Processor topology switching system and method and server

Abstract

The invention discloses a processor topology switching system and method and a server, and relates to the technical field of computers, a user is allowed to simply input an operation instruction in a management client, and the management client can generate a management request according to the operation instruction and send the management request to a management controller; the management controller generates a topology switching configuration file according to the management request so as to control the topology switching controller, and switching of the logic connection relation between the processors is achieved. In the embodiment of the invention, the target logic connection relationship in the topology switching configuration file is generated by the management controller, a user only needs to simply input the operation instruction, the operation complexity is relatively low, the cost is relatively low, and the switching error rate can be effectively reduced.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a processor topology switching system, method, and server. Background Art

[0002] With the rapid development of technologies like artificial intelligence and big data, server performance requirements are increasing. Configuring multiple GPUs (Graphics Processing Units) on a server can improve server performance. The logical connections between multiple GPUs form a server's GPU topology, and these topologies can be switched between.

[0003] In related technologies, the GPU topology mode of a server is implemented through hardware-level jumpers or BIOS (Basic Input / Output System) configuration. This approach suffers from poor flexibility, complex operation, switching errors, and high costs. Summary of the Invention

[0004] The present application provides a processor topology switching system, method, and server to at least solve the problems of poor flexibility, complex operation, and high cost in related technologies.

[0005] In a first aspect, the present application provides a processor topology switching system, which includes: a management controller and a topology switching controller, the management controller includes: a management request parsing unit and a switching profile generating unit, the switching profile generating unit is connected to the management request parsing unit and the topology switching controller; the management request parsing unit is used to: receive a management request to parse the request type of the management request, and when the request type corresponds to a topology switching request, control the switching profile generating unit to generate a topology switching profile corresponding to the management request, the management request is generated according to the operation instructions entered by the user on the management client, and the topology switching profile includes the target logical connection relationship between the processors on the server; the switching profile generating unit is used to: generate a topology switching profile corresponding to the management request, and send the topology switching profile to the topology switching controller; the topology switching controller is used to: switch the logical connection relationship between the processors according to the topology switching profile.

[0006] In a second aspect, the present application provides a processor topology switching method, which is applied to a processor topology switching system. The processor topology switching system includes: a management controller and a topology switching controller. The management controller includes: a management request parsing unit and a switching configuration file generation unit. The switching configuration file generation unit is connected to the management request parsing unit and the topology switching controller. The above-mentioned processor topology switching method includes:

[0007] The management request parsing unit receives a management request to parse the request type of the management request, which is generated based on the operation instructions entered by the user on the management client; when the request type corresponds to a topology switching request, the management request parsing unit controls the switching profile generation unit to generate a topology switching profile corresponding to the management request, and the topology switching profile includes the target logical connection relationship between the processors on the server; the switching profile generation unit generates a topology switching profile corresponding to the management request and sends the topology switching profile to the topology switching controller; the topology switching controller switches the logical connection relationship between the processors according to the topology switching profile.

[0008] In a third aspect, the present application provides a server comprising the aforementioned processor topology switching system.

[0009] Through this application, since users can simply enter an operation instruction in the management client, the management client can generate a management request based on the operation instruction and send it to the management controller. The management controller generates a topology switching configuration file based on the management request to control the topology switching controller, thereby switching the logical connection relationship between processors. In an embodiment of the present application, the target logical connection relationship in the topology switching configuration file is generated by the management controller, and the user only needs to enter a simple operation instruction. This reduces the complexity of the operation, reduces the cost, and effectively reduces the switching error rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] 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.

[0011] Figure 1 This is a schematic diagram of a processor topology structure of a server provided in an embodiment of the present application;

[0012] Figure 2 This is a schematic diagram of a processor topology structure of another server provided in an embodiment of the present application;

[0013] Figure 3 A schematic diagram of the structure of a processor topology switching system provided in an embodiment of the present application;

[0014] Figure 4 A schematic diagram of the structure of another processor topology switching system provided in an embodiment of the present application;

[0015] Figure 5 A flowchart of the steps of a processor topology switching method provided in an embodiment of the present application. DETAILED DESCRIPTION

[0016] 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.

[0017] 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.

[0018] The embodiments of the present application are intended to perform topological switching on one or more processors in a server. The server of the present application is generally suitable for application scenarios with relatively high data processing performance requirements, such as deep learning training, reasoning, scientific computing, and other application scenarios. In these application scenarios, the server is generally provided with multiple processors, and the processors here may include: GPU and CPU (Central Processing Unit). The logical connection relationship between the processors is used to indicate the GPU assigned to the CPU, and the CPU and the GPU assigned to the CPU can collaborate to process data.

[0019] Figure 1 This is a schematic diagram of a processor topology structure of a server provided in an embodiment of the present application, with reference to Figure 1 As shown, the server includes two central processing units (CPU1 and CPU2) and four graphics processing units (GPU1 to GPU4). When GPU1 and GPU2 are logically connected to CPU1, and GPU3 and GPU4 are logically connected to CPU2, CPU1, GPU1, and GPU2 collaborate to process data, and CPU2, GPU3, and GPU4 collaborate to process data.

[0020] It should be noted that in actual applications, the number of processors and topology of the server are not limited to Figure 1 shown.

[0021] The processor topology of the above servers can be switched dynamically and flexibly according to actual needs. Figure 1 The processor topology shown is adjusted to Figure 2The processor topology shown is as follows: GPU1 and GPU2, GPU3 are logically connected to CPU1, and GPU4 is logically connected to CPU2. At this time, CPU1, GPU1 and GPU2, GPU3 collaborate to process data, and CPU2 and GPU4 collaborate to process data.

[0022] Related technologies typically implement topology switching through hardware jumper settings or BIOS configuration. However, this approach requires manual configuration by professionals, is complex and costly, and is prone to topology switching errors. These errors can cause the GPU to malfunction or even be damaged, seriously impacting server service quality.

[0023] To address the aforementioned technical issues, an embodiment of the present application allows a user to simply enter an operation instruction in a management client. The management client can then generate a management request based on the operation instruction and send it to a management controller. The management controller then generates a topology switching configuration file based on the management request to control the topology switching controller, thereby switching the logical connection relationships between processors. In this embodiment of the present application, the target logical connection relationship in the topology switching configuration file is generated by the management controller, and the user only needs to enter the operation instruction. This reduces operational complexity and cost, and can effectively reduce the switching error rate.

[0024] Furthermore, the management controller of this application can be a BMC (Baseboard Management Controller), which runs independently of the server's operating system and can monitor the server's hardware status, control the server's power supply, and access the server's operating system logs through out-of-band management. Therefore, this application uses the management controller to implement switching, allowing the server's operating system to continue running, requiring only a restart after the switch. This avoids the time required to stop the server's operating system for the switch, effectively reducing the duration of service interruptions caused by processor topology switching.

[0025] Again, the management controller and switch controller of the present application are commonly used hardware for servers, and therefore, will not lead to an increase in hardware costs.

[0026] In order to enable those skilled in the art to better understand the present application, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0027] Figure 3 A schematic diagram of a processor topology switching system provided in an embodiment of the present application is shown in FIG. Figure 3As shown, the processor topology switching system 100 provided in an embodiment of the present application includes: a management controller 101 and a topology switching controller 102 , which are interconnected so that the management controller 101 implements processor topology structure switching through the topology switching controller 102 .

[0028] The management controller 101 includes a management request parsing unit 1011 and a handover profile generating unit 1012. The handover profile generating unit 1012 is connected to the management request parsing unit 1011 and the topology handover controller 102. Thus, the connection between the handover profile generating unit 1012 and the topology handover controller 102 realizes the connection between the management controller 101 and the topology handover controller 102.

[0029] The management request parsing unit 1011 is used to parse the management request and forward the management request to the business processing unit based on the parsing result. In the processor topology switching scenario, the management request parsing unit 1011 is used to: receive a management request, parse the request type of the management request, and when the request type corresponds to a topology switching request, control the switching profile generation unit 1012 to generate a topology switching profile corresponding to the management request. The management request is generated based on an operation instruction input by a user on the management client 103. The management client 103 and the management controller 101 are in communication connection. The management request is a request sent by the management client 103 to the management controller 101. The topology switching profile includes the target logical connection relationship between the processors on the server.

[0030] The request type may be represented by one or more flag bits in the header information of the management request. The number of flag bits is associated with the number of request types supported by the management controller 101. When the number of request types is M and the number of flag bits is N, the two satisfy the following relationship: For example, when the number of flag bits is 1, M=2, indicating that the management controller 101 supports two types of management requests, wherein flag bit 1 indicates that the management request is a topology switching request to implement processor topology switching.

[0031] Reference Figure 4 As shown, the management client 103 is connected to the management request parsing unit 1011. The parsing process of the above request type includes: when receiving the management request sent by the management client, the management request parsing unit 1011 extracts the above flag bit from the specified bit of the management request header information, and determines whether the request type is a topology switching request based on the specific value of the flag bit. If it is a topology switching request, the management request parsing unit 1011 sends the management request to the switching profile generation unit 1012.

[0032] After receiving the management request, the switching profile generating unit 1012 is configured to generate a topology switching profile corresponding to the management request and send the topology switching profile to the topology switching controller 102 . The topology switching profile includes the target logical connection relationship between the processors on the server.

[0033] Specifically, the topology switching configuration file may be extracted from the topology switching configuration file, or may be dynamically generated according to the topology mode carried in the management request. The embodiment of the present application does not limit the specific generation strategy of the topology switching configuration file.

[0034] Of course, the handover profile generation unit 1012 can generate a topology handover profile based on the priorities of various topology handover profile generation strategies. For example, the topology handover profile can be preferentially extracted from the management request. If no topology handover profile exists in the management request, the topology handover profile can be dynamically generated based on the topology model in the management request. In this way, the target logical connection relationship set by the user can be prioritized, ensuring that the post-handover topology accurately meets the user's needs. If the user does not set a target logical connection relationship, a dynamic generation can be further performed, which helps to improve the handover success rate.

[0035] Upon receiving the topology switching configuration file sent by the switching configuration file generation unit 1012, the topology switching controller 102 is configured to switch the logical connection relationship between the processors according to the topology switching configuration file. Specifically, the topology switching controller 102 may replace the configuration file in the target path with the received topology switching configuration file, and load the topology switching configuration file to make the topology switching configuration file effective. When executing a computing task, the processors can collaboratively process data according to the target logical connection relationship in the topology switching configuration file.

[0036] In summary, the embodiment of the present application allows the user to simply input an operation instruction in the management client 103. The management client 103 can then generate a management request based on the operation instruction input by the user and send it to the management request parsing unit 1011 of the management controller 101 for parsing. When the parsed request type indicates that the management request is a topology switching request, the management request parsing unit 1011 controls the switching profile generation unit 1012 to generate a topology switching profile based on the management request, so that the topology switching controller 102 implements the switching of the logical connection relationship between processors based on the topology switching profile. It can be seen that the target logical connection relationship in the topology switching profile is generated by the switching profile generation unit 1012. The user only needs to input a simple operation instruction and does not need to input complex content. Therefore, the operation complexity is low, the cost is low, and the switching error rate can be effectively reduced.

[0037] Of course, the embodiment of the present application also supports the user inputting a topology switching configuration file, and the management controller 101 can add it to the management request, so as to accurately meet the user's requirements for the processor topology structure.

[0038] In some possible implementations, the data format of the topology switching request is: vendor function number, request type, request data, and response data, wherein the vendor function number is used to indicate the developer of the management request, and each developer has one or more vendor function numbers.

[0039] Request data is used to indicate the relevant data required by the request.

[0040] The response data may include but is not limited to: handover success, handover failure, and the reason for handover failure.

[0041] In some possible implementations, Figure 4 This is a schematic diagram of another processor topology switching system provided in an embodiment of the present application, referring to Figure 4 As shown, the management controller 101 further includes a storage unit 1013, to which the handover profile generation unit 1012 is connected. The storage unit 1013 is used to store any information required by the management controller 101 to manage the server. The storage unit 1013 can be integrated into the management controller 101 or provided externally to the management controller 101 and connected to the handover profile generation unit 1012 in the management controller 101.

[0042] The management request parsing unit 1011 is further configured to parse a target topology pattern from the management request and send the target topology pattern to the handover configuration file generating unit 1012 .

[0043] The target topology mode may be generated by the management client 103 based on an operation instruction entered by the user in the management client 103. The user may enter the target topology mode in the management client 103. The management client 103 may provide a UI (User Interface) interactive interface to facilitate user input of operation instructions, or may only provide an instruction input interface.

[0044] In the UI interactive interface, users can implement control through the operation of UI controls. Specifically, a mode selection box can be displayed in the management client 103, and the user can select a topology mode as the target topology mode from the list of options in the mode selection box. Of course, the current topology mode can also be displayed next to the mode selection box to prompt the user to select a topology mode other than the current topology mode as the target topology mode. The management client 103 can also distinguish and display the current topology mode in the list of options in the mode selection box to facilitate users to distinguish the current topology mode according to different display methods. In this way, the user can select the remaining topology modes other than the current topology mode in the list of options as the target topology mode. It can be seen that the embodiment of the present application can assist the user in selecting a mode through the display of the above-mentioned current topology mode, thereby improving the effectiveness of topology switching and avoiding switching caused by invalid user operations.

[0045] In the instruction input interface, users can enter machine-recognizable instructions, which are usually composed of a series of characters, including but not limited to: letters and special characters.

[0046] The management client 103 can generate a management request based on the target topology mode input by the user. The target topology mode can be represented by a mode flag located in the first bit of the management request. Therefore, upon receiving the management request, the management request parsing unit 1011 can extract the mode flag from the first bit of the management request to determine the target topology mode and send it to the handover profile generation unit 1012.

[0047] Accordingly, the switching configuration file generating unit 1012 is further configured to:

[0048] A topology switching profile corresponding to the target topology pattern is determined from one or more configuration files stored in the storage unit 1013. Specifically, each configuration file in the storage unit 1013 corresponds to its own topology pattern. Therefore, the switching profile generation unit 1012 can match the target topology pattern with the topology patterns of each configuration file in the storage unit 1013, and use the configuration file whose topology pattern matches the target topology pattern as the topology switching profile.

[0049] When the topology switching configuration file does not exist in one or more configuration files in the storage unit 1013, the topology switching configuration file is dynamically generated according to the target topology mode. At this time, all processors can also be obtained to determine the target logical connection relationship between all processors according to the target topology mode, so as to generate the topology switching configuration file according to the target logical connection relationship. For example, when the target topology mode is the balance mode (Balance), all graphics processors can be evenly distributed to each central processor so that the number of graphics processors corresponding to each central processor is the same. This topology mode can make the pressure of each central processor more balanced, which is suitable for distributed training and multi-tasking scenarios.

[0050] It can be seen that the embodiment of the present application can generate a topology switching profile based on two strategies through the target topology mode. The user only needs to enter the topology mode to achieve switching, which can reduce the complexity of the operation instructions entered by the user. In addition, the generation strategies of the two topology switching profiles help to improve the switching success rate.

[0051] In another example of an embodiment of the present application, the management request parsing unit 1011 is also used to: parse a topology switching configuration file from a management request, and the topology switching configuration file is generated by the management client 103 according to the operation instructions input by the user. Accordingly, the switching configuration file generation unit 1012 is also used to: upon receiving the topology switching configuration file sent by the management request parsing unit 1011, send the topology switching configuration file to the topology switching controller 102; upon not receiving the topology switching configuration file sent by the management request parsing unit 1011, determine the topology switching configuration file according to the target topology mode. The embodiment of the present application uses the topology switching configuration file carried in the management request or dynamically generates the topology switching configuration file according to the priority, which can not only meet the user's switching needs as accurately as possible, but also avoid the topology switching failure caused by the failure to generate the topology switching configuration file.

[0052] The topology switching configuration file may be input by a user in the management client 103, or may be generated by the management client 103 according to an operation instruction input by the user. The process of the switching configuration file generation unit 1012 determining the topology switching configuration file according to the target topology mode includes the aforementioned steps: determining the topology switching configuration file corresponding to the target topology mode from one or more configuration files stored in the storage unit 1013; and dynamically generating the topology switching configuration file according to the target topology mode when the topology switching configuration file does not exist in the one or more configuration files in the storage unit 1013.

[0053] In one possible implementation, the switching profile generation unit 1012 is further configured to generate a topology switching profile in the following manner: first, calculating the number of graphics processors corresponding to each central processor based on the target topology pattern; then, establishing a target logical connection relationship between the central processor and the graphics processor based on the number of graphics processors corresponding to each central processor and the central processor's demand for graphics processors; and finally, generating a topology switching profile based on the target logical connection relationship. In this way, the present application, while allocating graphics processors according to the target topology pattern, satisfies the central processor's demand for graphics processors as much as possible, thereby helping to improve the processing performance of the central processor.

[0054] The number of GPUs associated with each CPU is associated with the target topology mode. For example, when the target topology mode is balanced, the number of GPUs associated with each CPU is the same. Therefore, the number of GPUs associated with each CPU is the rounded-up ratio of the total number of GPUs on the server to the total number of CPUs. This rounded-up ratio can be rounded up, down, or rounded to the nearest integer. When the target topology mode is cascade, all GPUs can be assigned to one CPU.

[0055] After obtaining the number of graphics processors for each central processing unit (CPU), the CPU can also be combined with the CPU's demand information for the GPU to allocate the GPU. This information includes the CPU's memory requirements, floating-point computing performance requirements, and parallel computing requirements for the GPU. For example, the memory requirement can be greater than or equal to a preset memory threshold, the floating-point computing performance requirement can be greater than or equal to a preset computing threshold for the number of floating-point operations that can be executed per unit time, and the parallel computing requirement can be greater than or equal to a preset data threshold for the number of parallel data that can be executed per unit time.

[0056] The allocation of graphics processors based on the CPU's demand information for graphics processors is specifically embodied as follows: for each CPU, graphics processors that meet its demand information are screened as candidate graphics processors for the CPU, and graphics processors with the required number of graphics processors are screened from the candidate graphics processors as the graphics processors for the CPU, so as to establish a target logical connection relationship between the two.

[0057] In one possible implementation, the CPU can be prioritized to assign a corresponding graphics processor to each CPU in descending order of priority. CPUs with higher priorities are given priority in assigning graphics processors that meet their needs. The CPU's priority is positively correlated with the CPU's workload and task priority. The greater the CPU's workload and the higher the task priority, the higher the CPU's priority. Conversely, the lower the CPU's workload and the lower the task priority, the lower the CPU's priority. This allows the needs of CPUs with higher workloads and more important tasks to be met first, ensuring a high success rate for their task execution.

[0058] In one embodiment of the present application, referring to Figure 4 As shown, the management controller 101 also includes: a feasibility verification unit 1014, the feasibility verification unit 1014 is connected to the storage unit 1013 and the switching profile generation unit 1012, and the feasibility verification unit 1014 is used to: obtain preset topology switching information from the storage unit 1013, the preset topology switching information includes the switchable relationship between one or more topology modes; perform feasibility verification on the management request according to the preset topology switching information to obtain a feasibility verification result; when the feasibility verification result is that the feasibility verification is passed, control the switching profile generation unit 1012 to determine the topology switching profile according to the management request. In this way, the switching between topology modes can be restricted by the preset topology switching information to determine whether it is feasible in advance, thereby avoiding topology switching failure, or poor topology switching results, and abnormal topology structure after switching, which leads to abnormal server operation.

[0059] In one example, the preset topology switching information includes multiple directed topology mode pairs, each of which includes a first topology mode and a second topology mode, indicating that switching from the first topology mode to the second topology mode is supported. If the preset topology switching information does not include a directed topology mode pair from the second topology mode to the first topology mode, then switching from the second topology mode to the first topology mode is not supported.

[0060] When implementing feasibility verification based on the above-mentioned preset topology switching information, the feasibility verification unit 1014 needs to obtain the current topology mode. The current topology mode can be carried in the management request, or can be obtained by the feasibility verification unit 1014 from the storage unit 1013. Therefore, the current topology mode and the target topology mode can be constructed as a pattern pair to be verified to determine whether the pattern pair to be verified exists in the preset topology switching information. If so, the feasibility verification result is feasibility verification passed, and a topology switching configuration file can be generated according to the management request. If not, the feasibility verification result is feasibility verification failed, and a topology switching configuration file can be not generated according to the management request, and this topology switching fails.

[0061] When the current topology mode is stored in the storage unit 1013, the amount of data required for management can be reduced. Figure 4 As shown, the feasibility verification unit 1014 is also connected to the management request parsing unit 1011. The feasibility verification unit 1014 is also used to:

[0062] Obtain the current topology mode from the storage unit 1013 and the target topology mode sent by the receiving management request parsing unit 1011, and determine whether the current topology mode can be switched to the target topology mode based on the preset topology switching information; when the current topology mode can be switched to the target topology mode, determine that the management request verification is successful; when the current topology mode cannot be switched to the target topology mode, determine that the management request verification has failed.

[0063] The current topology mode is written into the storage unit 1013 by the switching configuration file generating unit 1012 after the last topology switching is successful. The switching configuration file generating unit 1012 is connected to the storage unit 1013 .

[0064] In other embodiments, reference Figure 4As shown, the management controller 101 further includes: a switching mode verification unit 1015, the switching mode verification unit 1015 is connected to the management request parsing unit 1011 and the switching configuration file generation unit 1012, and the switching mode verification unit 1015 is used to: receive the topology switching configuration file sent by the management request parsing unit 1011, and parse the topology switching configuration file to obtain an actual topology mode; match the actual topology mode with the target topology mode to obtain a matching result; when the matching result indicates that the actual topology mode and the target topology mode match, control the switching configuration file generation unit 1012 to send the topology switching configuration file to the topology switching controller 102; when the matching result indicates that the actual topology mode and the target topology mode do not match, the switching configuration file generation unit 1012 will not send the topology switching configuration file to the topology switching controller 102. In the embodiment of the present application, when both the target topology mode and the topology switching configuration file are present in the management request, it is determined whether the two are consistent, so as to avoid performing topology switching when the management request is erroneous, which helps to improve the accuracy of topology switching.

[0065] In some optional implementations, when the switching mode verification unit 1015 determines that the actual topology mode and the target topology mode do not match, a request error indication message may be sent to the management client 103. Upon receiving the request error indication message, the management client 103 may display the request error indication message to prompt the user to modify the request error indication message. In this way, the success rate of topology switching can be improved.

[0066] In some embodiments of the present application, reference is made to Figure 4 As shown, the management controller 101 further includes a system restart unit 1016 connected to the topology switching controller 102. Upon receiving a switch success message from the topology switching controller 102, the system restart unit 1016 is configured to determine that the topology switch is successful and restart the server's operating system. If no switch success message is received from the topology switching controller 102 within a preset time period, the system restart unit 1016 determines that the topology switch has failed and does not restart the server's operating system. In this manner, the system can be accurately restarted based on the switch success message to implement the target logical connection relationship in the topology switching configuration file.

[0067] Specifically, when the handover profile generation unit 1012 sends the topology handover profile to the topology handover controller 102, it also sends a timer start message to the system restart unit 1016. The system restart unit 1016 starts the timer, and the timer timeout duration is the preset timeout duration. If the timer has not timed out and the system restart unit 1016 receives a handover success message, the system restart unit 1016 restarts the server's operating system. If the timer has not timed out and the handover success message has not been received, the system restart unit 1016 does not restart the server's operating system.

[0068] It is understood that the switching profile generation unit 1012 can also estimate a preset duration based on the data volume of the topology switching profile. The preset duration is positively correlated with the number of topology switching profiles. The switching profile generation unit 1012 sends the preset duration and the timer start message to the system restart unit 1016. In this way, the timeout duration is accurately estimated to set an appropriate timeout duration, which helps save time.

[0069] In order to further accurately implement the operating system restart, refer to Figure 4 As shown, the system restart unit 1016 is also connected to the feasibility verification unit 1014 and the switching mode verification unit 1015. The system restart unit 1016 is also used to: determine the topology switching result of the processor according to the feasibility verification result, the switching result message sent by the topology switching controller 102, or at least one of the matching results; when the topology switching result indicates that the switching is successful, restart the operating system. The successful switching includes: the feasibility verification result is that the feasibility verification passes, the matching result indicates that the actual topology mode and the target topology mode match, and the switching result message is a switching success message. When the feasibility verification result fails or the matching result is mismatched, at this time, there is no need to wait for the switching result message to determine that the topology switching result is a switching failure. The switching result can be determined as soon as possible to proceed to the next step of processing, saving time.

[0070] In some possible implementations, refer to Figure 4 As shown, the processor topology switching system 100 further includes a switching result feedback unit 1017 connected to the feasibility verification unit 1014, the switching mode verification unit 1015 and the management client 103. The switching result feedback unit 1017 is used to:

[0071] When the topology switching result indicates a switching failure, a switching failure message is generated and sent to the management client 103 for display. The switching failure includes at least one of the following: the feasibility verification result indicates a feasibility verification failure, the matching result indicates that the actual topology mode and the target topology mode do not match, and the switching result message is a switching failure message. In this way, when the feasibility verification fails or the actual topology mode and the target topology mode do not match, the switching failure is determined to be a failure without waiting for the switching result message. In this way, the switching failure message can be fed back as soon as possible to prompt the user.

[0072] Of course, the handover failure message may also include the handover failure reason, which includes: feasibility verification result failure, mismatch between the actual topology mode and the target topology mode, and handover failure by the topology handover controller 102. In this way, the user can adjust the operation instructions according to the handover reason to improve the success rate of the next topology handover.

[0073] Furthermore, in some possible implementations, referring to Figure 4 As shown, the handover result feedback unit 1017 is further configured to: determine the handover failure cause, and generate handover adjustment indication information based on the handover failure cause; and add the handover failure indication information and the handover adjustment indication information to the handover failure message, where the handover failure indication information is used to indicate the handover failure. In this way, the present application intelligently generates handover adjustment indication information to instruct the user to adjust the operation instruction based on the handover adjustment indication information. The user does not need to analyze the handover failure cause to determine the adjustment strategy for the operation instruction, which can further reduce the complexity of user operations and lower the user's operation threshold.

[0074] It should be noted that the above-mentioned handover failure indication information can be represented by a flag bit, for example, 0 is used to indicate handover failure. Of course, in the handover success message, the handover success indication information is represented by flag bit 1. The handover adjustment indication information is used to adjust the target topology mode or one or more target logical connection relationships in the topology handover configuration file.

[0075] Corresponding to the switching result feedback unit 1017 sending the switching adjustment instruction information to the management client 103, the management client 103 is further configured to: display the switching adjustment instruction information and update the management request based on the switching adjustment instruction information. Upon receiving a user confirmation operation on the management request, the management request is sent to the management request parsing unit 1011, so that the management request parsing unit 1011 parses the management request, the configuration unit generates a topology switching configuration file, and the topology switching controller 102 performs topology switching based on the topology switching configuration file. In this way, the management client 103 can automatically update the management request, further reducing the user's operation of updating the management request, simplifying the complexity of the user operation, and improving the user's operation efficiency.

[0076] Of course, in actual applications, there may be multiple switching adjustment indication information, so multiple updated management requests may be generated. The management client 103 can display multiple updated management requests at the same time to allow the user to select one as the management request to be executed.

[0077] In one embodiment of the present application, referring to Figure 4 As shown, the handover result feedback unit 1017 is further connected to the storage unit 1013. The handover result feedback unit 1017 is configured to: when the topology handover result indicates a handover failure, determine a handover failure level according to the handover failure cause, and record the handover failure cause in a log file in the storage unit 1013 according to the handover failure level. In this way, the handover failure cause can be recorded in the log file to facilitate subsequent analysis and tracing of the handover failure cause, avoid subsequent repeated handover failures due to the same cause, and help improve the handover success rate.

[0078] The handover failure level is used to describe the impact of the handover failure cause. The greater the impact, the higher the handover failure level, and the smaller the impact, the lower the handover failure level. One or more handover failure causes correspond to one handover failure level.

[0079] In one example, a log file can be set up for each switching failure level to store the switching failure records corresponding to the switching failure level. The switching failure records include not only the switching failure cause, but also the switching failure time and management request. In this way, the number of switching failure records in the log file corresponding to each switching failure level can be periodically counted, and the number of switching failure records for each switching failure level can be statistically analyzed. When the number of switching failure records for a higher switching failure level is high, it is necessary to prompt further analysis of the switching failure level to reduce the frequency of occurrence of the switching failure level. When the number of switching failure records for a higher switching failure level is low, further analysis of the switching failure level may not be performed.

[0080] In some possible implementations, refer to Figure 4 As shown, management controller 101 further includes a resource management control unit 1018, which is connected to topology switching controller 102 and switching profile generation unit 1012. Upon receiving a switching success message from topology switching controller 102, resource management control unit 1018 is configured to allocate target resources required for CPU operation to the server's CPU based on the topology switching profile sent by switching profile generation unit 1012. In this way, the target resources required for each CPU operation are associated with the corresponding graphics processor, ensuring sufficient target resources for the CPU without wasting them.

[0081] In some possible implementations, the resource management control unit 1018 is further used to: extract the number of graphics processors corresponding to each central processor from the topology switching configuration file; adjust resource parameters according to the number of graphics processors corresponding to the central processor, the resource parameters including power supply parameters and / or heat dissipation parameters, the power supply parameters and heat dissipation parameters being positively correlated with the number of graphics processors; adjust target resources of the central processor according to the resource parameters, the target resources including power supply resources and / or heat dissipation resources.

[0082] The power supply parameters above indicate the ability to power the CPU, including the power supply. The heat dissipation parameters indicate the cooling capacity provided to the CPU, including the number of heat sinks and fans. The power supply parameters are positively correlated with the number of GPUs in the CPU, and the heat dissipation parameters are positively correlated with the number of GPUs in the CPU. This ensures proper power supply and heat dissipation for the CPU while conserving resources.

[0083] In one embodiment of the present application, referring to Figure 4 As shown, the management controller 101 also includes: a server performance detection unit 1019, the server performance detection unit 1019 is connected to the management client 103 and the topology switching controller 102, and the server performance detection unit 1019 is used to: upon receiving the switching success message sent by the topology switching controller 102, detect the server performance to obtain performance parameters, and send the performance parameters to the management client 103. In this way, after the topology switching is successful, the server performance can be detected in real time to facilitate the user to decide whether to continue using the current processor topology based on the server performance. When the server performance is poor, the user can continue to switch the processor topology, and when the server performance is good, the current processor topology can be continued to be used. In this way, the server performance can be guaranteed as much as possible to avoid the degradation of service quality caused by topology switching.

[0084] Of course, when the client receives the performance parameters of multiple topology modes, it can also analyze the performance parameters of each topology mode to obtain the topology mode with the highest performance parameters as the recommended topology mode. Then, the next time the user switches the processor topology, the recommended topology mode can be displayed, allowing the user to directly select the recommended topology mode.

[0085] The topology modes may also be arranged in descending order according to their performance parameters, so that the user may preferentially select the topology mode with the highest performance parameter.

[0086] In some possible implementations, refer to Figure 4As shown, the management controller 101 further includes a link configuration unit 1010, which is connected to a management request parsing unit 1011. The management request parsing unit 1011 is further configured to send the management request to the link configuration unit 1010 when the request type corresponds to a link configuration request. The link configuration request can also be generated by an operation instruction input by a user in the management client 103. The user can input the request type and request content corresponding to the link configuration request in the management client 103 to generate the link configuration request. For example, when the management client 103 provides a UI interface, the user can select the link configuration request from the list of available options in the request type selection box in the UI interface. In this case, the management client 103 displays a configuration interface for the link configuration request, allowing the user to input the request content of the link configuration request in the configuration interface.

[0087] The link configuration unit 1010 is used to configure the link between the handover profile generation unit 1012 and the topology handover controller 102. The link configuration request carries the link's validity period and transmission performance parameters. This application can also ensure the success rate of subsequent topology handover profile transmission by allowing users to configure the link.

[0088] The above-mentioned valid time is used to indicate that the switching profile generating unit 1012 needs to transmit the topology switching profile to the topology switching controller 102 within the valid time.

[0089] The transmission performance parameter may generally be a transmission speed, which is used to instruct the handover profile generating unit 1012 to transmit data to the topology handover controller 102 according to the transmission speed.

[0090] In one possible implementation, referring to Figure 4 As shown, the management client 103 also includes: a link configuration estimation unit 1031; the link configuration estimation unit 1031 is used to: estimate the effective time and transmission performance parameters based on the number of processors on the server for display, and upon receiving the user's confirmation operation on the effective time and transmission performance parameters, generate a link configuration request based on the effective time and transmission performance parameters.

[0091] Specifically, the management client 103 can also estimate the data volume of the topology switching configuration file based on the number of processors in the server. The data volume of the topology switching configuration file is positively correlated with the number of processors. Based on the data volume of the topology switching configuration file, the effective time and transmission performance parameters are estimated for display. When the user enters a confirmation operation for the effective time and transmission performance parameters, a link configuration request can be generated based on the effective time and transmission performance parameters and sent to the management request parsing unit 1011.

[0092] The effective time is positively correlated with the data volume of the topology switching configuration file, and the transmission performance parameter is negatively correlated with the data volume of the topology switching configuration file.

[0093] In practical applications, a user-entered validity period can be first received to indicate that the user expects the topology switching profile to be transmitted between the switching profile generation unit 1012 and the topology switching controller 102 within the validity period. The ratio of the data volume of the topology switching profile to the validity period is then used as a transmission performance parameter, that is, the amount of data transmitted per unit time. This approach can ensure that the topology switching profile is transmitted within a certain period of time.

[0094] Alternatively, a user-entered transmission performance parameter can be received to indicate the user's desired transmission performance. The validity period is then calculated as the ratio of the topology switching profile's data volume to the transmission performance. This approach can maximize the transmission speed of the topology switching profile.

[0095] In addition, the user may also make adjustments based on the effective time and transmission performance parameters estimated by the management client 103 , and generate a link configuration request using the adjusted effective time and transmission performance parameters.

[0096] Figure 5 This is a flowchart of the steps of a processor topology switching method provided by an embodiment of the present application, which is applied to the aforementioned processor topology switching system 100. The processor topology switching system 100 includes: a management controller 101 and a topology switching controller 102. The management controller 101 includes: a management request parsing unit 1011 and a switching profile generating unit 1012. The switching profile generating unit 1012 is connected to the management request parsing unit 1011 and the topology switching controller 102. The connection between the switching profile generating unit 1012 and the topology switching controller 102 can be implemented by a UART (Universal Asynchronous Receiver / Transmitter). Among them, a transceiver is provided in the switching profile generating unit 1012, and a transceiver is provided in the topology switching controller 102. The two communicate through the two transceivers.

[0097] Based on the above-mentioned processor topology switching system 100, the processor topology switching method of the present application includes but is not limited to the following steps:

[0098] S201 : The management request parsing unit 1011 receives a management request to parse the request type of the management request. The management request is generated according to an operation instruction input by a user on the management client 103 .

[0099] Specifically, the user inputs an operation instruction on the management client 103 , and the management client 103 generates a management request according to the operation instruction and sends the management request to the management request parsing unit 1011 .

[0100] In the embodiment of the present application, the management request is any management request that requires the management controller 101 to manage the server, including but not limited to a topology switching request and a link configuration request.

[0101] The topology switching request may include a target topology mode and / or a topology switching configuration file. The topology switching configuration file may be a file imported by a user in the UI interaction interface of the management client 103 .

[0102] S202 : When the request type corresponds to a topology switching request, the management request parsing unit 1011 controls the switching profile generating unit 1012 to generate a topology switching profile corresponding to the management request. The topology switching profile includes a target logical connection relationship between processors on the server.

[0103] When controlling the handover profile generation unit 1012 to generate a topology handover profile corresponding to a management request, the management request parsing unit 1011 needs to send different information to the topology handover profile generation unit 1012. Specifically, the management request parsing unit 1011 may send the target topology mode parsed from the management request to the handover profile generation unit 1012, so that the handover profile generation unit 1012 generates the topology handover profile based on the target topology mode. The management request parsing unit 1011 may also send the topology handover profile parsed from the management request to the handover profile generation unit 1012. Of course, the management request parsing unit 1011 may also send the topology handover profile and the target topology mode to the handover profile generation unit 1012 at the same time.

[0104] S203 : The switching profile generating unit 1012 generates a topology switching profile corresponding to the management request, and sends the topology switching profile to the topology switching controller 102 .

[0105] When receiving the target topology mode, the switching profile generation unit 1012 can select the topology switching profile corresponding to the target topology mode from the profiles corresponding to one or more topology modes; or, the switching profile generation unit 1012 sends the received topology switching profile to the topology switching controller 102.

[0106] S204 : The topology switching controller 102 switches the logical connection relationship between the processors according to the topology switching configuration file.

[0107] In some possible implementations, the management controller 101 also includes: a storage unit 1013, and the switching profile generation unit 1012 is connected to the storage unit 1013; the above-mentioned processor topology switching method also includes: the management request parsing unit 1011 parses the target topology mode from the management request and sends the target topology mode to the switching profile generation unit 1012.

[0108] Accordingly, the switching profile generating unit 1012 determines the topology switching profile according to the management request, including:

[0109] A topology switching configuration file corresponding to the target topology mode is determined from one or more configuration files stored in the storage unit 1013 ; when no topology switching configuration file exists in the configuration files, a topology switching configuration file is dynamically generated according to the target topology mode.

[0110] In some possible implementations, the processor topology switching method further includes:

[0111] The management request parsing unit 1011 parses the topology switching configuration file from the management request. The topology switching configuration file is generated by the management client 103 according to an operation instruction input by a user.

[0112] Accordingly, the switching profile generating unit 1012 determines the topology switching profile according to the management request, including:

[0113] Upon receiving the topology switching profile sent by the management request parsing unit 1011, the switching profile generating unit 1012 sends the topology switching profile to the topology switching controller 102. If the switching profile generating unit 1012 does not receive the topology switching profile sent by the management request parsing unit 1011, it determines the topology switching profile based on the target topology mode. In other words, the topology switching profile does not exist in the management request.

[0114] In some possible implementations, the handover profile generating unit 1012 determines the topology handover profile according to the management request, including:

[0115] The number of graphics processors corresponding to each central processor is calculated according to the target topology mode; the target logical connection relationship between the central processor and the graphics processor is established according to the number of graphics processors corresponding to each central processor and the demand information of the central processor for the graphics processor; and the topology switching configuration file is generated according to the target logical connection relationship.

[0116] For example, a blank configuration file is first created, and the target logical connection relationship is added to the blank configuration file to obtain the above-mentioned topology switching configuration file. The target logical connection relationship can exist in the topology switching configuration file in the form of graphics and / or text. The existence mode of the target logical connection relationship can be set by the user in the management client 103.

[0117] The requirement information here describes the CPU's requirements for the GPU assigned to it. Assigning a GPU that doesn't meet this requirement information to the CPU may affect the collaborative processing performance of the CPU and GPU, thereby reducing server performance. Therefore, in this embodiment of the application, a GPU that meets this requirement information is assigned to the CPU to maximize server performance.

[0118] In some possible implementations, the management controller 101 further includes: a feasibility verification unit 1014, the feasibility verification unit 1014 is connected to the storage unit 1013 and the switching profile generation unit 1012, and the processor topology switching method further includes:

[0119] The feasibility verification unit 1014 obtains the preset topology switching information from the storage unit 1013, and the preset topology switching information includes the switchable relationship between one or more topology modes; the feasibility verification unit 1014 performs feasibility verification on the management request according to the preset topology switching information to obtain a feasibility verification result; when the feasibility verification result is that the feasibility verification is passed, the feasibility verification unit 1014 controls the switching profile generation unit 1012 to determine the topology switching profile according to the management request.

[0120] In some possible implementations, the feasibility verification unit 1014 is further connected to the management request parsing unit 1011; and the processor topology switching method further includes:

[0121] The feasibility verification unit 1014 obtains the current topology mode from the storage unit 1013 and the target topology mode sent by the management request parsing unit 1011, and determines whether the current topology mode can be switched to the target topology mode according to the preset topology switching information; the feasibility verification unit 1014 determines that the management request verification is successful when the current topology mode can be switched to the target topology mode; the feasibility verification unit 1014 determines that the management request verification fails when the current topology mode cannot be switched to the target topology mode.

[0122] In some possible implementations, the management controller 101 also includes: a switching mode verification unit 1015, the switching mode verification unit 1015 is connected to the management request parsing unit 1011 and the switching profile generation unit 1012, and the processor topology switching method also includes: the switching mode verification unit 1015 receives the topology switching profile sent by the management request parsing unit 1011, and parses the topology switching profile to obtain the actual topology mode; the switching mode verification unit 1015 matches the actual topology mode and the target topology mode to obtain a matching result; when the matching result indicates that the actual topology mode and the target topology mode match, the switching mode verification unit 1015 controls the switching profile generation unit 1012 to send the topology switching profile to the topology switching controller 102.

[0123] In some possible implementations, the management controller 101 also includes: a system restart unit 1016, which is connected to the topology switching controller 102; the processor topology switching method also includes: when the system restart unit 1016 receives a switching success message sent by the topology switching controller 102, it represents that the topology switching controller 102 has successfully executed the switching action, but it is also necessary to restart the server's operating system to make the target logical connection relationship after the switch take effect.

[0124] In some possible implementations, the system restart unit 1016 is further connected to the feasibility verification unit 1014 and the switching mode verification unit 1015. When the system restart unit 1016 receives the switching success message sent by the topology switching controller 102, it restarts the operating system of the server, including:

[0125] The system restart unit 1016 determines the topology switching result of the processor based on the feasibility verification result, the switching result message sent by the topology switching controller 102, or at least one of the matching results; the system restart unit 1016 restarts the operating system when the topology switching result indicates that the switching is successful. The successful switching includes: the feasibility verification result is that the feasibility verification passed, the matching result indicates that the actual topology mode and the target topology mode match, and the switching result message is a switching success message.

[0126] In some possible implementations, the above-mentioned topology switching system also includes a switching result feedback unit 1017, which is connected to the feasibility verification unit 1014, the switching mode verification unit 1015 and the management client 103. The processor topology switching method also includes: when the topology switching result indicates a switching failure, the switching result feedback unit 1017 generates a switching failure message to send to the management client 103 for display. The switching failure includes at least one of the following: the feasibility verification result is a feasibility verification failure, the matching result indicates that the actual topology mode and the target topology mode do not match, and the switching result message is a switching failure message.

[0127] In some possible implementations, the processor topology switching method further includes:

[0128] The handover result feedback unit 1017 determines the handover failure reason and generates handover adjustment indication information according to the handover failure reason; the handover result feedback unit 1017 adds the handover failure indication information and the handover adjustment indication information to the handover failure message, and the handover failure indication information is used to indicate the handover failure.

[0129] In some possible implementations, the switching result feedback unit 1017 is further connected to the storage unit 1013, and the processor topology switching method further includes:

[0130] When the topology switching result indicates a switching failure, the switching result feedback unit 1017 determines a switching failure level according to the switching failure cause, and records the switching failure cause in a log file in the storage unit 1013 according to the switching failure level.

[0131] In some possible implementations, the processor topology switching method further includes:

[0132] The management client 103 displays the switching adjustment instruction information and updates the management request according to the switching adjustment instruction information, so as to send the management request to the management request parsing unit 1011 when receiving the user's confirmation operation on the management request.

[0133] In some possible implementations, the management controller 101 also includes: a resource management control unit 1018, which is connected to the topology switching controller 102 and the switching profile generation unit 1012; the processor topology switching method also includes: when the resource management control unit 1018 receives a switching success message sent by the topology switching controller 102, it allocates the target resources required for the central processor operation to the central processor of the server according to the topology switching profile sent by the switching profile generation unit 1012.

[0134] Of course, when the resource management control unit 1018 does not receive the handover success message sent by the topology handover controller 102, it indicates that the handover has failed, including feasibility verification failure, mismatch between the current topology mode and the target topology mode, and handover failure of the topology handover controller 102. In this case, the resource management control unit 1018 does not need to allocate the target resources required for the CPU operation to the CPU of the server.

[0135] In some possible implementations, the processor topology switching method further includes:

[0136] The resource management control unit 1018 extracts the number of graphics processors corresponding to each central processor from the topology switching configuration file; the resource management control unit 1018 adjusts resource parameters according to the number of graphics processors corresponding to the central processor, and the resource parameters include power supply parameters and / or heat dissipation parameters, and the power supply parameters and heat dissipation parameters are positively correlated with the number of graphics processors; the resource management control unit 1018 adjusts the target resources of the central processor according to the resource parameters, and the target resources include power supply resources and / or heat dissipation resources.

[0137] The power supply parameters may be calculated using a first preset formula, which may be a formula established based on historical experience, or by searching a preset mapping table for power supply parameters corresponding to the number of graphics processors.

[0138] The heat dissipation parameter may be calculated using a second preset formula, which may be a formula established based on historical experience, or by searching a preset mapping table for heat dissipation parameters corresponding to the number of graphics processors.

[0139] In some possible implementations, the management controller 101 further includes: a server performance detection unit 1019, the server performance detection unit 1019 is connected to the management client 103 and the topology switching controller 102, and the processor topology switching method further includes:

[0140] When receiving the switching success message sent by the topology switching controller 102 , the server performance detection unit 1019 detects the server performance to obtain performance parameters, and sends the performance parameters to the management client 103 .

[0141] In some possible implementations, the management controller 101 also includes: a link configuration unit 1010, and the link configuration unit 1010 is connected to the management request parsing unit 1011; the processor topology switching method also includes: the management request parsing unit 1011 sends the management request to the link configuration unit 1010 when the request type corresponds to a link configuration request; the link configuration unit 1010 configures the link between the switching profile generation unit 1012 and the topology switching controller 102, and the link configuration request carries the effective time and transmission performance parameters of the link.

[0142] In some possible implementations, the management client 103 further includes: a link configuration estimation unit 1031, and the processor topology switching method further includes:

[0143] The link configuration estimating unit 1031 estimates the effective time and transmission performance parameters according to the number of processors on the server for display, and generates a link configuration request according to the effective time and transmission performance parameters upon receiving a user confirmation operation on the effective time and transmission performance parameters.

[0144] It is understood that the above-mentioned processor topology switching method corresponds to the aforementioned processor topology switching system 100, allowing the user to simply enter an operation instruction in the management client 103. The management client 103 can generate a management request based on the operation instruction and send it to the management controller 101. The management controller 101 generates a topology switching configuration file based on the management request to control the topology switching controller 102 to switch the logical connection relationship between the processors. In the embodiment of the present application, the target logical connection relationship in the topology switching configuration file is generated by the management controller 101, and the user only needs to enter a simple operation instruction. This reduces the operation complexity and cost, and can effectively reduce the switching error rate.

[0145] The present application also provides a server, including the aforementioned processor topology switching system 100 .

[0146] 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.

[0147] The above is a detailed introduction to a processor topology switching system, a processor topology switching method, and a server provided by the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and core ideas of the present application. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the scope of protection of the claims of the present application.

Claims

1. A processor topology switching system, characterized in that: include: A management controller and a topology switching controller, the management controller comprising: a management request parsing unit and a switching configuration file generating unit, the switching configuration file generating unit being connected to the management request parsing unit and the topology switching controller; The management request parsing unit is configured to: receive a management request, parse a request type of the management request, and, when the request type corresponds to a topology switching request, control the switching profile generating unit to generate a topology switching profile corresponding to the management request, wherein the management request is generated based on an operation instruction input by a user on a management client, and the topology switching profile includes a target logical connection relationship between processors on a server; The switching profile generating unit is configured to: generate a topology switching profile corresponding to the management request, and send the topology switching profile to the topology switching controller; The topology switching controller is configured to switch the logical connection relationship between processors according to the topology switching configuration file.

2. The system according to claim 1, wherein: The management controller further includes: a storage unit, the handover profile generating unit being connected to the storage unit; The management request parsing unit is further configured to: parse a target topology mode from the management request, and send the target topology mode to the handover configuration file generating unit; The handover configuration file generating unit is further configured to: Determining a topology switching configuration file corresponding to the target topology mode from one or more configuration files stored in the storage unit; When the topology switching configuration file does not exist in the configuration file, the topology switching configuration file is dynamically generated according to the target topology mode.

3. The system according to claim 2, characterized in that The management request parsing unit is further configured to: parsing the topology switching configuration file from the management request, where the topology switching configuration file is generated by the management client according to an operation instruction input by a user; The handover configuration file generating unit is further configured to: Upon receiving the topology switching configuration file sent by the management request parsing unit, sending the topology switching configuration file to the topology switching controller; When the topology switching configuration file sent by the management request parsing unit is not received, the topology switching configuration file is determined according to the target topology mode.

4. The system according to claim 3, characterized in that The handover configuration file generating unit is further configured to: Calculate the number of graphics processors corresponding to each central processing unit according to the target topology mode; Establishing a target logical connection relationship between the central processor and the graphics processor according to the number of the graphics processors corresponding to each central processor and the demand information of the central processor for the graphics processor; The topology switching configuration file is generated according to the target logical connection relationship.

5. The system according to claim 4, characterized in that The management controller further includes a feasibility verification unit connected to the storage unit and the handover profile generation unit, and configured to: Acquire preset topology switching information from the storage unit, where the preset topology switching information includes a switchable relationship between one or more topology modes; Performing feasibility verification on the management request according to the preset topology switching information to obtain a feasibility verification result; When the feasibility verification result is feasibility verification passed, the switching profile generating unit is controlled to determine the topology switching profile according to the management request.

6. The system according to claim 5, characterized in that The feasibility verification unit is also connected to the management request parsing unit; The feasibility verification unit is further configured to: Acquire a current topology mode from the storage unit and receive a target topology mode sent by the management request parsing unit, and determine whether the current topology mode can be switched to the target topology mode according to the preset topology switching information; When the current topology mode can be switched to the target topology mode, determining that the management request verification is successful; When the current topology mode cannot be switched to the target topology mode, it is determined that the management request verification fails.

7. The system according to claim 6, characterized in that The management controller further includes a switching mode verification unit connected to the management request parsing unit and the switching configuration file generating unit, and configured to: receiving a topology switching configuration file sent by the management request parsing unit, and parsing the topology switching configuration file to obtain an actual topology mode; Matching the actual topology pattern with the target topology pattern to obtain a matching result; When the matching result indicates that the actual topology pattern matches the target topology pattern, the switching profile generating unit is controlled to send the topology switching profile to the topology switching controller.

8. The system according to claim 7, characterized in that The management controller further includes: a system restart unit, the system restart unit being connected to the topology switching controller; The system restart unit is configured to restart the operating system of the server upon receiving a switching success message sent by the topology switching controller.

9. The system according to claim 8, characterized in that The system restart unit is further connected to the feasibility verification unit and the switching mode verification unit, and the system restart unit is further configured to: Determine a topology switching result of the processor according to at least one of the feasibility verification result, the switching result message sent by the topology switching controller, or the matching result; When the topology switching result indicates that the switching is successful, the operating system is restarted, and the switching success includes: the feasibility verification result is feasibility verification passed, the matching result indicates that the actual topology mode and the target topology mode match, and the switching result message is a switching success message.

10. The system according to claim 9, characterized in that The invention also includes a switching result feedback unit connected to the feasibility verification unit, the switching mode verification unit and the management client, and the switching result feedback unit is used to: When the topology switching result indicates a switching failure, a switching failure message is generated and sent to the management client for display. The switching failure includes at least one of the following: the feasibility verification result is a feasibility verification failure, the matching result indicates that the actual topology mode and the target topology mode do not match, or the switching result message is a switching failure message.

11. The system according to claim 10, wherein: The switching result feedback unit is further configured to: Determining a handover failure reason, and generating handover adjustment instruction information according to the handover failure reason; The handover failure indication information and the handover adjustment indication information are added to the handover failure message, where the handover failure indication information is used to indicate a handover failure.

12. The system according to claim 11, wherein: The switching result feedback unit is further connected to the storage unit, and is configured to: When the topology switching result indicates a switching failure, a switching failure level is determined according to the switching failure cause, and the switching failure cause is recorded in a log file in the storage unit according to the switching failure level.

13. The system according to claim 11, wherein: The management client is further configured to: The switching adjustment instruction information is displayed, and the management request is updated according to the switching adjustment instruction information, so as to send the management request to the management request parsing unit when a confirmation operation of the user on the management request is received.

14. The system according to any one of claims 1 to 4, characterized in that The management controller further comprises: a resource management control unit connected to the topology switching controller and the switching profile generating unit; The resource management control unit is used to: when receiving the switching success message sent by the topology switching controller, allocate target resources required for the operation of the central processor to the central processor of the server according to the topology switching profile sent by the switching profile generation unit, and the target resources include power supply resources and / or heat dissipation resources.

15. The system according to claim 14, wherein: The resource management control unit is further configured to: Extracting the number of graphics processors corresponding to each central processor from the topology switching configuration file; Adjusting resource parameters according to the number of graphics processors corresponding to the central processing unit, the resource parameters including power supply parameters and / or heat dissipation parameters, the power supply parameters and the heat dissipation parameters being positively correlated with the number of graphics processors; The target resource of the central processing unit is adjusted according to the resource parameter.

16. The system according to any one of claims 1 to 4, characterized in that The management controller further includes a server performance detection unit connected to the management client and the topology switching controller, and configured to: When receiving the switching success message sent by the topology switching controller, the server performance is detected to obtain performance parameters, and the performance parameters are sent to the management client.

17. The system according to any one of claims 1 to 4, characterized in that The management controller further includes: a link configuration unit, the link configuration unit being connected to the management request parsing unit; The management request parsing unit is further configured to: when the request type corresponds to a link configuration request, send the management request to the link configuration unit; The link configuration unit is used to configure the link between the handover configuration file generation unit and the topology handover controller, and the link configuration request carries the effective time and transmission performance parameters of the link.

18. The system according to claim 17, wherein: The management client includes: a link configuration estimation unit; The link configuration estimation unit is used to: estimate the effective time and the transmission performance parameters according to the number of processors on the server for display, and upon receiving a user's confirmation operation on the effective time and the transmission performance parameters, generate the link configuration request according to the effective time and the transmission performance parameters.

19. A processor topology switching method, characterized in that: The invention is applied to a processor topology switching system, comprising: a management controller and a topology switching controller, wherein the management controller comprises: a management request parsing unit and a switching profile generating unit, wherein the switching profile generating unit is connected to the management request parsing unit and the topology switching controller; the method comprises: The management request parsing unit receives a management request to parse a request type of the management request, wherein the management request is generated according to an operation instruction input by a user on a management client; The management request parsing unit controls the switching profile generating unit to generate a topology switching profile corresponding to the management request when the request type corresponds to a topology switching request, wherein the topology switching profile includes a target logical connection relationship between processors on the server; The switching profile generating unit generates a topology switching profile corresponding to the management request, and sends the topology switching profile to the topology switching controller; The topology switching controller switches the logical connection relationship between processors according to the topology switching configuration file.

20. A server, characterized in that: A processor topology switching system comprising any one of claims 1 to 18.

Images