Host partition management architecture, server and method

Abstract

The invention discloses a host partition management architecture, a server and a method, and relates to the technical field of computers, a management unit is also divided into a plurality of management partitions, the management partitions correspond to host partitions of the server, and a host and the management unit are connected through a logic processing unit. And the logic processing unit is responsible for data interaction between the host and the management unit. Therefore, interface expansion can be realized through the logic processing unit, so that sufficient interfaces are provided for the target equipment of the plurality of host partitions of the host, and the target equipment requirements of the plurality of host partitions are met. In addition, the host partition management can be realized only by one management unit, so that the hardware complexity, the cost and the resource waste can be reduced. Besides, the logic processing unit can realize logic processing, and the management partition does not need complex logic processing, so that the design complexity of the management partition can be reduced, the production cycle of the management partition can be shortened, and the cost of the management partition can be reduced.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a host partition management architecture, a server, and a method. Background Art

[0002] A server is a device that processes data to provide data services, and its processing power has a significant impact on the service quality. To improve its processing power, a conventional solution is to physically divide the server into multiple partitions, and each partition can act as an independent system. To enable the partitions to run better, it is necessary to manage devices such as computing, storage, network, and heat dissipation of the partition through a Baseboard Management Controller (BMC).

[0003] In the related art, there are two types of device management architectures for partitions. In the first architecture, a BMC is equipped for each partition, but this solution has high hardware complexity, high cost, and low resource utilization. In the second architecture, a BMC is configured for one or more partitions, or even all partitions share a BMC, but this solution will result in high design complexity, long production cycle, and high cost of the BMC. Summary of the Invention

[0004] This application provides a host partition management architecture, a server, and a method to at least solve the problems of high hardware complexity, high cost, low BMC resource utilization, and long BMC production cycle in the related art.

[0005] This application provides a host partition management architecture, including: a host, a logical processing unit, and a management unit;

[0006] The host includes multiple host partitions, and target devices are set in the host partitions;

[0007] The management unit includes multiple management partitions, and different management partitions operate independently;

[0008] The logical processing unit includes a first transmission component, a routing component, and a second transmission component;

[0009] The first transmission component is connected to the routing component and the management partition, and the second transmission component is connected to the routing component and the host partition;

[0010] When the management partition manages the target devices in the host partition, the first transmission component and the second transmission component are used to transmit data, and the routing component is used to parse the data transmitted by the first transmission component and the second transmission component.

[0011] This application also provides a server, including the foregoing host partition management architecture.

[0012] The present application also provides a host partition management method, which is applied to a host partition management architecture. The host partition management architecture includes: a host, a logical processing unit, and a management unit; the host includes multiple host partitions, the management unit includes multiple management partitions, and the logical processing unit includes a first transmission component, a routing component, and a second transmission component; the method includes:

[0013] The management partition generates a first management command and sends the first management command to the routing component through the first transmission component; the routing component generates a second management command according to the first management command and sends the second management command to the target device of the host partition through the second transmission component; the target device executes the second management command.

[0014] Through the present application, the management unit is also divided into multiple management partitions, corresponding to the host partitions of the server, and is connected between the host and the management unit through a logical processing unit. The logical processing unit is responsible for data interaction between the host and the management unit. In this way, interface expansion can be achieved through the logical processing unit to provide sufficient interfaces for the target devices of multiple host partitions of the host to meet the requirements of the target devices of multiple host partitions. In addition, the present application only needs one management unit to implement host partition management, thereby reducing hardware complexity, cost, and resource waste. In addition, the logical processing unit of the present application can implement logical processing, and the management partition does not need to perform complex logical processing, thereby reducing the design complexity of the management partition, shortening its production cycle, and reducing its cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present application, the drawings required for the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0016] Figure 1 is a schematic diagram of a host partition management architecture in the related art;

[0017] Figure 2 is a schematic diagram of another host partition management architecture in the related art;

[0018] Figure 3 is a schematic diagram of a host partition management architecture structure provided by an embodiment of the present application;

[0019] Figure 4 is a schematic diagram of another host partition management architecture structure provided by an embodiment of the present application;

[0020] Figure 5It is a schematic structural diagram of a host partition management method provided by an embodiment of the present application;

[0021] Figure 6 It is another schematic structural diagram of a host partition management method provided by an embodiment of the present application. Detailed implementation manners

[0022] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the protection scope of the present application.

[0023] It should be noted that in the description of the present application, the terms "including", "comprising" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. The terms "first", "second", etc. in the present application are used to distinguish similar objects, rather than to describe a specific order or sequence.

[0024] In recent years, servers have been widely used in various data processing scenarios, such as cloud computing, artificial intelligence, big data and other scenarios. The server can be used in combination with front-end technologies such as clients. The front-end technology implements the client for data display. The client is communicatively connected to the server. The client can send a data request to the server so that the server processes the data and returns it to the client for display.

[0025] In order to improve the data processing ability of the server, a conventional solution is to divide the hardware of the server into multiple partitions. Each partition can independently run a set of host systems to independently carry services and provide services. Each partition needs to have complete computing, storage, network drivers, radiators and other target device supports, and the management of these target devices depends on the BMC. As the number of partitions increases, the number of target devices of the server will increase significantly compared with the traditional non-partitioned server system. This poses a greater challenge to the BMC of the server because the controller resources provided by the current general BMC are limited and cannot provide enough interfaces to manage the target devices of multiple host partitions, making it difficult to meet the growing device management requirements.

[0026] In the embodiments of the present application, since the partitions of the server are partitions for providing services and data processing, they can be called host partitions. In the related art, there are two common host partition management architectures, which are described in detail below.

[0027] In the first host partition management architecture, in order to better manage the target devices of multiple host partitions, a BMC needs to be set for each host partition. The number of BMCs is the same as the number of host partitions, and each BMC is used to manage all the target devices in the corresponding host partition. For example, Figure 1 is a schematic diagram of a host partition management architecture. Refer to Figure 1 As shown, in the host partition management architecture, two host partitions HA1 and HA2, and two BMCs: BMC1 and BMC2 are exemplarily given. Among them, BMC1 corresponds to host partition HA1, and BMC1 is used to manage each target device in host partition HA1. BMC2 corresponds to host partition HA2, and BMC2 is used to manage each target device in host partition HA2.

[0028] From the above Figure 1 It can be seen from the shown host partition management architecture that the number of BMCs is the same as the number of host partitions. Thus, when there are many host partitions, more BMCs are needed, resulting in higher costs and higher hardware complexity. And one BMC manages one host partition. When the host partition is not running, it may cause waste of the BMC resources, and the resource utilization rate is low. For example, for Figure 1 the shown host partition management architecture, when host partition HA1 is not running, BMC1 is idle, causing waste of the resources of BMC1, and the BMC resource utilization rate of the server is low.

[0029] In the second host partition management architecture, a high-specification BMC is provided to provide management for multiple host partitions. There is only one BMC, and one BMC is used to manage all the target devices in multiple host partitions. For example, Figure 2 is a schematic diagram of another host partition management architecture. Refer to Figure 2 As shown, in the host partition management architecture, two host partitions HA1 and HA2 and one BMC are exemplarily given. Among them, this BMC corresponds to host partitions HA1 and HA2, and is used to manage all the target devices in host partitions HA1 and HA2.

[0030] From the above Figure 2 It can be seen from the shown host partition management architecture that the number of BMCs is not related to the number of host partitions. Even if the number of host partitions is very large, the management of target devices can still be achieved with fewer BMCs. In order to achieve the management of the target devices of multiple host partitions by one BMC, the performance requirements for the BMC are relatively high, and a higher-specification BMC is required to ensure the management of the target devices of multiple host partitions. When the number of host partitions is large, the performance requirements for the BMC are even higher, resulting in higher costs, higher design complexity, and longer production cycles for a single BMC.

[0031] To solve the above technical problems, in this application, the BMC is also divided into multiple partitions, called management partitions, corresponding to the host partitions of the server, and a logical processing unit is set between the host and the BMC. The logical processing unit is responsible for data interaction between the host and the BMC. In this way, interface expansion can be achieved through the logical processing unit to provide sufficient interfaces for the target devices in multiple host partitions of the host, meeting the requirements of the target devices in multiple host partitions. In addition, this application only requires one BMC to achieve management, thereby reducing hardware complexity, cost, and resource waste. In addition, the management partition of this application interacts with the host partition through a routing component, and the BMC does not need to perform complex logical processing. The complex logical processing can be executed by the routing component, thereby reducing the design complexity of the BMC, shortening its production cycle, and reducing its cost.

[0032] To enable those skilled in the art of this technology to better understand the solution of this application, the following further elaborates on this application in combination with the accompanying drawings and specific implementation manners.

[0033] Figure 3 It is a schematic structural diagram of a host partition management architecture provided by an embodiment of this application. Refer to Figure 3 As shown, the above host partition management architecture includes a host, a logical processing unit, and a management unit. The logical processing unit is connected to the host and the management unit.

[0034] Among them, the host includes multiple host partitions. Figure 3 Exemplarily, two host partitions HA1 and HA2 are given. In practice, the number of host partitions can be flexibly set according to actual scenario requirements. For example, it can be 3 or more host partitions. When the data processing requirement is higher, more host partitions can be divided. Each host partition can execute an independent host physical system, and the host physical systems of different host partitions run independently without affecting each other. Thus, the host of the server can provide more powerful processing capabilities through multiple host partitions, improving the service quality. Here, the host can be a partition obtained by hardware partitioning of the server.

[0035] A target device is set in each of the above host partitions. Figure 3 Exemplarily, one target device is given, but in actual applications, there can be multiple target devices in the host partition. For example, the target device can be a hard disk, a network card, a GPU (Graph Processing Unit) acceleration card, various types of sensors, etc.

[0036] Corresponding to the above host, this application also divides the management unit into multiple management partitions. Refer to Figure 3As shown, the management unit includes two management partitions MA1 and MA2, and different management partitions operate independently. The management unit here can be the aforementioned BMC or any integrated circuit with management functions. Each management partition can correspond to one or more host partitions.

[0037] To achieve the interaction between the above-mentioned host partition and the management partition, in the embodiment of the present application, a logic processing unit is set between the host and the management unit for connecting the host partition and the management partition. The logic processing unit can perform data transmission and data parsing between the host partition and the management partition. Thus, the management partition does not need to perform complex logic processing and data parsing, and these complex logic processing and data parsing can be executed by the logic processing unit, reducing the complexity, cost, and production cycle of the management unit. The logic processing unit can parse the first management command sent by the management partition and generate a second management command according to the parsing result and send it to the target device in the host partition. In addition, the logic processing unit can also receive the response data returned by the target device in the host partition and generate a response message according to the response data to return the response message to the management partition. In this way, through the connection of the logic processing unit, the management of the target device by the management partition can be realized.

[0038] The logic processing unit can be implemented by any integrated circuit with logic processing capabilities. For example, common circuits can be CPLD (Complex Programmable Logic Device), FPGA (Field Programmable Gate Array), and ASIC (Application Specific Integrated Circuits).

[0039] Specifically, the above-mentioned logic processing unit includes a first transmission component, a routing component, and a second transmission component. The first transmission component is connected to the routing component and the management partition, and the second transmission component is connected to the routing component and the host partition. Each management partition is connected to the above-mentioned first transmission component, realizing the connection between the management unit and the logic processing unit. Each target device in each host partition is connected to the second transmission component, realizing the connection between the host and the logic processing unit.

[0040] When the management partition manages the target device in the host partition, the first transmission component and the second transmission component are used for data transmission, and the routing component is used to process the data transmitted by the first transmission component and the second transmission component.

[0041] For example, the first transmission component can transmit the first management command sent by the management partition to the routing component, and the routing component parses the first management command and generates a second management command. The second transmission component sends the second management command to the target device in the host partition so that the target device executes the second management command. After executing the second management command, the target device can send response data to the second transmission component, and the second transmission component can forward the response data to the routing component. The routing component generates a response message based on the response data and sends the response message to the first transmission component, and the first transmission component returns the response message to the management partition.

[0042] In some embodiments, as shown in Figure 4 The first transmission component includes multiple groups of first transmission processing interfaces. The multiple groups of first transmission processing interfaces are connected to the routing component. One group of first transmission processing interfaces is connected to one management partition, and different management partitions are connected to different first transmission processing interfaces. The first transmission processing interfaces are used for data transmission between the routing component and the management partition corresponding to the first transmission processing interfaces. In this way, the management data of multiple management partitions can be transmitted through multiple groups of first transmission processing interfaces respectively, ensuring the independence of different management partitions.

[0043] The connection between the first transmission processing interface and the routing component realizes the connection between the first transmission component and the routing component. The connection between the first transmission processing interface and the management partition realizes the connection between the logical processing unit and the host.

[0044] It can be understood that each group of first transmission processing interfaces may include multiple first transmission processing interfaces. The number of groups of first transmission processing interfaces is the same as the number of host partitions and the number of management partitions. When the number of host partitions is N, the number of management partitions and the number of groups of first transmission processing interfaces are both N. N is an integer greater than or equal to 1.

[0045] In practical applications, the number of first transmission processing interfaces included in each group of first transmission processing interfaces can be the same. When the number of first transmission processing interfaces in the logic processing unit is M and the number of management partitions is N, the number of first transmission processing interfaces for each management partition can be the integer part of M / N. For example, the first transmission processing interfaces of the logic processing unit can be I1_1, I1_2, ..., I1_15. When the number of management partitions is 3, 5 first transmission processing interfaces I1_1, I1_2, ..., I1_5 form the first group of first transmission processing interfaces and serve as the first transmission processing interfaces of the first management partition. 5 first transmission processing interfaces I1_6, I1_7, ..., I1_10 form the second group of first transmission processing interfaces and serve as the first transmission processing interfaces of the second management partition. 5 first transmission processing interfaces I1_11, I1_12, ..., I1_15 form the third group of first transmission processing interfaces and serve as the first transmission processing interfaces of the third management partition.

[0046] However, for management partitions with a relatively large amount of management data, more first transmission processing interfaces can be set. For management partitions with a relatively small amount of management data, fewer first transmission processing interfaces can be set. In this way, on the basis of ensuring the normal operation of each management partition, the waste of first transmission processing interfaces can be avoided as much as possible, which helps to fully improve their utilization rate.

[0047] In some embodiments, the number of first transmission processing interfaces corresponding to the management partition can be determined according to parameters such as the running duration of the host partition corresponding to the management partition and the number of target devices. For example, for a host partition with a relatively short running duration and a relatively small number of target devices, fewer first transmission processing interfaces can be set for its corresponding management partition. For a host partition with a relatively long running duration and a relatively large number of target devices, more first transmission processing interfaces can be set for its corresponding management partition.

[0048] It can be understood that the first transmission processing interfaces in the same management partition can be arranged at different positions, which may or may not be adjacent. For example, the physical position order of the first transmission processing interfaces I1_1, I1_2,..., I1_5 of the first management partition, the first transmission processing interfaces I1_6, I1_7,..., I1_10 of the second management partition, and the first transmission processing interfaces I1_11, I1_12,..., I1_15 of the third management partition can be I1_1, I1_2, I1_3, I1_4, I1_5, I1_6, I1_7, I1_8, I1_9, I1_10, I1_11, I1_12, I1_13, I1_14, I1_15. Of course, it can also be I1_2, I1_12, I1_5, I1_6, I1_10, I1_7, I1_8, I1_3, I1_9, I1_11, I1_4, I1_13, I1_14, I1_1, I1_15.

[0049] Corresponding to the division of the above first transmission component into multiple first transmission processing interfaces, referring to Figure 4 As shown, each management partition is also divided into multiple management interfaces. The multiple management interfaces are all connected to the first transmission component to realize the connection between the management partition and the first transmission component, and further realize the connection between the management unit and the logic processing unit. The multiple management interfaces of the same management partition are connected to multiple continuously distributed positions on the first transmission component, and the first transmission component is used for data transmission between each management interface of each management partition and the routing component. It can be understood that compared with the multiple management interfaces of the same management partition being connected to multiple discretely distributed positions on the first transmission component, in this application, the same management partition is connected to multiple continuously distributed positions on the first transmission component, so as to shorten the connection line as much as possible, improve the circuit neatness, and also facilitate the inspection and maintenance of the connection relationship.

[0050] Among them, the management interface is used to generate and send a first management command to the first transmission component, or receive a response message sent by the first transmission component for the first management command. The first management command here is a management command of the management partition for the target device, which is used to write the parameters of the target device into the target device memory or read the parameters of the target device from the target device memory.

[0051] In the embodiment of this application, for a management partition, the number of management interfaces can be flexibly set according to the actual application scenario. By default, the number of management interfaces in different management partitions can be the same. However, it can also be increased or decreased according to the actual needs of different management partitions. For example, if the management data volume of a management partition is large, more management interfaces can be set for this management partition; if the management data volume of a management partition is small, fewer management interfaces can be set for this management partition.

[0052] In some embodiments, referring to Figure 4 As shown, the multiple management interfaces are connected to multiple first transmission processing interfaces. The multiple first transmission processing interfaces connected by the multiple management interfaces in the same management partition are adjacent, and the first transmission processing interfaces connected by the multiple management interfaces in different management partitions are arranged in sequence. The management interfaces are used to perform data interaction with the routing component through the first transmission processing interfaces. In this way, the length of the connection line can be shortened as much as possible, the circuit neatness can be improved, and the circuit maintenance can be facilitated.

[0053] Specifically, a first transmission processing interface can be connected to one or more management interfaces to realize the connection between the first transmission component and the management partition, and further realize the connection between the logic processing unit and the management unit.

[0054] In the embodiment of the present application, the multiple management interfaces in the same management partition are connected to adjacent first transmission processing interfaces. For example, if the first transmission processing interfaces I1_1, I1_2, I1_3, I1_4, I1_5, I1_6, I1_7, I1_8, I1_9, I1_10, I1_11, I1_12, I1_13, I1_14, I1_15 are arranged in sequence, then the multiple management interfaces of the first management partition are respectively connected to the first transmission processing interfaces I1_1, I1_2,..., I1_5 arranged adjacent to each other in sequence, the multiple management interfaces of the second management partition are respectively connected to the first transmission processing interfaces I1_6, I1_7,..., I1_10 arranged adjacent to each other in sequence, and the multiple management interfaces of the third management partition are respectively connected to the first transmission processing interfaces I1_11, I1_12,..., I1_15 arranged adjacent to each other in sequence. Compared with the multiple management interfaces of the first management partition being respectively connected to the first transmission processing interfaces I1_6, I1_7,..., I1_10 arranged adjacent to each other in sequence, the multiple management interfaces of the second management partition being respectively connected to the first transmission processing interfaces I1_11, I1_12,..., I1_15 arranged adjacent to each other in sequence, and the multiple management interfaces of the third management partition being respectively connected to the first transmission processing interfaces I1_1, I1_2,..., I1_5 arranged adjacent to each other in sequence, the above connection method of the present application can shorten the connection line as much as possible, reduce costs, and improve the circuit neatness and maintainability.

[0055] Further, for the same management partition, its management interfaces are also arranged adjacent to each other in sequence, so that the sequence of the management interfaces can be the same as that of the first transmission processing interfaces connected thereto, which can further shorten the connection lines and avoid line crossovers. For example, if the five management interfaces I2_1 to I2_5 of the first management partition are arranged in sequence, the corresponding first transmission processing interfaces I1_1, I1_2,..., I1_5 are also arranged in sequence, and the first transmission processing interface I1_1 is closest to the management interface I2_1, the first transmission processing interface I1_2 is closest to the management interface I2_2, the first transmission processing interface I1_3 is closest to the management interface I2_3, the first transmission processing interface I1_4 is closest to the management interface I2_4, and the first transmission processing interface I1_5 is closest to the management interface I2_5. Therefore, the first transmission processing interface I1_1 is connected to the management interface I2_1, the first transmission processing interface I1_2 is connected to the management interface I2_2, the first transmission processing interface I1_3 is connected to the management interface I2_3, the first transmission processing interface I1_4 is connected to the management interface I2_4, and the first transmission processing interface I1_5 is connected to the management interface I2_5.

[0056] In some embodiments, as shown in Figure 4 multiple management interfaces of the management partition include multiple service management interfaces, and different service management interfaces operate independently to ensure the independence and accuracy of service transmission; one service management interface is connected to one first transmission processing interface, and the service management interface is used to send a first management command for a target device to the first transmission processing interface, and / or receive a response message for the first management command from the first transmission processing interface. In the embodiments of the present application, the service management interface and the first transmission processing interface can be connected one-to-one to ensure the accurate transmission of the first management command.

[0057] The connection between the above service management interface and the first transmission processing interface forms a connection channel between the logic processing unit and the host partition for transmitting the first management command and the response message, and thus can be referred to as a service transmission channel. Each service transmission channel corresponds to one first transmission processing interface and one first management command.

[0058] Further, as shown in Figure 4As shown, the multiple management interfaces further include a first channel management interface. For a management partition, the first transmission processing interface connected to the first channel management interface is located before or after the first transmission processing interface connected to the service management interface. The first channel management interface is used to manage the connection channel from the first transmission processing interface to the service management interface. In this application, before the routing component sends a response message to the management partition, the first channel management interface can manage the connection channel from the first transmission processing interface to the service management interface to allocate a connection channel for the response message, so as to transmit the response message from the routing component to the management partition through the allocated connection channel. The allocated connection channel is an idle channel or a channel with a small data volume, thereby effectively improving the sending success rate of the response message and reducing its sending delay.

[0059] Here, the first transmission processing interface connected to the first channel management interface can be a second channel management interface, which is used to manage the connection channel between the first transmission processing interface and the service management interface together with the first channel management interface.

[0060] In the embodiment of this application, for the same management partition, its first channel management interface is located before or after the service management interface. The first transmission processing interfaces connected to its service management interface are adjacent, and the first transmission processing interface connected to its first channel management interface is located before or after the first transmission processing interface connected to the service management interface. In this way, the crossing of connection lines is avoided as much as possible, and the length of the connection lines is shortened.

[0061] For example, if the five management interfaces I2_1 to I2_5 of the first management partition are arranged in sequence, where I2_1 to I2_4 are service management interfaces and I2_5 is the first channel management interface. Their corresponding first transmission processing interfaces I1_1, I1_2,..., I1_5 are also arranged in sequence. Therefore, the first transmission processing interface I_{1_1} is connected to the management interface I_{2_1}, the first transmission processing interface I_{1_2} is connected to the management interface I_{2_2}, the first transmission processing interface I_{1_3} is connected to the management interface I_{2_3}, the first transmission processing interface I_{1_4} is connected to the management interface I_{2_4}, and the first transmission processing interface I_{1_5} is connected to the management interface I_{2_5}. The first transmission processing interface I_{1_5} connected to the first channel management interface I_{2_5} is located after the first transmission processing interfaces I_{1_2} to I_{1_4} respectively connected to the service management interfaces I_{2_1} to I_{2_4}.

[0062] Of course, when the first channel management interface I2_5 is before the service management interfaces I2_1 to I2_4, that is, if the five management interfaces I2_5, I2_1, I2_2, I2_3, I2_4 of the first management partition are arranged in sequence, where I2_1 to I2_4 are service management interfaces and I2_5 is the first channel management interface. Their corresponding first transmission processing interfaces I1_1, I1_2,..., I1_5 are also arranged in sequence. Therefore, the first transmission processing interface I1_1 is connected to the management interface I2_5, the first transmission processing interface I1_2 is connected to the management interface I2_1, the first transmission processing interface I1_3 is connected to the management interface I2_2, the first transmission processing interface I1_4 is connected to the management interface I2_3, and the first transmission processing interface I1_5 is connected to the management interface I2_4. The first transmission processing interface I1_1 connected to the first channel management interface I2_5 is before the first transmission processing interfaces I1_2 to I1_5 respectively connected to the service management interfaces I2_1 to I2_4.

[0063] In summary, the position of the first transmission processing interface connected to the first channel management interface corresponds to the relative position relationship between the first channel management interface and the service management interface, so as to shorten the line length as much as possible and avoid line crossing. When the first channel management interface is before the service management interface, the first transmission processing interface connected to the first channel management interface is before the first transmission processing interfaces connected to the service management interface. When the first channel management interface is after the service management interface, the first transmission processing interface connected to the first channel management interface is after the first transmission processing interfaces connected to the service management interface.

[0064] Corresponding to the division of the management interfaces of the above management partition into service management interfaces and the first channel management interface, referring to Figure 4 As shown, the present application divides multiple first transmission processing interfaces into multiple first service processing interfaces and a second channel management interface. Multiple first service processing interfaces are connected to the routing component, and one first service processing interface is connected to one service management interface of one management partition. The first service processing interface is used to transmit the first management command and / or response message. The embodiment of the present application divides the first transmission processing interface into the first service processing interface and assigns a first service processing interface to each service management interface, so as to ensure that the management data sent by the service management interface is managed through an exclusive channel, which helps to improve the transmission accuracy of the management data.

[0065] The above-mentioned second channel management interface is connected to the routing component and the first channel management interface. The second channel management interface is used to manage the connection channel from the service management interface to the first transmission processing interface through the routing component. After generating a response message, the routing component of the logic processing unit can call the second channel management interface to apply for a connection channel for transmitting the response message to the first channel management interface. In this way, it can be ensured that the response message is transmitted through an idle or less-loaded connection channel, and the transmission delay of the response message can be reduced as much as possible.

[0066] For example, I2_5 of a management partition is the first channel management interface, and I1_5 is the second channel management interface of the logic processing unit. Therefore, I2_5 and I1_5 are connected.

[0067] In some possible implementation manners, the second channel management interface is connected to the first channel management interfaces of one or more management partitions. In this way, the number of second channel management interfaces can be reduced, and costs can be saved.

[0068] In one example, the logic processing unit may set multiple second channel management interfaces, and each second channel management interface is connected to the first channel management interface of a management partition. In another example, the logic processing unit may set one second channel management interface, and this second channel management interface is connected to the first channel management interfaces of all management partitions, so that the first channel management interfaces of multiple management partitions share one second channel management interface. For example, I1_5 is the second channel management interface of the logic processing unit, and it can be connected to the first channel management interface I2_5 of the first management partition, the first channel management interface I2_10 of the second management partition, and the first channel management interface I2_15 of the third management partition.

[0069] It can be seen that the foregoing description introduces the connection relationship between the logic processing unit and the management unit. This connection relationship is further implemented through the connection relationship between the first transmission component and the management partition. The connection relationship between the first transmission component and the management partition is further implemented through the connection relationship between the first transmission processing interface and the management interface. The connection relationship between the first transmission processing interface and the management interface further includes: the connection relationship between the first service processing interface and the service management interface, and the connection relationship between the second channel management interface and the first channel management interface.

[0070] For a management partition, the number of service management interfaces of the management partition is the same as the number of first service processing interfaces connected to the management partition, and the number of first service processing interfaces connected to the management partition is less than the number of target devices of the host partitions managed by the management partition, so that multiple target devices of the host partitions managed by the management partition share the first service processing interfaces connected to the management partition. Therefore, through the above architecture, embodiments of the present application can minimize service management interfaces and first service processing interfaces as much as possible without affecting management performance, which helps reduce costs, avoid waste of service management interfaces and first service processing interfaces, make full use of service management interfaces and first service processing interfaces, and improve resource utilization rate.

[0071] The connection relationship between the logic processing unit and the host is introduced in detail below.

[0072] In some possible implementation manners, as shown in Figure 4 a second transmission component includes a plurality of second service processing interfaces, the plurality of second service processing interfaces are all connected to a routing component, and one second service processing interface is connected to one or more target devices. The second service processing interface is configured to send a second management command sent by the routing component to the connected target device, or send response data of the target device for the second management command to the routing component. In the present application, the connection between the target device and the logic processing unit can be implemented through a plurality of second service processing interfaces, and each second service processing interface serves one or more target devices, which can ensure the success rate of service transmission as much as possible, avoid data transmission failure caused by the busy state of the second service processing interface, and help improve the management success rate of the target device.

[0073] It can be understood that the plurality of second service processing interfaces of the second transmission component are used to provide services to all host partitions, and the second service processing interfaces of all host partitions are connected to the routing component to make full use of the routing component and improve the utilization rate of the routing component. When expanding the interface through the logic processing circuit, more second service processing interfaces need to be provided, and the total number of second service processing interfaces of the logic processing unit may be greater than the total number of service management interfaces of each management partition in the management unit. In this way, when the service management interfaces provided by the management unit are limited, more second service processing interfaces can still be used to connect the target devices of more host partitions to meet the management requirements of the target devices of multiple host partitions.

[0074] In practical applications, the above-mentioned multiple second service processing interfaces can also be divided into multiple groups of second service processing interfaces, and each group of second service processing interfaces includes multiple second service processing interfaces. Each group of second service processing interfaces is connected to one or more target devices in a host partition. For example, the second transmission component includes 12 second service processing interfaces I3_1 to I3_12, and there are three host partitions. The target devices in each host partition include network cards, hard disks, GPUs, and temperature sensors. The 12 second service processing interfaces I3_1 to I3_12 can be divided into 3 groups: I3_1 to I3_4, I3_5 to I3_8, and I3_9 to I3_12. Among them, the first group of second service processing interfaces I3_1 to I3_4 corresponds to the first host partition, the second group of second service processing interfaces I3_5 to I3_8 corresponds to the second host partition, and the third group of second service processing interfaces I3_9 to I3_12 corresponds to the third host partition.

[0075] Within the first host partition, the network card is connected to the second service processing interface I3_1, the hard disk is connected to the second service processing interface I3_2, the GPU is connected to the second service processing interface I3_3, and the temperature sensor is connected to the second service processing interface I3_4.

[0076] Within the second host partition, the network card is connected to the second service processing interface I3_5, the hard disk is connected to the second service processing interface I3_6, the GPU is connected to the second service processing interface I3_7, and the temperature sensor is connected to the second service processing interface I3_8.

[0077] Within the third host partition, the network card is connected to the second service processing interface I3_9, the hard disk is connected to the second service processing interface I3_10, the GPU is connected to the second service processing interface I3_11, and the temperature sensor is connected to the second service processing interface I3_12.

[0078] It should be noted that the connection relationships between the target devices in different host partitions and the second service processing interfaces can be the same or different. For example, the connection relationships in the second and third host partitions remain unchanged as above, but within the second host partition, the network card can be connected to the second service processing interface I3_5, the hard disk and the GPU can be connected to I3_6 together, some temperature sensors are connected to the second service processing interface I3_7, and some other temperature sensors are connected to the second service processing interface I3_8.

[0079] In some possible implementation manners, a second service processing interface is connected to one or more target devices in the same host partition. The target devices in the same host partition are connected to a plurality of second service processing interfaces that are continuously distributed. The target devices in the host partition are connected to the second service processing interface that is the nearest and available to the target device. In this way, it can be ensured that the routing component accurately sends the second management command to the corresponding host partition through the second service processing interface, avoiding sending the command to the wrong host partition, and improving the transmission accuracy and management accuracy of the second management command. Moreover, since the target devices in the same host partition are connected to a plurality of consecutive second service processing interfaces, the connection line crossing can be avoided, and the length of the connection line between the host and the logic processing unit can be shortened as much as possible.

[0080] In one example, the plurality of second service processing interfaces included in the second transmission component are the following interfaces arranged in sequence: I3_1 to I3_12. Among them, the first host partition can be connected to a plurality of consecutive second service processing interfaces I3_1 to I3_4 that are the nearest to the first host partition, the second host partition can be connected to a plurality of consecutive second service processing interfaces I3_5 to I3_8 that are the nearest to the second host partition, and the third host partition can be connected to a plurality of consecutive second service processing interfaces I3_9 to I3_12 that are the nearest to the third host partition.

[0081] Within the first host partition, the second service processing interface that is the nearest to the network card is I3_1. Therefore, the network card is connected to the second service processing interface I3_1. The second service processing interface that is the nearest to the hard disk is I3_2. Then, the hard disk is connected to the second service processing interface I3_2. The second service processing interface that is the nearest to the GPU is I3_2. However, I3_2 has been connected to the hard disk and is unavailable. Therefore, the GPU can be connected to the second service processing interface I3_3 that is the second nearest. The second service processing interface that is the nearest to the temperature sensor is I3_4. Thus, it can be connected to the second service processing interface I3_4.

[0082] Within the second host partition, the second service processing interface that is the nearest to the network card is I3_5. Therefore, the network card is connected to the second service processing interface I3_5. The second service processing interface that is the nearest to the hard disk is I3_6. Then, the hard disk is connected to the second service processing interface I3_6. The second service processing interface that is the nearest to the GPU is I3_6. However, I3_6 has been connected to the hard disk and is unavailable. Therefore, the GPU can be connected to the second service processing interface I3_7 that is the second nearest. The second service processing interface that is the nearest to the temperature sensor is I3_8. Thus, it can be connected to the second service processing interface I3_8.

[0083] In the third host partition, the second service processing interface closest to the network card is I3_9, so the network card is connected to the second service processing interface I3_9. The second service processing interface closest to the hard disk is I3_10, so the hard disk is connected to the second service processing interface I3_10. The second service processing interface closest to the GPU is I3_10, but I3_10 is already connected to the hard disk and is unavailable. Therefore, the GPU can be connected to the second closest second service processing interface I3_11. The second service processing interface closest to the temperature sensor is I3_12, so it can be connected to the second service processing interface I3_12.

[0084] In some possible implementation manners, a second service processing interface is connected to one or more target devices of the same type in the same host partition. Thereby, the management of the target devices can be facilitated. By recording the mapping relationship between the device type and the second service processing interface in the routing component, the second service processing interface for sending the second management command can be quickly located, which helps to further improve the transmission efficiency of the second management command.

[0085] For example, the type of the target device can be flexibly divided according to the actual application scenario. For example, the temperature sensor and the image sensor can be classified into one type, that is, both are sensors, and the network card, the hard disk, and the GPU are three types respectively. Thus, all sensors can be connected to a second service processing interface I3_1, the network card is connected to a second service processing interface I3_2, the hard disk is connected to a second service processing interface I3_3, and the GPU is connected to a second service processing interface I3_4. The temperature sensor and the image sensor in the sensors can also be further classified as one type respectively, so that all temperature sensors are connected to a second service processing interface and the image sensors are connected to another second service processing interface.

[0086] In some possible implementation manners, a second service processing interface is connected to one or more target devices adjacent in location in the same host partition. Thereby, not only the number of second service processing interfaces can be reduced, but also the poor connection lines between the host partition and the logic processing unit can be avoided as much as possible, and the length of the connection lines between the host partition and the logic processing unit is shortened.

[0087] In some possible implementation manners, a second service processing interface is connected to one or more target devices with different running times in the same host partition. Thereby, not only the number of second service processing interfaces can be reduced, but also the situation that multiple target devices of the same second service processing interface need to receive the second management command or send response data simultaneously can be avoided as much as possible, reducing the blocking rate of the second service processing interface. In this way, the transmission success rate of the second management command and the response data can be improved as much as possible, or the transmission delay of the second management command and the transmission delay of the response data can be shortened.

[0088] In some possible implementation manners, the overlap degree of the running times of different target devices in the same host partition may be calculated to indicate different degrees of the running times of different target devices. When the overlap degree is larger, it represents that the different degrees of the running times of different target devices are lower; when the overlap degree is smaller, it represents that the different degrees of the running times of different target devices are higher. Thus, different target devices with the lowest overlap degree are connected to the same second service processing interface, while different target devices with the highest overlap degree are connected to different second service processing interfaces.

[0089] For example, in the first host partition, among five target devices V1 to V5, if the overlap degree of the running times of target devices V1 and V3 is relatively small, and the overlap degrees of the running times of target devices V2, V4, and V5 are all relatively large, then target devices V1 and V3 can be connected to the same second service processing interface I3_1, target device V2 is connected to a second service processing interface I3_2, target device V4 is connected to a second service processing interface I3_3, and target device V5 is connected to a second service processing interface I3_4.

[0090] In the second host partition, among five target devices V1 to V5, if the overlap degree of the running times of target devices V1 and V3 is relatively small, and the overlap degrees of the running times of target devices V2, V4, and V5 are all relatively large, then target devices V1 and V3 can be connected to the same second service processing interface I3_5, target device V2 is connected to a second service processing interface I3_6, target device V4 is connected to a second service processing interface I3_7, and target device V5 is connected to a second service processing interface I3_8.

[0091] In the third host partition, among five target devices V1 to V5, if the overlap degree of the running times of target devices V1 and V3 is relatively small, and the overlap degrees of the running times of target devices V2, V4, and V5 are all relatively large, then target devices V1 and V3 can be connected to the same second service processing interface I3_9, target device V2 is connected to a second service processing interface I3_10, target device V4 is connected to a second service processing interface I3_11, and target device V5 is connected to a second service processing interface I3_12.

[0092] As can be seen from the above examples, the second service processing interfaces corresponding to different host partitions are different. The first host partition, the second host partition, and the third host partition respectively correspond to a set of continuously distributed second service processing interfaces. That is to say, the second service processing interfaces I3_1 to I3_4 of the first host partition are continuously distributed and are the closest to the first host partition. The second service processing interfaces I3_5 to I3_8 of the second host partition are continuously distributed and are the closest to the second host partition. The second service processing interfaces I3_9 to I3_12 of the third host partition are continuously distributed and are the closest to the third host partition. However, within each host partition, the second service processing interfaces can be allocated according to the above time overlap degree. In this way, not only can the connection lines between different host partitions and the logical processing unit be avoided from crossing, but also the data transmission success rate can be increased within the host partition, and the data transmission delay can be shortened.

[0093] In some embodiments, within the host partition, the second service processing interfaces can also be allocated according to the management frequency of the management unit for the target device. For a target device with a high management frequency, an exclusive second service processing interface can be allocated to it. For a target device with a low management frequency, it can be connected to the same second service processing interface as the other target devices with a low management frequency. In this way, the second service processing interfaces can be utilized as fully as possible, and the situation where the blocking rate of some second service processing interfaces is very high while some other second service processing interfaces are idle can be avoided.

[0094] For example, within the first host partition, if among the five target devices V1 to V5, the management frequency of target device V1 is high, the management frequencies of target devices V2 and V3 are low, and the management frequencies of V4 and V5 are also high, then target device V1 and V2 can be connected to the same second service processing interface I3_1, target device V3 can be connected to a second service processing interface I3_2, target device V4 can be connected to a second service processing interface I3_3, and target device V5 can be connected to a second service processing interface I3_4.

[0095] Within the second host partition, if among the five target devices V1 to V5, the management frequency of target device V1 is high, the management frequencies of target devices V2 and V3 are low, and the management frequencies of V4 and V5 are also high, then target device V1 and V2 can be connected to the same second service processing interface I3_5, target device V3 can be connected to a second service processing interface I3_6, target device V4 can be connected to a second service processing interface I3_7, and target device V5 can be connected to a second service processing interface I3_8.

[0096] In the third host partition, among the five target devices V1 to V5, if the management frequency of target device V1 is relatively high, the management frequencies of target devices V2 and V3 are relatively low, and the management frequencies of V4 and V5 are also relatively high, then target device V1 and V2 can be connected to the same second service processing interface I3_9, target device V3 is connected to a second service processing interface I3_10, target device V4 is connected to a second service processing interface I3_11, and target device V5 is connected to a second service processing interface I3_12.

[0097] It can be seen that in the host partition, the target devices and the second service processing interfaces can be connected according to the above management frequencies. However, between host partitions, the second service processing interfaces are still allocated according to the positional relationship between the host partition and the second service processing interface. The second service processing interfaces closest to the above first host partition are second service processing interfaces I3_1 to I3_4; the second service processing interfaces closest to the above second host partition are second service processing interfaces I3_5 to I3_8; the second service processing interfaces closest to the above third host partition are second service processing interfaces I3_9 to I3_12. In this way, it is possible to avoid the connection lines between different host partitions and the logic processing unit from crossing as much as possible, and it is also possible to avoid the situation where the blocking rate of some second service processing interfaces is very high while some other second service processing interfaces are idle.

[0098] In some possible implementation manners, the number of management partitions is the same as the number of host partitions. One management partition is used to manage the target devices in one host partition, and the management partition is used to manage the host partition closest to the management partition; the management partition and the host partition corresponding to the management partition correspond to the same identifier. In this way, not only can the corresponding host partition be quickly found according to the management partition identifier, but it is not necessary to establish a partition mapping relationship between the management partition and the host partition and store it in the routing component. It can save storage resources and reduce the size of the routing component.

[0099] Of course, in actual applications, one management partition can also manage multiple host partitions.

[0100] Finally, it should be noted that the above first channel management interface, service management interface, first service processing interface, second channel management interface, and second service processing interface in the embodiments of the present application can be I2C (Inter-Integrated Circuit, integrated circuit bus) interfaces. Therefore, the target device can be called an I2C device.

[0101] The present application also provides a server, including the foregoing host partition management architecture, where the host partition management architecture includes: a host, a logical processing unit, and a management unit; the host includes multiple host partitions, and a target device is set in the host partition; the management unit includes multiple management partitions, and different management partitions operate independently.

[0102] The logical processing unit includes a first transmission component, a routing component, and a second transmission component; the first transmission component is connected to the routing component and the management partition, and the second transmission component is connected to the routing component and the host partition; when the management partition manages the target device in the host partition, the first transmission component and the second transmission component are used to transmit data, and the routing component is used to process the data transmitted by the first transmission component and the second transmission component.

[0103] An embodiment of the present application also provides a host partition management method, which is applied to a host partition management architecture, where the host partition management architecture includes: a host, a logical processing unit, and a management unit; the host includes multiple host partitions, the management unit includes multiple management partitions, and the logical processing unit includes a first transmission component, a routing component, and a second transmission component. Figure 5 This is a host partition management method provided by an embodiment of the present application, including:

[0104] S201: The management partition generates a first management command and sends the first management command to the routing component through the first transmission component.

[0105] Among them, the first management command is used to manage one or more target devices in the host partition corresponding to this management partition. Here, the management includes: reading relevant parameters of the target device, and / or writing relevant parameters of the target device. For example, reading the temperature of the network card, reading the temperature of the hard disk, reading the GPU acceleration card identifier, the GPU acceleration card manufacturer identifier, reading the sensor temperature, writing the network card operation parameters, hard disk operation parameters, GPU acceleration card operation parameters, sensor operation parameters. Here, the operation parameters may include, but are not limited to: restart parameters, operation duration, data acquisition frequency, etc.

[0106] Therefore, the foregoing first management command may carry a command identifier and a host partition identifier. Among them, the host partition identifier is used to indicate the host partition managed by the first management command, and the command identifier is used to indicate the target device of the operation and the specific management operation.

[0107] S202: The routing component generates a second management command according to the first management command and sends the second management command to the target device of the host partition through the second transmission component.

[0108] The first management command and the second management command include different contents. The second management command is a command that the target device can directly recognize and execute, and is of the same type as the first management command. For example, if the first management command is a read command, the second management command is also a read command. If the first management command is a write command, the second management command is also a write command.

[0109] It is understandable that the routing component needs to determine the target device to send based on the first management command. Therefore, the first management command needs to include the host partition identifier. However, when the routing component determines the target device for the first management command, it only needs to send the second management command to the target device. Therefore, the second management command no longer needs to carry the host partition identifier. The second management command only needs to instruct the target device to execute command information. The execution command information here can include the target device memory address to be accessed and the command type, which can be a read command or a write command. In this way, the target device can access the target device memory address based on the command type carried in the second management command.

[0110] For example, a read command reads data from the target device's memory address, while a write command writes data to the target device's memory address. A target device can correspond to one or more target device memory addresses to store different parameters related to the target device. For example, if the target device is a hard drive, target device memory address ADD1 can store the hard drive temperature, and target device memory address ADD2 can store the hard drive type.

[0111] S203: The target device executes the second management command.

[0112] The target device can read data from the target device memory address or write data to the target device memory address by executing the second management command.

[0113] The embodiments of the present application not only utilize a second transmission component to expand interfaces to meet the target device management requirements of multiple host partitions, but also utilize a routing component to implement the management process for different management partitions and corresponding host partitions. The routing component is used to route and parse data during the management process, ensuring that management commands from the management partition accurately reach the corresponding host partition, and that response data from the host partition accurately reaches the management partition. In this way, accurate management is achieved based on the shared logical processing unit between different management partitions and different host partitions. Furthermore, the shared logical processing unit can fully utilize the resources of the logical processing unit, minimizing resource waste and improving resource utilization.

[0114] In some possible implementations, the above method further includes:

[0115] S204: The target device sends the response data of the second management command to the routing component through the second transmission component.

[0116] In the embodiments of the present application, when the first management command and the second management command are read commands, the target device also needs to send the read data, that is, the response data, to the routing component.

[0117] Of course, in some scenarios, the response data can also be an identifier indicating whether the management command is executed successfully. The response data can be flexibly set according to the actual application scenario.

[0118] S205: The routing component generates a response message based on the response data, and sends the response message to the management partition through the first transmission component.

[0119] The response data is the actual data replied by the target device, which does not include the command information related to the first management command. Therefore, if the response data is directly returned to the management partition, the management partition cannot recognize the response data and thus cannot perform subsequent processing on the response data.

[0120] To enable the management partition to accurately recognize the response data and perform subsequent processing, the routing component needs to establish a mapping relationship between the response data and the first management command. Therefore, the routing component splices the response data and the command information used to uniquely identify the first management command into the response message. In this way, after receiving the response message, the management partition can determine the software interface for calling the first management command based on the command information of the first management command in it, and then perform subsequent processing on its response data. For example, the management partition calls the first management command in interface I1 to obtain the hard disk temperature. When the management partition receives the response message, it extracts the read hard disk temperature and the command information of the first management command from it, determines that the calling interface is I1 based on the command information of the first management command, and then continues to execute the subsequent processing code of interface I1 to determine whether to cool the hard disk according to the hard disk temperature.

[0121] In some possible implementation manners, the first transmission component includes a first service processing interface and a second channel management interface, and the management interface includes a service management interface and a first channel management interface. Among them, the service management interface is connected to the first service processing interface, and the service management interface is used to send the first management command for the target device and receive the response message. The first service processing interface is used to forward the first management command to the routing component, or forward the response message generated by the routing component to the service management interface. The first channel management interface is connected to the second channel management interface, and the first channel management interface and the second channel management interface are used to manage the connection channel between the service management interface and the first service processing interface.

[0122] Correspondingly, before the management partition sends the above first management command, S201 includes: the first channel management interface sends a first channel request command to the second channel management interface and receives a second channel identifier sent by the second channel management interface. The second channel identifier is used to indicate the service management interface used to transmit the first management command; the service management interface corresponding to the second channel identifier sends the first management command to the routing component.

[0123] Among them, the second channel identifier can be an idle interface or the interface with the smallest amount of data to be sent. Therefore, through the above channel allocation process, flexible allocation and multiplexing of the connection channels between the first service processing interface and the service management interface can be achieved, and the transmission success rate of the first management command is also guaranteed.

[0124] It can be understood that for the same management partition, multiple connection channels can be set between it and the routing component, and each connection channel is a channel formed by connecting a first service processing interface and a service management interface. Therefore, when the management partition needs to send the first management command, it can select a connection channel from multiple connection channels. In this way, channel multiplexing is achieved instead of exclusivity, which can effectively improve channel utilization. Therefore, the second channel identifier is used to uniquely represent a connection channel and the corresponding first service processing interface and service management interface of the connection channel.

[0125] In some possible implementation manners, sending the response message to the management partition through the first transmission component includes:

[0126] The second channel management interface sends a second channel request command to the first channel management interface and receives a first channel identifier sent by the first channel management interface. The first channel identifier is used to indicate the first service processing interface used to transmit the response message; the response message is sent to the management partition through the first service processing interface corresponding to the first channel identifier.

[0127] Among them, the first channel identifier can be an idle interface or an interface with a relatively small amount of data to be sent. Therefore, through the above channel allocation process, flexible allocation and multiplexing of the connection channels between the first service processing interface and the service management interface can be achieved, and the transmission success rate of the response message is also guaranteed.

[0128] Similarly, when the first transmission component needs to send a response message to the management partition, it can select a connection channel from multiple connection channels. Reusing a set of connection channels for sending the first management command and response data can effectively improve channel utilization. Therefore, the first channel identifier is used to uniquely represent a connection channel and the corresponding first service processing interface and service management interface of the connection channel.

[0129] The above-mentioned host partition management method of the present application is implemented based on the above management architecture. Under the action of the routing component, it can ensure that management commands are accurately sent to the target device on this management architecture, and the response data is accurately sent from the target device to the management partition.

[0130] Figure 6 It is a flowchart of the steps of another host partition management method provided by an embodiment of the present application. Refer to Figure 6 As shown, the above-mentioned host partition management method includes:

[0131] S301: The management partition generates a first management command.

[0132] S302: The first channel management interface sends a first channel request command to the second channel management interface.

[0133] S303: The second channel management interface sends a second channel identifier to the first channel management interface. The second channel identifier is used to indicate the service management interface used to transmit the first management command.

[0134] [[ID=S18]]S: The first management command is sent to the routing component through the service management interface corresponding to the second channel identifier.

[0135] S305: The routing component generates a second management command according to the first management command, and sends the second management command to the target device of the host partition through the second transmission component.

[0136] S306: The target device executes the second management command.

[0137] S307: The target device sends the response data of the second management command to the routing component through the second transmission component.

[0138] S308: The routing component generates a response message according to the response data.

[0139] S309: The second channel management interface sends a second channel request command to the first channel management interface.

[0140] S310: The first channel management interface sends a first channel identifier to the second channel management interface. The first channel identifier is used to indicate the first service processing interface used to transmit the response message.

[0141] S311: The response message is sent to the management partition through the first service processing interface corresponding to the first channel identifier.

[0142] The steps of S301 to S311 above can be specifically referred to the host partition method shown above, and will not be elaborated here. Figure 5 shown, and will not be elaborated here.

[0143] Through the description of the above embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be implemented by means of software plus a necessary general hardware platform. Of course, it can also be implemented by hardware, but in many cases the former is a better implementation method.

[0144] Those skilled in the art can further realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this article can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described according to functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0145] The above has introduced in detail a host partition management architecture, a server, and a method provided by this application. Specific examples are used in this article to elaborate on the principle and implementation manner of this application. The description of the above embodiments is only used to help understand the method and its core idea of this application. It should be noted that for those of ordinary skill in the art in this technical field, without departing from the principle of this application, several improvements and modifications can be made to this application, and these improvements and modifications also fall within the protection scope of the claims of this application.

Claims

1. A host partition management architecture, characterized in that Comprising: A host, a logic processing unit, and a management unit; The host includes multiple host partitions, and target devices are provided in the host partitions; The management unit includes multiple management partitions, and different management partitions operate independently; The logic processing unit includes a first transmission component, a routing component, and a second transmission component; The first transmission component is connected to the routing component and the management partitions, and the second transmission component is connected to the routing component and the host partitions; When the management partitions manage the target devices in the host partitions, the first transmission component and the second transmission component are used to transmit data, and the routing component is used to process the data transmitted by the first transmission component and the second transmission component.

2. The architecture according to claim 1, characterized in that, The first transmission component includes multiple groups of first transmission processing interfaces, the multiple groups of first transmission processing interfaces are connected to the routing component, one group of first transmission processing interfaces is connected to one management partition, different management partitions are connected to different first transmission processing interfaces, and the first transmission processing interfaces are used for data transmission between the routing component and the management partition corresponding to the first transmission processing interfaces.

3. The architecture according to claim 2, characterized in that The management partitions include multiple management interfaces, the multiple management interfaces are all connected to the first transmission component, the multiple management interfaces of the same management partition are connected to multiple continuously distributed positions on the first transmission component, and the first transmission component is used for data transmission between each management interface of each management partition and the routing component.

4. The architecture according to claim 3, characterized in that, The multiple management interfaces are connected to multiple first transmission processing interfaces, the multiple first transmission processing interfaces connected by the multiple management interfaces of the same management partition are adjacent, the first transmission processing interfaces connected by the multiple management interfaces of different management partitions are arranged in sequence, and the management interfaces are used for data interaction with the routing component through the first transmission processing interfaces.

5. The architecture according to claim 4, characterized in that, The multiple management interfaces include multiple service management interfaces, and different service management interfaces operate independently; One service management interface is connected to one first transmission processing interface, and the service management interface is used to send a first management command for the target device to the first transmission processing interface, and / or receive a response message for the first management command from the first transmission processing interface.

6. The architecture according to claim 5, characterized in that, The multiple management interfaces further include a first channel management interface. For one management partition, the first transmission processing interface connected by the first channel management interface is located before or after the first transmission processing interface connected by the service management interface, and the first channel management interface is used to manage the connection channel from the first transmission processing interface to the service management interface.

7. The architecture according to claim 6, characterized in that, The multiple first transmission processing interfaces include multiple first service processing interfaces, the multiple first service processing interfaces are connected to the routing component, one first service processing interface is connected to one service management interface of one management partition, and the first service processing interface is used to transmit the first management command and / or the response message.

8. The architecture according to claim 7, wherein The multiple first transmission processing interfaces further include a second channel management interface, which is connected to the routing component and the first channel management interface, and is used to manage the connection channel from the service management interface to the first transmission processing interface through the routing component.

9. The architecture according to claim 8, characterized in that, The second channel management interface is connected to the first channel management interfaces of one or more management partitions.

10. The architecture according to claim 1, characterized in that, The second transmission component includes multiple second service processing interfaces, all of which are connected to the routing component. One second service processing interface is connected to one or more of the target devices, and the second service processing interface is used to send the second management command sent by the routing component to the connected target device, or send the response data of the target device for the second management command to the routing component.

11. The architecture according to claim 10, characterized in that, One second service processing interface is connected to one or more target devices in the same host partition. The target devices in the same host partition are connected to multiple second service processing interfaces distributed continuously, and the target devices in the host partition are connected to the nearest and available second service processing interface to the target device.

12. The architecture according to claim 11, wherein One second service processing interface is connected to one or more target devices of the same type in the same host partition.

13. The architecture according to claim 11, wherein One second service processing interface is connected to one or more adjacent target devices in the same host partition.

14. The architecture according to claim 11, characterized in that, One second service processing interface is connected to one or more target devices with different running times in the same host partition.

15. The architecture according to claim 1, characterized in that, The number of the management partitions is the same as that of the host partitions. One management partition is used to manage the target devices in one host partition, and the management partition is used to manage the host partition closest to the management partition; The management partition and the corresponding host partition have the same identifier.

16. A server, characterized in that, Including the host partition management architecture according to any one of claims 1 to 15.

17. A host partition management method, characterized in that Applied to a host partition management architecture, the host partition management architecture includes: a host, a logic processing unit, and a management unit; the host includes multiple host partitions, the management unit includes multiple management partitions, and the logic processing unit includes a first transmission component, a routing component, and a second transmission component; the method includes: The management partition generates a first management command and sends the first management command to the routing component through the first transmission component; The routing component generates a second management command according to the first management command and sends the second management command to the target device of the host partition through the second transmission component; The target device executes the second management command.

18. The method according to claim 17, wherein It further includes: The target device sends the response data of the second management command to the routing component through the second transmission component; The routing component generates a response message according to the response data and sends the response message to the management partition through the first transmission component.

19. The method according to claim 18, characterized in that, The first transmission component includes a first service processing interface and a second channel management interface, and the management interface includes a service management interface and a first channel management interface; Sending the first management command to the routing component through the first transmission component includes: The first channel management interface sends a first channel request command to the second channel management interface and receives a second channel identifier sent by the second channel management interface, where the second channel identifier is used to indicate the service management interface for transmitting the first management command; The first management command is sent to the routing component through the service management interface corresponding to the second channel identifier.

20. The method according to claim 19, characterized in that, Sending the response message to the management partition through the first transmission component includes: The second channel management interface sends a second channel request command to the first channel management interface and receives a first channel identifier sent by the first channel management interface, where the first channel identifier is used to indicate the first service processing interface for transmitting the response message; The response message is sent to the management partition through the first service processing interface corresponding to the first channel identifier.

Images