Electrical interconnection link parameter adjustment system, server and method
By introducing processing components and storage components into the host device and combining the neural network model to dynamically adjust the completion timeout value of the slave device, the problem of low efficiency in adjusting the parameters of the electrical interconnection link is solved, and the stability and reliability of the system are improved.
Patent Information
- Application Number
- CN202511195179.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-25
- Publication Date
- 2025-09-26
- Estimated Expiration
- 2045-08-25
AI Technical Summary
In the prior art, the parameter adjustment efficiency of the electrical interconnection link is low and the configuration method is fixed, resulting in a large number of completion timeout errors and being unable to effectively ensure the continuous and stable operation of the system.
By introducing processing components and storage components into the host device, the operating parameters of the processing unit are monitored in real time, the completion timeout value of the slave device is dynamically adjusted, and a pre-trained neural network model is used to predict the most suitable completion timeout value based on the operating parameters to achieve adaptive optimization.
It improves the stability and reliability of the electrical interconnection link, reduces system errors and interruptions, and ensures the continuous and stable operation of the system.
Smart Images

Figure CN120704935A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of computer systems and information interaction, and in particular to a system, server, and method for adjusting parameters of an electrical interconnection link. Background Art
[0002] PCIe (Peripheral Component Interconnect Express), a high-speed serial point-to-point dual-channel computer expansion bus standard, is divided into a host and a slave side. After a slave device issues a request, it must wait for feedback from the host before continuing. Under high traffic volumes and slow host processing, if the default completion timeout expires, data may not be fully returned, resulting in an unrecoverable error. This can cause the slave device to wait for an abnormally long time, impacting stability and, in severe cases, even causing system downtime.
[0003] However, in the related art, the device location is manually found and the completion timeout value of the electrical interconnection device is set using commands or scripts. However, this method is inefficient, labor-intensive, and has a fixed configuration, making it difficult to adapt to complex scenarios. Therefore, there are a large number of completion timeout errors, which cannot effectively ensure the continuous and stable operation of the system. Summary of the Invention
[0004] The present application provides a system, server, method, electronic device, storage medium, and program product for adjusting parameters of an electrical interconnection link, to at least address the problems in the related art of adjusting parameters of an electrical interconnection link, such as low efficiency and fixed configuration, the continued occurrence of a large number of completion timeout errors, and the inability to effectively ensure the continuous and stable operation of the system.
[0005] The present application also provides an electrical interconnection link parameter adjustment system, wherein the two ends of the electrical interconnection link are respectively connected to a host device and at least one slave device, the host device includes: a processing component for processing a request message of the slave device and returning the message processing result to the slave device, wherein the processing component includes a processing unit, a first register and a second register, the first register stores error information of a request message completion timeout, the second register stores a completion timeout value of each slave device, and when the processing unit completes processing of the request message for a timeout exceeding the completion timeout value, generates error information of a request message completion timeout; a storage component for obtaining operating parameters of the processing unit, adjusting the completion timeout value corresponding to the slave device according to the operating parameters of the processing unit, reading the first register and the second register, and if the error information of the request message completion timeout is read, determining the corresponding slave device according to the error information of the request message completion timeout, and adjusting at least one of the completion timeout value corresponding to the slave device and the operating parameters of the processing unit.
[0006] The present application also provides a server, comprising the above-mentioned electrical interconnection link parameter adjustment system.
[0007] The present application provides a method for adjusting parameters of an electrical interconnection link, which is applied to the above-mentioned electrical interconnection link parameter adjustment system. The two ends of the electrical interconnection link are respectively connected to a host device and at least one slave device. The host device includes a processing component and a storage component, wherein the storage component is configured to perform the following steps: obtaining operating parameters of a processing unit; adjusting a completion timeout value corresponding to the slave device according to the operating parameters of the processing unit; reading a first register and a second register, wherein the first register stores error information of a request message completion timeout, and the second register stores a completion timeout value of each slave device. When the processing unit completes processing a request message for a time exceeding the completion timeout value, an error message of a request message completion timeout is generated; if the error message of a request message completion timeout is read, the corresponding slave device is determined according to the error message of the request message completion timeout, and at least one of the completion timeout value corresponding to the slave device and the operating parameters of the processing unit is adjusted.
[0008] The present application also provides a method for adjusting parameters of an electrical interconnection link, which is applied to the above-mentioned electrical interconnection link parameter adjustment system, wherein the two ends of the electrical interconnection link are respectively connected to a host device and at least one slave device, and the host device includes a processing component and a storage component, wherein the storage component is configured to perform the following steps: obtaining the completion timeout value corresponding to the slave device when the processing unit has different operating parameters; generating a correspondence between the operating parameters and the completion timeout value according to the different operating parameters and the completion timeout value corresponding to the slave device, and generating a training data set according to the operating pressure value, the completion timeout value corresponding to the slave device, and the corresponding relationship; using the training data set to train a neural network model, and using the trained neural network model and the operating parameters of the processing unit to adjust the completion timeout value corresponding to the slave device.
[0009] The present application also provides an electronic device, comprising: a memory for storing a computer program; and a processor for implementing the steps of any of the above-mentioned methods for adjusting parameters of an electrical interconnection link when executing the computer program.
[0010] The present application also provides a non-volatile computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned method for adjusting the parameters of the electrical interconnection link are implemented.
[0011] The present application also provides a computer program product, including a computer program, which implements the steps of the above-mentioned method for adjusting the parameters of the electrical interconnection link when the computer program is executed by a processor.
[0012] Through the present application, the embodiments of the present application can monitor the operating parameters of the processing unit in real time, dynamically adjust the completion timeout value corresponding to the slave device based on the real-time monitored parameters, and reduce system errors and interruptions caused by completion timeout value timeouts. Therefore, the completion timeout value can be dynamically adjusted according to the operating status of the processing unit, and the stability and reliability of the electrical interconnection link can be improved through real-time dynamic adjustment. After adjusting the completion timeout value, the first register is monitored in real time. If an error message still exists, the completion timeout error message is corrected by adjusting at least one of the completion timeout value corresponding to the slave device and the operating parameters of the processing unit, further reducing system errors and interruptions caused by completion timeout value timeouts. Therefore, the problem of low efficiency and fixed configuration of electrical interconnection link parameter adjustment in related technologies can be solved, which still exists in a large number of completion timeout error messages and cannot effectively ensure the continuous and stable operation of the system. BRIEF DESCRIPTION OF THE DRAWINGS
[0013] In order to more clearly illustrate the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.
[0014] Figure 1 A flowchart of a method for configuring a timeout for completing electrical interconnection provided in the related art; Figure 2 A schematic diagram of the structure of an electrical interconnection link parameter adjustment system provided in an embodiment of the present application; Figure 3 A schematic diagram of the overall solution of the electrical interconnection link parameter adjustment system provided in an embodiment of the present application; Figure 4 A flowchart of a method for adjusting parameters of an electrical interconnection link provided in one embodiment of the present application; Figure 5 A schematic diagram of the implementation process of the method for adjusting the parameters of an electrical interconnection link provided in an embodiment of the present application; Figure 6 A flowchart of a method for adjusting parameters of an electrical interconnection link provided in another embodiment of the present application; Figure 7 A schematic structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0015] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.
[0016] It should be noted that, in the description of this application, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. The terms "first," "second," etc., in this application are used to distinguish similar objects, and are not used to describe a particular order or sequence.
[0017] In related technologies, such as Figure 1 As shown, by adding single-device and global device options to the system settings, the Basic Input / Output System (BIOS) enumerates interconnected devices at system startup and sets the optimal completion timeout value they support to improve device stability. However, this technology is only configured at the initial system startup and cannot adjust to dynamic changes in business volume and CPU (Central Processing Unit) load during operation. This can still cause timeouts and errors, making it difficult to ensure continuous and stable system operation.
[0018] To this end, the embodiments of the present application propose electrical interconnection link parameter adjustment systems, devices, methods, media and products. In order to enable technicians in this technical field to better understand the application scheme, the application is further described in detail below with reference to the accompanying drawings and specific implementation methods.
[0019] Figure 2 A schematic diagram of the structure of an electrical interconnection link parameter adjustment system provided in an embodiment of the present application is shown in FIG. Figure 2 As shown, both ends of the electrical interconnection link of the electrical interconnection link parameter adjustment system 10 are connected to a host device 400 and at least one slave device 100 respectively. The host device 400 specifically includes: a processing component 200 and a storage component 300.
[0020] Among them, the processing component 200 is used to process the request message of the slave device 100 and return the message processing result to the slave device 100. Among them, the processing component 200 includes a processing unit 201, a first register 202 and a second register 203. The first register 202 stores the error information of the request message completion timeout, and the second register 203 stores the completion timeout value of each slave device 100. When the completion time of the request message processed by the processing unit 201 exceeds the completion timeout value, an error information of the request message completion timeout is generated; the storage component 300 is used to obtain the operating parameters of the processing unit 201, adjust the completion timeout value corresponding to the slave device 100 according to the operating parameters of the processing unit 201, read the first register 202 and the second register 203, and if the error information of the request message completion timeout is read, determine the corresponding slave device 100 according to the error information of the request message completion timeout, and adjust at least one of the completion timeout value corresponding to the slave device 100 and the operating parameters of the processing unit 201.
[0021] The electrical interconnection link is a data transmission path connecting the host device 400 and the slave device 100. In specific applications, it can be a PCIe link, used to enable communication and information exchange between the devices at both ends. The host device 400 is the device at one end of the electrical interconnection link, responsible for managing and processing communication requests with the slave device 100. The slave device 100 can be a PCIe device. The host device 400 typically includes a processing component 200 and a storage component 300. The slave device 100 is the device at the other end of the electrical interconnection link, sending request messages to the host device 400 and awaiting processing results, acting as the communication partner of the host device 400. The processing component 200, consisting of a processing unit 201 and registers, is responsible for receiving and processing request messages from the slave device 100 and returning the processing results to the corresponding slave device 100. The processing unit 201 can be a CPU module, responsible for executing instructions, processing data, and controlling system operation. The first register 202 can be an Uncorrectable Error Status Register, which is used to store error information indicating a request message completion timeout. When the request processing time exceeds the preset completion timeout value, the processing unit 201 records a timeout error status in this register. The second register 203, which can be called a Device Control 2 Register, is used to store the completion timeout value corresponding to each slave device 100. The completion timeout value is the maximum time threshold allowed by the system for completing the processing of a request message. The error message is generated by the processing unit 201 when the processing time exceeds the preset completion timeout value, reflecting the abnormal request processing timeout. The storage component 300 is the module in the system responsible for storing and managing data. Specifically, it obtains the operating parameters of the processing unit 201, reads the contents of the first register 202 and the second register 203, and adjusts the completion timeout value of the slave device 100 and the operating parameters of the processing unit 201 based on the error message and operating parameters. The operating parameters are indicators of the current operating status of the processing unit 201 and are used to guide the dynamic adjustment of the completion timeout value. The request message is a data communication request sent by the slave device 100 to the host device 400, used to implement information exchange and operation instruction transmission between devices. A timeout error is an abnormal state caused by the failure to complete the processing of a request message within the maximum allowable time specified by the system. It usually triggers the error handling mechanism.
[0022] It can be understood that the embodiments of the present application can monitor the operating parameters of the processing unit in real time, dynamically adjust the completion timeout value corresponding to the slave device according to the parameters monitored in real time, and reduce system errors and interruptions caused by the completion timeout value timing out. Therefore, the completion timeout value can be dynamically adjusted according to the operating status of the processing unit, and the stability and reliability of the electrical interconnection link can be improved through real-time dynamic adjustment. After the completion timeout value is adjusted, the first register is monitored in real time. If the error message still exists, the completion timeout error is corrected by adjusting at least one of the completion timeout value corresponding to the slave device and the operating parameters of the processing unit, thereby further reducing system errors and interruptions caused by the completion timeout value timing out.
[0023] In one embodiment of the present application, the storage component 300 is provided with a pre-trained neural network model, the operating parameters of the processing unit 201 are input into the neural network model, and the neural network model outputs the target value of the completion timeout value corresponding to the slave device 100, wherein the neural network model includes a correspondence between the operating parameters and the completion timeout value, and the target value of the completion timeout value corresponding to the operating parameters is determined based on the correspondence.
[0024] Among them, the neural network model is a model formed after the deep learning model is trained with historical data, and can predict or output the corresponding completion timeout value target based on the input operating parameters. The operating parameter input is to use the current working status data of the processing unit 201, such as CPU load, task queue length, etc., as the input of the neural network for the model to calculate and predict. The target value of the completion timeout value is the completion timeout threshold value that is most suitable for the current system state, which is predicted by the neural network model based on the operating parameters, and is used to dynamically adjust the system parameters. The corresponding relationship is the mapping rule between the operating parameters and the completion timeout value learned within the neural network.
[0025] As will be appreciated, storage component 300 incorporates a pre-trained neural network model. By studying a large amount of historical operational data, the model has mastered the complex correspondence between processing unit 201 operating parameters and the completion timeout value of slave device 100. By inputting the current operating parameters of processing unit 201 into this neural network model, the model intelligently predicts the target completion timeout value that best suits the current operating state. This enables dynamic and precise timeout adjustment, avoiding the performance bottlenecks and error risks associated with traditional static configurations, and thus reducing the need for manual adjustments.
[0026] In one embodiment of the present application, the host device 400 further includes at least one expansion component, which is connected to the slave device 100 , and the processing component 200 is connected to the expansion component via an electrical interconnection link.
[0027] The expansion component is an additional module or device added to the host device 400, used to expand system functionality and connect more slave devices 100, thereby improving the system's scalability and flexibility. The electrical interconnection link is a physical and logical connection established between the processing component 200 of the host device 400 and the expansion component via the electrical interconnection link, supporting the transmission of data and control signals.
[0028] It can be understood that the introduction of the expansion component enhances the master device 400's access capabilities to multiple slave devices 100, improving the system's scalability and flexibility. The expansion component is connected to the processing component 200 via an electrical interconnect link, enabling efficient data transmission and collaborative processing, supporting the access of more peripherals, meeting complex application requirements, and improving overall system performance and maintainability.
[0029] like Figure 3 As shown, the electrical interconnection circuit parameter adjustment system includes multiple PCIe add-in cards, a processing unit, and a storage component 300. The left side of the figure shows PCIe add-in cards 1 through N, each connected to the central processing unit via a PCIe bus. The processing unit is responsible for enumerating each peripheral device and collecting its hardware and link information (such as device identification, current configuration, and link status). It then provides this PCIe add-in card information to the storage component 300 on the right. Based on this collected information and system policies or rules, the storage component 300 calculates the parameters of the second register of the PCIe device on the processing unit that needs adjustment and sends the adjustment instructions back to the processing unit, which then writes the corresponding device register to update the configuration. Through this closed loop of "information-decision-dispatching-effectiveness," the system can dynamically optimize PCIe-related parameters when multiple cards coexist and the load varies, thereby improving the stability and reliability of device operation.
[0030] The electrical interconnection link parameter adjustment system of the embodiment of the present application is composed of a host device connected to at least one slave device, the host device includes a processing component and a storage component, the processing component is composed of a processing unit, a first register and a second register, the processing component is used to process the request message of the slave device and return the result, the first register stores the error information of the request message completion timeout, and the second register stores the completion timeout value of each slave device. When the processing unit detects that the request message completion time exceeds the set value, a timeout error message is generated; the storage component is used to obtain the operating parameters of the processing unit and adjust the completion timeout value of the slave accordingly, and can read the first register and the second register at the same time. When the timeout error message is detected, the corresponding slave device is determined, and at least one of the completion timeout value of the slave device and the operating parameters of the processing unit is adjusted.
[0031] An embodiment of the present application further provides a server comprising the above-mentioned electrical interconnection link parameter adjustment system.
[0032] The embodiment of the present application also provides a method for adjusting parameters of an electrical interconnection link, such as Figure 4 As shown, the method is applied to the above-mentioned electrical interconnection link parameter adjustment system, where the two ends of the electrical interconnection link are respectively connected to a host device and at least one slave device, and the host device includes a processing component and a storage component, wherein the storage component is configured to perform the following steps: In step S101 , the operating parameters of the processing unit are obtained.
[0033] Among them, the operating parameters may include register values at the hardware level, such as performance counter registers, error status registers, timeout counter registers, etc., as well as monitoring indicators at the software level, such as task queue length, processing delay, CPU occupancy, etc.
[0034] It can be understood that by obtaining the operating parameters of the processing unit, the embodiment of the present application can grasp the current working status of the processing unit in real time, including information such as operating pressure, error status and resource occupancy, which is helpful to make predictive adjustments before performance degradation or link abnormalities occur.
[0035] In step S102 , the completion timeout value corresponding to the slave device is adjusted according to the operating parameters of the processing unit.
[0036] It is understandable that when the processing unit load is high, the timeout value is appropriately extended to reduce completion timeout errors; when the load is low, the timeout value is shortened to speed up anomaly detection and retry, improve link utilization, and adaptively adjust the completion timeout value to avoid the problem of waiting too long or too little due to a fixed timeout value.
[0037] In one embodiment of the present application, the completion timeout value corresponding to the slave device is adjusted according to the operating parameters of the processing unit, including: a pre-trained neural network model; inputting the operating parameters of the processing unit into the neural network model, and the neural network model outputting a target value of the completion timeout value corresponding to the slave device, wherein the neural network model includes a correspondence between the operating parameters and the completion timeout value, and the target value of the completion timeout value corresponding to the operating parameters is determined based on the correspondence.
[0038] The neural network model is a dynamic processing unit stress model, used to monitor and analyze the stress state of the processing unit under different workloads in real time. By inputting various current operating parameters of the processing unit (such as utilization, task queue length, temperature, etc.), the trained neural network model estimates the stress level of the processing unit, helping the system dynamically adjust relevant configurations (such as slave device timeout parameters). The correspondence is a functional mapping or data association between operating parameters and completion timeout values, reflecting the dependence and influence between the two. The neural network makes predictions based on this relationship.
[0039] It can be understood that by matching the real-time operating parameters of the processing unit with the rules extracted from historical data through a pre-trained neural network model, the most appropriate target value for the completion timeout of the slave device can be predicted quickly and accurately, avoiding the inaccuracy and lag caused by manually setting parameters, and automatically optimizing the timeout configuration according to the dynamic changes in the system operating status.
[0040] In one embodiment of the present application, before inputting the operating parameters of the processing unit into the neural network model, it also includes: obtaining the completion timeout value corresponding to the slave device when the processing unit has different operating parameters; generating a correspondence between the operating parameters and the completion timeout value based on the different operating parameters and the completion timeout value corresponding to the slave device, and generating a training data set based on the operating pressure value, the completion timeout value corresponding to the slave device and the correspondence; and training the neural network model using the training data set.
[0041] As can be understood, by collecting the completion timeout values corresponding to processing units in different operating states, establishing a mapping relationship between operating parameters and completion timeout values, and generating a high-quality training dataset based on this mapping relationship, a neural network model capable of accurately predicting optimal completion timeout values was trained. This improves the accuracy and adaptability of completion timeout adjustment, enabling the system to dynamically optimize parameter configuration based on real-time operating pressure.
[0042] During the operation of the storage component, the system dynamically trains the neural network model of the processing unit and monitors the status of the first registers in all slave devices under the processing unit in real time. Specifically, if the first register does not detect an error, that is, the completion timeout status bit is 0, it indicates that the slave device's request has normally received a response from the processing unit within the configured completion timeout, the electrical interconnection link is communicating normally, and the external card is functioning normally. At this time, the storage component records the current processing unit pressure value and the corresponding completion timeout value and generates corresponding data for establishing a training model between the processing unit pressure and the completion timeout value. If the first register in the processing unit detects an error, that is, the completion timeout status bit is 1, indicating that a completion timeout error has occurred, indicating that the external card has not received a response from the processing unit within its configured completion timeout range, then step S103 is executed.
[0043] In step S103, the first register and the second register are read, wherein the first register stores error information of request message completion timeout, and the second register stores the completion timeout value of each slave device. When the processing unit processes the request message completion timeout when the completion time exceeds the completion timeout value, it generates error information of request message completion timeout.
[0044] It is understandable that by reading the request message timeout error information stored in the first register and the completion timeout value stored in the second register, real-time monitoring of the request processing time can be achieved.
[0045] The storage component obtains the value of the first register in the processing unit in real time and checks the completion timeout status bit therein. When the first register of the processing unit detects an error, the storage component reads the value of the second register of the slave device where the error occurred and obtains the currently configured completion timeout value from it. This embodiment of the application constructs a specific encoding correspondence between the completion timeout value in the register according to the PCIe4.0 protocol specification, and the specific content is shown in Table 1.
[0046] Table 1
[0047] According to the PCIe4.0 protocol specification, the timeout value may include the following ranges: 50µs-50ms, 50µs-100µs, 1ms-10ms, 16ms-55ms, 65ms-210ms, 260ms-900ms, 1s-3.5s, 4s-13s, and 17s-64s.
[0048] By analyzing the obtained completion timeout value, the storage component adjusts the parameters of the second register in the external card and appropriately extends the completion timeout to ensure that the processing unit has enough time to process the request.
[0049] In addition, the system dynamically adjusts the completion timeout of the external card according to the real-time pressure of the processing unit, ensuring that the request sent by the slave device can be processed by the processing unit in time within the waiting time, thereby improving the stability of the system and communication efficiency.
[0050] In step S104, if error information indicating that the request message has timed out is read, the corresponding slave device is determined based on the error information indicating that the request message has timed out, and at least one of the completion timeout value corresponding to the slave device and the operating parameters of the processing unit is adjusted.
[0051] It's understandable that by monitoring the processing unit's response time to request messages and promptly identifying timeouts, data communication failures and system performance degradation caused by processing delays can be effectively prevented. By identifying the specific slave device experiencing a timeout and dynamically adjusting its completion timeout value and the processing unit's operating parameters, the system achieves adaptive optimization, improving overall communication stability and reliability, avoiding service interruptions caused by timeout errors, and ensuring the efficient operation of storage components.
[0052] In one embodiment of the present application, before reading the error information of the request message completion timeout, the method further includes: obtaining the electrical interconnection protocol error information of the electrical interconnection link; parsing the error type in the electrical interconnection protocol error information; and when determining the error of the request message completion timeout according to the error type, reading the error information of the request message completion timeout.
[0053] Among them, the electrical interconnection protocol can be a PCIe protocol. In the electrical interconnection protocol, the error types include the request not being completed within the set timeout value, the transaction layer packet format being illegal, the extended cyclic redundancy check failing, and the receipt of an unsupported request type.
[0054] It is understood that before reading the error message related to the request message completion timeout, first obtaining and parsing the electrical interconnection protocol error message of the electrical interconnection link can preemptively determine the nature and source of the link anomaly. By parsing the error type, the corresponding timeout error message can be read only after confirming that the error is a completion timeout issue. This reduces wasted reading and resource consumption, and identifies the fault type as soon as it occurs.
[0055] In one embodiment of the present application, adjusting at least one of the completion timeout value corresponding to the slave device and the operating parameters of the processing unit includes: obtaining the maximum value of the completion timeout value corresponding to the slave device; if the completion timeout value corresponding to the slave device is less than the maximum value, adjusting the completion timeout value corresponding to the slave device; if the completion timeout value corresponding to the slave device is the maximum value, adjusting the operating parameters of the processing unit.
[0056] It is understandable that the embodiments of the present application can implement a hierarchical and flexible optimization strategy when processing the request message completion timeout problem. By preferentially adjusting the completion timeout value corresponding to the slave device, more time can be reserved for the processing of the request, reducing the risk of timeouts caused by link jitter, instantaneous congestion, etc. When the completion timeout value has reached the maximum value, the system can intelligently switch to a solution for adjusting the operating parameters of the processing unit, reducing the load pressure of the processing unit from the source, thereby indirectly shortening the request response time. The dual adjustment mechanism can not only ensure that the optimal response measures are taken under different operating environments, but also avoid system instability caused by the failure of a single means.
[0057] In one embodiment of the present application, adjusting the completion timeout value corresponding to the slave device includes: calculating the difference between the completion timeout value corresponding to the slave device and the maximum value; if the difference is less than a preset adjustment step, increasing the completion timeout value corresponding to the slave device to the maximum value; if the difference is greater than or equal to the preset adjustment step, increasing the completion timeout value corresponding to the slave device according to the adjustment step.
[0058] It is understandable that when the difference between the completion timeout value and the maximum value is small, directly increasing it to the maximum value can quickly extend the request waiting time, ensure that the system can respond to extreme load conditions in a timely manner, and avoid the performance overhead caused by repeated small adjustments; when the difference is large, gradually increasing the completion timeout value according to the preset adjustment step size will help to smooth the transition and reduce the instability or abnormalities that may be caused by the rapid change of the timeout parameters.
[0059] In one embodiment of the present application, adjusting the operating parameters of the processing unit includes: identifying an operating pressure value in the operating parameters; obtaining a minimum value of the operating pressure value of the processing unit; if the operating pressure value of the processing unit is greater than the minimum value, reducing the operating pressure value of the processing unit; if the operating pressure value of the processing unit is the minimum value, generating an error alarm signal of the electrical interconnection link.
[0060] It can be understood that the embodiment of the present application monitors and actively reduces the operating pressure of the processing unit to avoid performance degradation and request timeouts due to excessive load, thereby ensuring stable communication of the slave device; when the pressure has been reduced to the lowest level but is still abnormal, the system automatically alarms and reminds maintenance in time to improve the security and reliability of the link.
[0061] Specifically, during the process of adjusting the completion timeout value, it is first determined whether the completion timeout value of the slave device has reached the maximum value. If the completion timeout value has not reached the maximum value, the storage component gradually increases the completion timeout period by writing a new configuration value to the second register of the slave device to extend the time the request message waits for the processing unit to respond, and continues to perform subsequent operations to verify the adjustment effect. Subsequently, the error information in the first register is cleared, and step S102 is re-executed. During operation, the error status of the slave device is monitored in real time, and the processing unit neural network model is dynamically trained based on the processing unit pressure value and completion timeout value data collected during normal communication. If the completion timeout value of the slave device reaches its maximum value, the system cannot adjust the value by increasing it. Instead, it can only adjust the operating parameters of the processing unit. The steps for adjusting the operating parameters of the processing unit include: first, the system identifies the operating pressure value in the current operating parameters and obtains the minimum allowable operating pressure value of the processing unit. Then, if the current operating pressure value is greater than the minimum value, the system takes measures to reduce the operating pressure to reduce the load on the processing unit and improve processing efficiency. Conversely, if the operating pressure value has been reduced to the minimum value but the completion timeout problem has not been resolved, the system proactively generates an error alarm signal for the electrical interconnection link.
[0062] That is, if it is found during the adjustment process that the pressure value of the processing unit is still high, that is, not the minimum value, the storage component will actively reduce the business volume of the storage device, thereby reducing the load of the processing unit, and clear the error information of the first register, and then re-execute step S102. The storage component monitors the error status of the slave device in real time during operation, and dynamically trains the processing unit neural network model based on the processing unit pressure value and completion timeout value data collected during normal communication.
[0063] If the pressure on the processing unit has been reduced to the minimum value and a completion timeout error still occurs, it means that even when the processing unit is not under pressure, the completion timeout value of the slave device has reached the maximum value and the problem still exists. At this time, the storage component will issue an alarm signal, indicating that developer intervention is required to check and maintain the stability and reliability of the electrical interconnection link to ensure the normal operation of the system.
[0064] In one embodiment of the present application, after adjusting the completion timeout value corresponding to the slave device, the method further includes: clearing error information of completion timeout of the request message corresponding to the slave device from the first register.
[0065] It is understood that the embodiments of the present application can ensure that the system status is restored to a normal monitoring state in a timely manner, avoiding the system repeatedly identifying and processing the same error during subsequent monitoring processes due to historical error information left in the register, thereby causing unnecessary resource consumption and logical interference. By clearing this error information, it can be ensured that only the most recently generated anomaly is stored in the register, improving the real-time and accuracy of fault detection.
[0066] According to the electrical interconnection link parameter adjustment method proposed in an embodiment of the present application, first, the operating parameters of the processing unit are obtained; second, the completion timeout value of the corresponding slave device is adjusted based on the parameters; a first register is read to obtain request message completion timeout error information, and a second register is read to obtain the completion timeout value of each slave device; when the processing unit detects that the request message completion time exceeds the completion timeout value, a corresponding error message is generated; if this error message is detected, the relevant slave device is identified and at least one of the completion timeout value corresponding to the slave device and the operating parameters of the processing unit is adjusted. This solves the problems of low efficiency and fixed configuration methods in related technologies for electrical interconnection link parameter adjustment, avoids a large number of completion timeout error reports, effectively ensures the continuous and stable operation of the system, and achieves the technical effect of dynamically adjusting timeout values to ensure stable system operation.
[0067] The following will illustrate the implementation process of the method for adjusting the parameters of the electrical interconnection link through a specific implementation. Figure 5 As shown, the following steps are included: In step S501, when the storage component is in operation, the pressure on the processing unit is continuously increased to simulate a high-load scenario; In step S502, the system monitors the values in the first buffer of all PCIe add-in cards under the processing unit to check whether the completion timeout status bit is reported. If the detection result is no, the pressure of the processing unit at that time and the completion timeout value of the corresponding slave device are recorded and a corresponding relationship is established; if yes, the next step is processed; As shown in step S503 and step S504, when a completion timeout error is reported, the value of the request processing unit resource of the PCIe add-in card and the corresponding cache delay are obtained, and then the value cached in the second buffer of the abnormal PCIe add-in card is obtained, and the completion timeout value is queried; In step S505, if the current completion timeout value is the maximum value, it is determined whether the device service volume is the maximum value. If not, the system adjusts the parameters of the second register to increase the completion timeout value. If it is the maximum value, it proceeds to step S506. In step S506, the storage component obtains the business volume of the storage device and determines whether it is the minimum, that is, whether the pressure on the processing unit is the minimum. If not, the business volume is reduced to relieve the pressure on the processing unit. If it is the maximum, the process proceeds to step S507. In step S507, when the pressure of the processing unit has reached the minimum value and the system still reports an error, an alarm is reported and manual intervention is required.
[0068] The above embodiment describes the method for adjusting the parameters of the electrical interconnection link from the perspective of online application. Based on the above embodiment, the method for adjusting the parameters of the electrical interconnection link will be further described from the perspective of offline training. Figure 6 A flow chart of a method for adjusting parameters of an electrical interconnect link provided in an embodiment of the present application is provided. The method is applied to the above-mentioned electrical interconnect link parameter adjustment system. The two ends of the electrical interconnect link are respectively connected to a host device and at least one slave device. The host device includes a processing component and a storage component, wherein the storage component is configured to perform the following steps: In step S201 , the completion timeout value corresponding to the slave device under different operating parameters of the processing unit is obtained.
[0069] In step S202, a correspondence between operating parameters and completion timeout values is generated according to different operating parameters and completion timeout values corresponding to slave devices, and a training data set is generated according to the operating pressure value, the completion timeout value corresponding to the slave device and the correspondence.
[0070] In step S203, the neural network model is trained using the training data set, and the completion timeout value corresponding to the slave device is adjusted using the trained neural network model and the operating parameters of the processing unit.
[0071] For descriptions of features in the embodiment corresponding to the above-mentioned method for adjusting parameters of an electrical interconnection link, reference may be made to the descriptions of the embodiment corresponding to the system for adjusting parameters of an electrical interconnection link, which will not be described in detail here.
[0072] According to the electrical interconnection link parameter adjustment method proposed in the embodiment of the present application, first, the slave completion timeout value of the processing unit under different operating parameters is obtained; second, a correspondence is established between the operating parameters and the corresponding completion timeout value, and a training data set is generated in combination with the operating pressure value; finally, a neural network model is trained using the training data set, and the slave completion timeout value is adjusted based on the model output and the operating parameters. This solves the problems of low efficiency and fixed configuration of electrical interconnection link parameter adjustment in related technologies, avoids a large number of completion timeout errors, effectively ensures the continuous and stable operation of the system, and achieves the technical effect of dynamically adjusting the timeout value to ensure stable system operation.
[0073] The embodiment of the present application also provides an electronic device, such as Figure 7 As shown, the electronic device includes a memory 701 and a processor 702. The memory 701 stores a computer program, and the processor 702 is configured to run the computer program to execute the steps in the embodiment of the method for adjusting parameters of an electrical interconnection link.
[0074] An embodiment of the present application further provides a non-volatile computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned method for adjusting the parameters of the electrical interconnection link are implemented.
[0075] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media that can store computer programs, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk, or an optical disk.
[0076] An embodiment of the present application further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps in any of the above-mentioned electrical interconnection link parameter adjustment method embodiments are implemented.
[0077] An embodiment of the present application further provides another computer program product, including a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned embodiment of the method for adjusting the parameters of the electrical interconnection link are implemented.
[0078] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.
[0079] The above describes in detail the electrical interconnection link parameter adjustment system, server, and method provided by this application. This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only intended to help understand the method and core concept of this application. It should be noted that for ordinary technicians in this technical field, without departing from the principles of this application, several improvements and modifications can be made to this application, and these improvements and modifications also fall within the scope of protection of the claims of this application.
Claims
1. An electrical interconnection link parameter adjustment system, characterized in that: The two ends of the electrical interconnection link are respectively connected to a host device and at least one slave device, and the host device includes: a processing component, configured to process the request message of the slave device and return the message processing result to the slave device, wherein the processing component includes a processing unit, a first register, and a second register, wherein the first register stores error information of the request message completion timeout, and the second register stores a completion timeout value of each of the slave devices, and when the completion time of the request message processed by the processing unit exceeds the completion timeout value, generates error information of the request message completion timeout; A storage component is used to obtain the operating parameters of the processing unit, adjust the completion timeout value corresponding to the slave device according to the operating parameters of the processing unit, read the first register and the second register, and if the error information of the request message completion timeout is read, determine the corresponding slave device according to the error information of the request message completion timeout, and adjust at least one of the completion timeout value corresponding to the slave device and the operating parameters of the processing unit.
2. The electrical interconnection link parameter adjustment system according to claim 1, characterized in that: The storage component is provided with a pre-trained neural network model, the operating parameters of the processing unit are input into the neural network model, and the neural network model outputs the target value of the completion timeout value corresponding to the slave device, wherein the neural network model includes the correspondence between the operating parameters and the completion timeout value, and the target value of the completion timeout value corresponding to the operating parameters is determined based on the correspondence.
3. The electrical interconnection link parameter adjustment system according to claim 1, characterized in that: The host device further includes at least one expansion component, the expansion component is connected to the slave device, and the processing component and the expansion component are connected via the electrical interconnection link.
4. A server, characterized in that: The invention comprises the electrical interconnection link parameter adjustment system according to any one of claims 1 to 3.
5. A method for adjusting parameters of an electrical interconnection link, characterized in that: The method is applied to the electrical interconnection link parameter adjustment system according to any one of claims 1 to 3, wherein two ends of the electrical interconnection link are respectively connected to a host device and at least one slave device, the host device includes a processing component and a storage component, wherein the storage component is configured to perform the following steps: Obtaining operating parameters of the processing unit; adjusting a completion timeout value corresponding to the slave device according to an operating parameter of the processing unit; Reading the first register and the second register, wherein the first register stores error information of a request message completion timeout, and the second register stores a completion timeout value of each of the slave devices, and when the processing unit completes processing the request message for a time exceeding the completion timeout value, generating error information of the request message completion timeout; If error information of the request message completion timeout is read, the corresponding slave device is determined according to the error information of the request message completion timeout, and at least one of the completion timeout value corresponding to the slave device and the operating parameters of the processing unit is adjusted.
6. The method for adjusting parameters of an electrical interconnection link according to claim 5, wherein: The adjusting the completion timeout value corresponding to the slave device according to the operating parameters of the processing unit includes: Pre-trained neural network models; The operating parameters of the processing unit are input into the neural network model, and the neural network model outputs the target value of the completion timeout value corresponding to the slave device, wherein the neural network model includes a correspondence between the operating parameters and the completion timeout value, and the target value of the completion timeout value corresponding to the operating parameters is determined based on the correspondence.
7. The method for adjusting parameters of an electrical interconnection link according to claim 6, wherein: Before inputting the operating parameters of the processing unit into the neural network model, the method further includes: Obtain the completion timeout value corresponding to the slave device when the processing unit has different operating parameters; Generate a correspondence between the operating parameters and the completion timeout values corresponding to the slave devices according to different operating parameters, and generate a training data set according to the operating pressure value, the completion timeout value corresponding to the slave devices, and the correspondence; The neural network model is trained using the training data set.
8. The method for adjusting parameters of an electrical interconnection link according to claim 5, wherein: The adjusting at least one of a completion timeout value corresponding to the slave device and an operating parameter of the processing unit includes: Obtain the maximum value of the completion timeout value corresponding to the slave device; If the completion timeout value corresponding to the slave device is less than the maximum value, adjusting the completion timeout value corresponding to the slave device; If the completion timeout value corresponding to the slave device is the maximum value, the operating parameters of the processing unit are adjusted.
9. The method for adjusting parameters of an electrical interconnection link according to claim 8, wherein: The adjusting the completion timeout value corresponding to the slave device includes: Calculating the difference between the completion timeout value corresponding to the slave device and the maximum value; If the difference is smaller than a preset adjustment step, increasing the completion timeout value corresponding to the slave device to the maximum value; If the difference is greater than or equal to the preset adjustment step, the completion timeout value corresponding to the slave device is increased according to the adjustment step.
10. The method for adjusting parameters of an electrical interconnection link according to claim 8 or 9, characterized in that: After adjusting the completion timeout value corresponding to the slave device, the method further includes: Clear the error information of the request message completion timeout corresponding to the slave device from the first register.
11. The method for adjusting parameters of an electrical interconnection link according to claim 6, wherein: Before reading the error message indicating that the request message has timed out, the following steps are also included: Obtaining electrical interconnection protocol error information of the electrical interconnection link; parsing the error type in the electrical interconnection protocol error information; When determining, according to the error type, that the request message is timed out, the error information of the request message is read.
12. A method for adjusting parameters of an electrical interconnection link, characterized in that: The method is applied to the electrical interconnection link parameter adjustment system according to any one of claims 1 to 3, wherein two ends of the electrical interconnection link are respectively connected to a host device and at least one slave device, the host device includes a processing component and a storage component, wherein the storage component is configured to perform the following steps: Obtain the completion timeout value corresponding to the slave device when the processing unit has different operating parameters; Generate a correspondence between the operating parameters and the completion timeout values corresponding to the slave devices according to different operating parameters, and generate a training data set according to the operating pressure value, the completion timeout value corresponding to the slave devices, and the correspondence; The neural network model is trained using the training data set, and the completion timeout value corresponding to the slave device is adjusted using the trained neural network model and the operating parameters of the processing unit.
13. An electronic device, characterized in that: include: memory for storing computer programs; A processor, configured to implement the steps of the method for adjusting parameters of an electrical interconnection link according to any one of claims 5 to 12 when executing the computer program.
14. A non-volatile computer-readable storage medium, characterized in that: The non-volatile computer-readable storage medium stores a computer program, wherein the computer program, when executed by a processor, implements the steps of the method for adjusting parameters of an electrical interconnection link according to any one of claims 5 to 12.
15. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method for adjusting parameters of an electrical interconnection link according to any one of claims 5 to 12 are implemented.
Citation Information
Patent Citations
PCIe completion timeout error processing method and control unit
CN117992385A
Method, device and equipment for adjusting timeout time of gateway interface
CN118827375A
Financial terminal application and SP timeout protocol negotiation method, system, device and medium
CN119484488A
Memory management method and memory controller
CN120469645A
Adaptive timeout mechanism
US10592322B1