is frequently triggered, or PFC deadlock is caused. Therefore, the ECN of the first device can be avoided from being affected by the abnormal PFC state, and ECN error marking is performed, thereby improving the fault tolerance of congestion marking.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a congestion fault tolerance method, network equipment, storage medium and computer program product. Background Art

[0002] At present, in order to solve the problem of network congestion. When the bandwidth of the forwarded data packet exceeds the port forwarding capacity, the congestion marking based on explicit congestion notification (ECN) is performed, and a back pressure frame based on priority flow control (PFC) is sent to the previous hop network device. However, in large-scale networking, there is a risk of PFC deadlock. Frequent triggering of PFC back pressure frames or causing PFC deadlock will increase the probability of ECN congestion mechanism being mismarked, causing the ECN congestion mechanism to fall into a chaotic state. Therefore, how to avoid the ECN congestion mechanism from being mismarked and improve the fault tolerance of congestion marking is a problem that needs to be solved urgently. Summary of the invention

[0003] The present application provides a congestion fault tolerance method, a network device, a storage medium and a computer program product, which solve the problem of how to avoid the ECN congestion mechanism from being incorrectly marked and improve the fault tolerance of congestion marking.

[0004] In a first aspect, a congestion fault tolerance method is provided, the method comprising: receiving a priority-based flow control (PFC) back pressure frame sent by a second device; determining a PFC state of the second device according to the PFC back pressure frame; and when the PFC state of the second device is abnormal, not triggering an explicit congestion indication (ECN) congestion mark.

[0005] In combination with the first aspect, in a possible implementation manner, the frequency and / or number of PFC back pressure frames received within a preset period are counted; when the frequency and / or number of PFC back pressure frames is greater than a PFC threshold, it indicates that the PFC state of the second device is abnormal.

[0006] In combination with the first aspect, in another possible implementation, when the PFC state of the second device is normal, the target queue depth of the egress port of the first device is detected; when the target queue depth of the egress port of the first device exceeds the ECN threshold, the ECN congestion mark is triggered.

[0007] In combination with the first aspect, in another possible implementation manner, when the PFC state of the second device is normal, sending data to the second device is stopped.

[0008] In a second aspect, a congestion fault tolerance device is provided, wherein the congestion fault tolerance device includes a receiving module, a determining module and a processing module.

[0009] The receiving module is used to receive a priority-based flow control (PFC) back pressure frame sent by the second device.

[0010] The determination module is used to determine the PFC state of the second device according to the PFC back pressure frame.

[0011] The processing module is used for not triggering a congestion indication (ECN) congestion mark when the PFC state of the second device is abnormal.

[0012] In combination with the second aspect, in a possible implementation manner, the determination module is further used to count the frequency and / or number of PFC back pressure frames received within a preset period; when the frequency and / or number of PFC back pressure frames is greater than the PFC threshold, it indicates that the PFC state of the second device is abnormal.

[0013] In combination with the second aspect, in another possible implementation, the processing module is also used to detect the target queue depth of the egress port of the first device when the PFC state of the second device is normal; when the target queue depth of the egress port of the first device exceeds the ECN threshold, trigger the ECN congestion mark.

[0014] In combination with the second aspect, in another possible implementation manner, the processing module is further configured to stop sending data to the second device when the PFC state of the second device is normal.

[0015] In a third aspect, a computer device is provided, the computer device comprising a processor, an ECN module, a communication interface and a memory, the memory being used to store computer instructions; the communication interface being used to receive PFC back pressure frames; the processor being used to determine the PFC state of a second device according to the frequency and / or number of PFC back pressure frames received within a preset period, and when the processor executes a set of computer instructions, the functions of each module of the method in the first aspect or any possible implementation of the first aspect are executed; the ECN module being used to determine the ECN congestion mark of the network device according to the PFC state of the second device, when the PFC state of the second device is abnormal, the ECN congestion mark is not triggered, and when the PFC state of the second device is normal, the target queue depth of the egress port of the first device is detected; when the target queue depth of the egress port of the first device exceeds the ECN threshold, the ECN congestion mark is triggered.

[0016] In a fourth aspect, a computer-readable storage medium is provided, comprising computer software instructions; when the computer software instructions are executed in a computer, the computer executes a method as described in any one of the first aspect or any possible implementation of the first aspect.

[0017] In a fifth aspect, a computer program product comprising instructions is provided, which, when executed on a computer, enables the computer to execute the method described in the first aspect or any one of the implementations of the first aspect.

[0018] Based on the implementations provided in the above aspects, this application can also be further combined to provide more implementations. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A schematic diagram of a PFC back pressure provided in this application;

[0020] Figure 2 A flowchart of a congestion fault tolerance method provided by this application;

[0021] Figure 3 A schematic diagram of a switching chip provided for this application;

[0022] Figure 4 A schematic diagram of a transmission message provided for this application;

[0023] Figure 5 A schematic diagram of the structure of a congestion fault-tolerant device provided in this application;

[0024] Figure 6 A computer device is provided for this application. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0026] To facilitate understanding of the solutions of the embodiments of the present application, a brief introduction to the relevant concepts is first given as follows:

[0027] Network congestion refers to the situation where the bandwidth for forwarding data packets in the network exceeds the port forwarding capacity. As a result, the data packet forwarding delay increases due to the limited queue buffer resources of the network switch. In severe cases, packet loss and retransmission may occur, further exacerbating the congestion and causing service interruption.

[0028] Priority Flow Control (PFC): When a downstream device in the network finds that its traffic receiving capacity is less than the sending capacity of the first device, it will actively send a Pause frame to the first device, requiring the first device to suspend traffic sending and wait for a period of time before continuing to send.

[0029] like Figure 1As shown, the egress port of the first device is divided into 8 priority queues, and the ingress port of the second device has 8 receive buffers, and the egress port of the first device corresponds to the ingress port of the second device one by one. When a buffer queue on the ingress port of the second device is congested, such as the first queue, and the ingress port queue length of the first queue in the second device exceeds a preset threshold value, the second device triggers PFC and sends a PFC back pressure frame to the first device. The PFC back pressure frame allows the first queue of the first device to suspend sending data packets, while other queues are still sending data normally. When the ingress port queue length of the second device is lower than another set threshold value, the first queue of the second device sends a RESUME frame to allow the first queue of the first device to resume sending data packets.

[0030] However, after the first device receives the PFC back pressure frame, it will stop sending or delay sending data according to the back pressure information indicated by the PFC back pressure frame, and store the data in the local port cache queue. If the local port cache queue consumption exceeds the threshold, it will continue to apply back pressure to the upper-hop device, and so on, until the back pressure points to the source server, thereby eliminating packet loss caused by congestion in network nodes.

[0031] In order to solve the problem of how to avoid the ECN congestion mechanism from being incorrectly marked and improve the fault tolerance of congestion marking, an embodiment of the present application provides a congestion fault tolerance method, that is, when the second device sends congestion, a PFC back pressure frame sent by the second device is received, and the PFC state of the second device is determined according to the PFC back pressure frame. When the PFC state of the second device is abnormal, the ECN congestion marking of the first device is not triggered. The PFC state of the second device can be determined according to the PFC back pressure frame sent by the second device. When the PFC state of the second device is abnormal, it means that in a large-scale network, there is a risk of PFC deadlock, and PFC back pressure frames are frequently triggered, or PFC deadlock is caused. Therefore, the ECN of the first device can be avoided from being affected by the abnormal PFC state, and ECN error marking is performed, thereby improving the fault tolerance of congestion marking.

[0032] Next, the congestion fault tolerance method is described in detail with reference to the accompanying drawings. Figure 2 A flowchart of a congestion fault tolerance method is provided for an embodiment of the present application. Figure 3 A schematic diagram of a switching chip provided in the present application is provided. Here, the first device and the second device are switches as an example for explanation.

[0033] Step 210: Receive a PFC back pressure frame sent by the second device.

[0034] When the bandwidth of the data packet sent by the source server exceeds the port forwarding capability of the second device, the second device makes a determination based on the priority of the received message to determine how to process the message.

[0035] In one implementation, if the priority of the received message has PFC enabled, the message is received, and a PFC back pressure frame is sent to the first device to notify the first device to temporarily stop sending such messages. After receiving the PFC back pressure frame, the first device will temporarily stop sending such messages to the local end, and the length of the pause time is carried by the PFC back pressure frame. When congestion still exists, this process will be repeated until the congestion is relieved.

[0036] In another implementation, if the priority of the received message does not enable PFC, the message is directly discarded.

[0037] like Figure 4 As shown, the source server sends 10Gbps data to the destination server, which is first forwarded by the first device. Since the output port bandwidth of the first device is 10Gbps, the first device is not congested and the ECN congestion mark is not triggered. The first device continues to forward the 10Gbps data to the second device. Since the output port bandwidth of the second device is only 1Gbps, the second device will trigger the ECN congestion mark. Since the output port of the second device continues to be congested, the second device will also send PFC back pressure frames to the first device in the input direction.

[0038] PFC allows the creation of 8 virtual channels on an Ethernet link (that is, 8 virtual channels are created between the first device and the second device), and specifies a corresponding priority for each virtual channel, allowing any virtual channel to be paused and restarted individually, while allowing traffic on other virtual channels to pass uninterrupted.

[0039] Although PFC can implement queue-based flow control by mapping different priorities to different queues, it also introduces new problems, such as PFC deadlock. PFC deadlock refers to a network state in which data flows on all switches are permanently blocked when congestion occurs simultaneously between multiple switches due to loops or other reasons, and the cache consumption of each port exceeds the threshold, while the switches are waiting for each other to release resources.

[0040] Therefore, after receiving the PFC back pressure frame sent by the second device, the first device needs to determine the PFC state of the second device according to the PFC back pressure frame to avoid the first device from making an erroneous ECN congestion marking in the event of PFC deadlock or second device failure.

[0041] Step 220: Determine the PFC state of the second device according to the PFC back pressure frame.

[0042] Under normal circumstances, when the second device is congested, the ingress port of the second device will send PFC back pressure frames to the first device at a preset frequency. When the congestion of the second device is relieved, it will stop sending PFC back pressure frames to the first device. In the event of PFC deadlock or failure of the second device, the second device will continue to send PFC back pressure frames to the first device, causing the first device to perform ECN congestion marking each time it receives a PFC back pressure frame sent by the second device, thereby causing the ECN congestion mechanism to enter a chaotic state. Therefore, after receiving the PFC back pressure frame sent by the second device, the PFC state of the second device is determined according to the PFC back pressure frame.

[0043] Specifically, the first device counts the frequency and / or number of PFC back pressure frames received within a preset period, and the first device receives the PFC back pressure frames sent by the second device at a normal frequency, so that it is normal for the first device to have a small number of PFC back pressure frames. When the number of PFC back pressure frames is obviously too large, and the port traffic accumulated by the PFC is weakened or even stopped, that is, when the frequency and / or number of PFC back pressure frames is greater than the PFC threshold, it indicates that the PFC state of the second device is abnormal. When the frequency and number of PFC back pressure frames are less than the PFC threshold, it indicates that the PFC state of the second device is normal.

[0044] Step 230: When the PFC state of the second device is abnormal, the ECN congestion mark is not triggered.

[0045] Since the PFC state of the second device is abnormal, it continuously sends PFC back pressure frames to the first device. If the first device continuously responds to the PFC back pressure frames sent by the second device, ECN congestion marking is performed. When PFC and ECN are used simultaneously in the network, since the effectiveness time of ECN is longer, PFC usually takes precedence over ECN. However, by performing hop-by-hop back pressure on the first device based on the port, other flows with the same priority are also affected, which in turn causes the ECN of the same priority flow of the previous hop device to be incorrectly marked, causing the ECN congestion to enter a chaotic state. Therefore, when the PFC state of the second device is abnormal, the ECN congestion marking is not triggered, which can effectively improve the fault tolerance of ECN congestion and prevent ECN congestion from entering a chaotic state.

[0046] Step 240: When the PFC state of the second device is normal, perform ECN congestion marking according to the target queue depth of the egress port of the first device.

[0047] When the frequency and number of PFC back pressure frames are less than the PFC threshold, it indicates that the PFC state of the first device is normal, and the first device can respond to the PFC back pressure frame sent by the first device and determine the processing method of the message according to the priority configuration information and state information of the second device.

[0048] If the first device has the PFC function of the corresponding priority enabled and has not yet suspended sending the message of the corresponding priority, the sending of the message of the corresponding priority is suspended, and a timer is started according to the corresponding suspension time in the PFC back pressure frame. When the timer expires, the sending of the message of the corresponding priority is resumed.

[0049] If the first device has enabled the PFC function of the corresponding priority and has suspended sending messages of the corresponding priority, the expiration time of the corresponding timer is updated according to the corresponding pause time in the PFC back pressure frame.

[0050] If the corresponding pause time in the PFC back pressure frame is 0, it is equivalent to causing the corresponding pause timer to expire immediately, and immediately resuming the sending of the corresponding priority message.

[0051] If the corresponding pause time in the PFC back pressure frame is not 0, it is equivalent to resetting the corresponding pause timer. In other words, as long as the second device is always congested, the first device will continue to suspend sending messages of the corresponding priority due to continuous receipt of PFC back pressure frames.

[0052] If the first device does not enable the PFC function of the corresponding priority, the sending of the message of the corresponding priority will not be suspended.

[0053] At the same time, it is also necessary to perform ECN congestion marking according to the target queue depth of the egress port of the first device. When the target queue depth of the egress port of the first device exceeds the threshold, the congestion information of the first device is carried to the second device through the ECN field in the IP message, forwarded by the network devices along the way, and carried to the destination server. The destination server adjusts the sending rate of the source server by sending congestion notification messages to the source server according to the frequency of the received messages carrying the ECN mark.

[0054] The DSCP field in the IP packet header has 2 bits for identifying ECN. These 2 bits represent: support for ECN transmission (ECN Capable Transport, ECT) and congestion (Congestion Experienced, CE). Specifically, when ECT is 0 and CE is 0, it means that the IP packet does not support ECN; when ECT is 0 and CE is 1, it means that the IP packet supports ECN; when ECT is 1 and CE is 0, it means that the IP packet supports ECN; when ECT is 1 and CE is 0, it means that the IP packet supports ECN; when ECT is 1 and CE is 1, it means that the IP packet supports ECN and congestion occurs.

[0055] ECN means that when a message is congested at the exit of a network device, the ECN field in the ECN field of the IP message header that is enabled with ECN (when the ECN field of an IP message is 01 or 10, it indicates that ECN is enabled) will be marked as ECN=11, indicating that the IP message has encountered network congestion and the IP message will not be discarded by the WRED mechanism. If the destination server finds that the ECN field of the IP message is marked as 11, it will immediately generate a congestion notification message and send the message to the source server. The congestion notification message contains information about the congested data flow. After receiving it, the remote server will reduce the corresponding data flow sending rate to resolve network device congestion, thereby avoiding packet loss.

[0056] It is understandable that in order to implement the functions in the above embodiments, the computer includes hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should easily realize that, in combination with the units and method steps of each example described in the embodiments disclosed in this application, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application scenario and design constraints of the technical solution.

[0057] Figure 5 This is a schematic diagram of the structure of the congestion fault tolerance device provided by the embodiment of the present application. These congestion fault tolerance devices can be used to implement the functions of the first device or the second device in the above method embodiment, and thus can also achieve the beneficial effects possessed by the above method embodiment.

[0058] like Figure 5 As shown, the congestion fault tolerance device 500 includes a receiving module, a determining module and a processing module. The congestion fault tolerance device 500 is used to implement the above Figure 2 The functions of the first device or the second device in the method embodiments shown in FIG.

[0059] When the congestion tolerance device 500 is used to implement Figure 2 The functions of the first device or the second device in the method embodiment shown are: a receiving module 501 , a determining module 502 and a processing module 503 .

[0060] The receiving module 501 is used to receive a priority-based flow control (PFC) back pressure frame sent by the second device.

[0061] The determination module 502 is configured to determine a PFC state of the second device according to the PFC back pressure frame.

[0062] The processing module 503 is configured to not trigger a congestion indication (ECN) congestion mark when the PFC state of the second device is abnormal.

[0063] The determination module 502 is further configured to count the frequency and / or number of PFC back pressure frames received within a preset period; when the frequency and / or number of PFC back pressure frames is greater than a PFC threshold, it indicates that the PFC state of the second device is abnormal.

[0064] The processing module 503 is further configured to detect the target queue depth of the egress port of the first device when the PFC state of the second device is normal; and trigger an ECN congestion mark when the target queue depth of the egress port of the first device exceeds an ECN threshold.

[0065] The processing module 503 is further configured to stop sending data to the second device when the PFC state of the second device is normal.

[0066] For more detailed description of the receiving module 501, the determining module 502 and the processing module 503, please refer to Figure 2 The relevant description in the method embodiment shown is directly obtained and will not be repeated here.

[0067] Figure 6 A computer device is provided. Figure 6 The computer device 600 shown can be used to implement the above Figure 5 Functions of the congestion tolerance device 500 in the illustrated embodiment.

[0068] Computer device 600 includes bus 601, processor 602, communication interface 603 and memory 604. Processor 602, memory 604 and communication interface 603 communicate through bus 601. Bus 601 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 6 The communication interface 603 is used for communicating with the outside, for example, receiving IP packets or PFC back pressure frames.

[0069] The processor 602 may be a central processing unit (CPU), and the processor 602 is used to receive a PFC back pressure frame sent by the second device when the second device sends congestion, determine the PFC state of the second device according to the PFC back pressure frame, and when the PFC state of the second device is abnormal, do not trigger the ECN congestion mark of the first device, and when the PFC state of the second device is normal, detect the target queue depth of the egress port of the first device; when the target queue depth of the egress port of the first device exceeds the ECN threshold, trigger the ECN congestion mark. The memory 604 may include a volatile memory, such as a random access memory (RAM). The memory 604 may also include a non-volatile memory, such as a read-only memory (ROM), a flash memory, an HDD or an SSD.

[0070] The memory 604 stores executable codes, and the processor 602 executes the executable codes to perform the aforementioned congestion fault tolerance method.

[0071] Specifically, in implementing Figure 5 In the case of the embodiment shown, and Figure 5 When each module described in the embodiment is implemented by software, the memory 604 stores the execution Figure 5 The software or program code required for the functions of the receiving module 501, the determining module 502 and the processing module 503, the processor 602 is used to execute the instructions in the memory 604, and execute the congestion fault tolerance method applied to the congestion fault tolerance device 500.

[0072] The present application also provides a computer-readable storage medium, comprising instructions, which, when executed on a computer, enables the computer to execute the above-mentioned congestion fault tolerance method applied to the congestion fault tolerance device 400.

[0073] The present application also provides a computer program product, and when the computer program product is executed by a computer, the computer executes any of the aforementioned methods. The computer program product may be a software installation package, and when any of the aforementioned methods is needed, the computer program product may be downloaded and executed on a computer.

[0074] It should also be noted that the device embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the drawings of the device embodiments provided by the present application, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines.

[0075] Through the description of the above implementation mode, the technicians in the field can clearly understand that the present application can be implemented by means of software plus necessary general hardware, and of course, it can also be implemented by special hardware including special integrated circuits, special CPUs, special memories, special components, etc. In general, all functions completed by computer programs can be easily implemented by corresponding hardware, and the specific hardware structure used to implement the same function can also be various, such as analog circuits, digital circuits or special circuits. However, for the present application, software program implementation is a better implementation mode in more cases. Based on such an understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a readable storage medium, such as a computer floppy disk, a U disk, a mobile hard disk, a ROM, a RAM, a disk or an optical disk, etc., including a number of instructions to enable a computer device (which can be a personal computer, a training device, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0076] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments may be implemented in the form of a computer program product.

[0077] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website site, a computer, a training device, or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, training device, or data center. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a training device, a data center, etc. that includes one or more available media integrations. The available medium may be a magnetic medium, (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)), etc.

Claims

1. A network device, It is characterized in that The network device comprises a processor and an ECN module for displaying congestion indication, wherein the processor is used for: Receiving a priority-based flow control PFC back pressure frame sent by the second device; Determine the PFC state of the second device by counting the frequency and / or number of the PFC back pressure frames received within a preset period, wherein when the frequency and / or number is greater than a preset PFC threshold, it is determined that the PFC state of the second device is abnormal, indicating that there is a PFC deadlock or a second device failure; when the frequency and / or number is less than the preset PFC threshold, it is determined that the PFC state of the second device is normal; When the PFC state of the second device is normal, if the network device has enabled the PFC function of the corresponding priority and has not yet suspended sending the message of the corresponding priority, suspend sending the message of the priority, and start the timer according to the pause time in the PFC back pressure frame, and resume sending after the timer expires; if sending has been suspended, update the expiration time of the timer according to the pause time; if the pause time is 0, resume sending immediately; if the PFC function of the corresponding priority is not enabled, do not suspend sending; The ECN module is used to: determine the ECN congestion mark of the network device according to the PFC state of the second device; When the PFC state of the second device is abnormal, the congestion indication ECN congestion mark is not triggered; When the PFC state of the second device is normal, an ECN congestion mark is triggered according to a target queue depth of an egress port of the network device, wherein when the target queue depth exceeds a preset depth threshold, an ECN field in a header of an IP message supporting ECN is set to a congestion state, and the message is forwarded to a destination server through network devices along the way. The destination server sends a congestion notification message to a source server according to a received ECN mark frequency, so as to adjust a sending rate of the source server.

2. A congestion tolerance method, It is characterized in that Applied to a first device, comprising: When the ECN of the first device is not marked, receiving a priority-based flow control PFC back pressure frame sent by the second device; Determine the PFC state of the second device by counting the frequency and / or number of the PFC back pressure frames received within a preset period, wherein when the frequency and / or number is greater than a preset PFC threshold, it is determined that the PFC state of the second device is abnormal, indicating that there is a PFC deadlock or a second device failure; when the frequency and / or number is less than the preset PFC threshold, it is determined that the PFC state of the second device is normal; When the PFC state of the second device is abnormal, the congestion indication ECN congestion mark is not triggered; When the PFC state of the second device is normal, an ECN congestion mark is triggered according to the target queue depth of the egress port of the first device, wherein when the target queue depth exceeds a preset depth threshold, the ECN field of the ECN-supporting IP message header is set to a congestion state, and the message is forwarded to the destination server through the network devices along the way. The destination server sends a congestion notification message to the source server according to the received ECN mark frequency to adjust the sending rate of the source server; When the PFC state of the second device is normal, if the first device has enabled the PFC function of the corresponding priority and has not yet suspended sending the message of the corresponding priority, then suspend sending the message of the priority, and start the timer according to the pause time in the PFC back pressure frame, and resume sending after the timer expires; if sending has been suspended, update the expiration time of the timer according to the pause time; if the pause time is 0, resume sending immediately; if the PFC function of the corresponding priority is not enabled, then do not suspend sending.

3. A computer-readable storage medium, It is characterized in that Used to store computer instructions, which, when executed on a computer device, cause the computer to execute the method as claimed in claim 2.

4. A computer program product, comprising a computer program, which implements the method of claim 2 when executed by a processor.

Images