Single board management system, method, device and equipment

Abstract

The invention discloses a single board management system, method, device and equipment. A substrate management controller can be connected with a computing equipment single board through a management bus. The computing device single board comprises a memory and a device manager, and management information of the computing device single board is recorded in the memory. In the computing device single board, the memory and the device manager can be connected with the substrate management controller through a management bus. And the substrate management controller acquires the management information from the memory through the management bus, and interacts with the device manager to manage the computing equipment single board based on the management information. The connection relation between the substrate management controller and the computing device is simple, the substrate management controller is adaptive to the single boards of the computing devices of different structures, and the management mode of the single boards of the computing devices can be effectively simplified. And the management mode of the computing equipment single board is more efficient.

Description

[0001] This application is a divisional application. The application number of the original application is 202210188470.X, and the original application date is February 28, 2022. The entire contents of the original application are incorporated into this application by reference. Technical Field

[0002] The present application relates to the field of server technology, and in particular to a single board management system, method, device and equipment. Background Art

[0003] Since the 1980s, Microsoft and Intel have formed the Wintel Alliance to promote the development of the personal computer (PC) industry. The two companies have worked closely in the PC industry to drive faster development of the computing industry and gradually affect other computing devices such as servers. Computing devices such as servers have many application scenarios, many configuration types, and relatively high reliability requirements. At the same time, servers have a huge commercial volume and are the focus of open industry ecosystem construction.

[0004] Taking traditional servers as an example, the current industrial ecology of traditional servers has the following characteristics:

[0005] Low degree of standardization: Traditional servers already have a certain basis for component standardization. For example, memory sticks, solid-state drives (SSDs), and peripheral component interconnect express (PCIe) cards have their own standards. Component standardization has made great contributions to the industry ecology and resource sharing, reducing the development work of some server manufacturers. However, standardized components account for a small proportion of the entire server, which means that the development of server motherboards still requires more manpower to complete the adaptation of standardized and non-standardized components.

[0006] To this end, it is necessary to enhance the standardization of server components. The standardization of server components involves all aspects of the server. Out-of-band management of servers refers to the maintenance of servers and other equipment through independent management channels. Out-of-band management of servers allows system administrators to remotely monitor and manage servers. Out-of-band management of servers mainly involves the management and monitoring of the working environment of devices on server boards (such as processors, memory, and hard disks). The working environment of the devices includes but is not limited to information such as temperature, operating voltage, fans, and power supply status, to ensure that the server devices can work in a suitable working environment.

[0007] Out-of-band management of servers is usually implemented by a baseboard management controller (BMC). In order to implement out-of-band management, the baseboard management controller needs to be connected to the server board to connect to the various devices on the server board. However, due to the increasing types and varieties of processors, the architectures of server boards deployed with different processors are also different. The interfaces connected to the baseboard management controller in these different server boards are not unified. This results in the baseboard management controller needing to perform a lot of adaptation work for the out-of-band management of each server board, and the reuse rate of baseboard management controllers of different types of server boards is low. Summary of the invention

[0008] The present application provides a single board management system, method, apparatus and device, which are used to provide a more adaptable out-of-band management BMC and method.

[0009] In a first aspect, an embodiment of the present application provides a single board management system, which includes a baseboard management controller and a computing device single board. The single board management system can be deployed in a computing device, which can be a server, a personal computer, etc.

[0010] The baseboard management controller can be connected to the computing device board through a management bus. The computing device board includes a memory and a device manager, and the memory records the management information of the computing device board. Inside the computing device board, the memory and the device manager can be connected to the baseboard management controller through a management bus.

[0011] The baseboard management controller can obtain management information from the memory through the management bus, and manage the computing device board in interaction with the device manager based on the management information.

[0012] Through the above system, the connection relationship between the baseboard management controller and the computing device is simple, and the single board of the computing device with different structures can be adapted, which can effectively simplify the management method of the single board of the computing device. The management method of the single board of the computing device is also more efficient.

[0013] In a possible implementation, the computing device board further includes a first type of device, the first type of device is connected to a device manager, and the device manager can obtain working information of the first type of device. The baseboard management controller can obtain working information of the first type of device from the device manager via a management bus.

[0014] Through the above system, the baseboard management controller can easily obtain the working information of the first category devices through the device manager without connecting to the first category devices. The method of obtaining the working information of the first category devices is simple and efficient, which avoids the large amount of adaptation work required by the baseboard management controller to match different computing device boards and simplifies the out-of-band management process.

[0015] In one possible implementation, the computing device board also includes a second type of device, which may be connected to the baseboard management controller without going through a device manager, and may be directly connected to the baseboard management controller through a management bus; the baseboard management controller may directly interact with the second type of device through the management bus to obtain working information of the second type of device.

[0016] Through the above system, the management bus can not only hang the memory and device manager, but also hang the second type of devices. This connection method is relatively simple, and the baseboard management controller also does not need to perform too much adaptation work, which effectively expands the application scenarios.

[0017] In a possible implementation, the management information is the information required by the baseboard management controller to manage the computing device single board. That is, the management information can be pre-stored in the memory. The specific content of the management information is not limited in the embodiment of the present application, and all the information required to manage the computing device single board is applicable to the embodiment of the present application. For example, the management information includes part or all of the following: attribute information of the computing device single board, topology information of the computing device single board, attribute information of the first type of device, and attribute information of the second type of device.

[0018] Through the above system, the management information is pre-stored in the memory, and the baseboard management controller only needs to perform a simple loading operation to obtain the management information, and the method of obtaining the management information is simpler.

[0019] In a possible implementation, the baseboard management controller can interact with the device manager, and the embodiment of the present application does not limit the interaction method between the baseboard management controller and the device manager. For example, the baseboard management controller can interact with the device manager based on command words, which can ensure efficient interaction efficiency. Different computing device boards can set a common command word method. In this way, the baseboard management controller can be adapted to different computing device boards, improving the adaptability of the baseboard management controller and the management method.

[0020] In a possible implementation, the baseboard management controller can control the first type of device. For example, the baseboard management controller can send a control command to the device manager to instruct the device manager to control the first type of device. The baseboard management controller can directly control the first type of device, and the baseboard management controller can send a control command to the second type of device through the management bus to control the second type of device.

[0021] The baseboard management controller upgrades the first-category device or the device manager. For example, the baseboard management controller can pass the upgrade file of the first-category device to the device manager, instructing the first-category device to be upgraded. After obtaining the upgrade file of the first-category device, the device manager uses the upgrade file of the first-category device to upgrade the first-category device. The baseboard management controller can also pass the upgrade file of the device manager to the device manager, instructing the device manager to be upgraded. Of course, the baseboard management controller can also directly upgrade the second-category device through the management bus.

[0022] Through the above system, the baseboard management controller controls or upgrades the device through the management bus or the device manager, which simplifies the control and upgrade methods and ensures that the single board of the computing device can be managed efficiently.

[0023] In a possible implementation, the embodiment of the present application does not limit the type of memory. For example, the memory may be an electrically erasable programmable read-only memory, which has a small size and a higher degree of integration.

[0024] In a possible implementation, the embodiment of the present application does not limit the specific structure of the device manager, and any module that can implement device management is applicable to the embodiment of the present application. For example, the device manager is a complex programmable logic device or a microcontroller unit. The specific structure of the device manager is diverse and applicable to different computing device boards, effectively expanding the application scenarios.

[0025] In a possible implementation, the management bus may be an internal integrated circuit bus or a serial peripheral interface bus, or may be other types of buses. The type of management bus is relatively flexible, so that the baseboard management controller can be connected to different types of computing device boards through the management bus, thereby improving the adaptability of the baseboard management controller to different types of computing devices.

[0026] In the second aspect, the embodiment of the present application provides a single board management method, which is used to manage a computing device single board. The beneficial effects can be found in the description of the first aspect and will not be repeated here. The computing device single board includes a memory and a device manager, and the memory records the management information of the computing device single board. In this method, the baseboard management controller can obtain the management information from the memory through the management bus. After obtaining the management information, the baseboard management controller can interact with the device manager through the management bus based on the management information to manage the computing device single board.

[0027] In a possible implementation, the computing board includes a first type of device, the first type of device may be connected to a device management device, and the baseboard management controller may obtain working information of the first type of device from the device manager through a management bus.

[0028] In a possible implementation, the computing board includes second-category devices, the first-category devices may be directly connected to the baseboard management controller via a management bus, and the baseboard management controller may obtain working information of the second-category devices from the second-category devices via the management bus.

[0029] In a possible implementation, the management information includes part or all of the following: attribute information of a computing device board, topology information of a computing device board, attribute information of first-category devices, and attribute information of second-category devices.

[0030] In a possible implementation, when the baseboard management controller interacts with the device manager through the management bus, the baseboard management controller may interact with the device manager through the management bus in a manner based on command words.

[0031] In a possible implementation, the baseboard management controller controls the first-category device through the device manager, and can also upgrade the first-category device through the device manager. For example, the baseboard management controller transmits the upgrade file of the first-category device to the device manager, indicating that the first-category device is to be upgraded. After receiving the upgrade file of the first-category device, the device manager can upgrade the first-category device using the upgrade file of the first-category device. The baseboard management controller can also upgrade or control the device management controller or the second-category device.

[0032] In a possible implementation, the management bus is an I2C bus or an SPI bus.

[0033] In the third aspect, the embodiment of the present application further provides a single board management device, which has the function of implementing the behavior in the method example of the second aspect above. The beneficial effects can be found in the description of the first aspect and will not be repeated here. The functions can be implemented by hardware, or by hardware executing corresponding software implementations. The hardware or software includes one or more modules corresponding to the above functions. In one possible design, the structure of the single board management device includes a request acquisition unit, a management unit, and optionally, an upgrade unit. These units can execute the corresponding functions in the method example of the second aspect above. Please refer to the detailed description in the method example for details, which will not be repeated here.

[0034] In a fourth aspect, an embodiment of the present application further provides a baseboard management controller, which has the function of implementing the behavior in the method example of the second aspect above. The beneficial effects can be found in the description of the second aspect and will not be repeated here. The structure of the device includes a processor. Optionally, it may also include a memory. The processor is configured to support the single board management device to execute the corresponding method in the method of the second aspect above. Optionally, the baseboard management controller may also include a memory. The memory is coupled to the processor, which stores the computer program instructions necessary for the communication device. The processor can call the computer program instructions to execute the corresponding method in the method of the second aspect above.

[0035] In a fifth aspect, an embodiment of the present application further provides a computing device, which includes a baseboard management controller and a computing device single board, and the computing device single board may include devices such as a processor and a memory. The baseboard management controller has the function of implementing the behavior in the method example of the second aspect above, and the beneficial effects can be referred to the description of the first aspect and will not be repeated here.

[0036] In a sixth aspect, the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the method described in the above-mentioned second aspect and each possible implementation manner of the second aspect.

[0037] In a seventh aspect, the present application further provides a computer program product comprising instructions, which, when executed on a computer, enables the computer to execute the method described in the above-mentioned first aspect and various possible implementations of the first aspect.

[0038] In an eighth aspect, the present application further provides a computer chip, wherein the chip is connected to a memory, and the chip is used to read and execute a software program stored in the memory, and to execute the method described in the above second aspect and each possible implementation manner of the second aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1A schematic diagram of the architecture of a single board management system provided for this application;

[0040] Figure 2 A schematic diagram of the architecture of another single board management system provided for this application;

[0041] Figure 3 A schematic diagram of the structure of a command word provided for this application;

[0042] Figure 4A A schematic diagram of the structure of a read request provided by this application;

[0043] Figure 4B A schematic diagram of a read response structure provided by this application;

[0044] Figure 4C A schematic diagram of a write request structure provided by this application;

[0045] Figure 5A to Figure 5B A schematic diagram of the structure of an expansion board provided for this application;

[0046] Figure 6A to Figure 6C A schematic diagram of the architecture of a single board management system provided for this application;

[0047] Figure 7 A schematic diagram of a single board management method provided in this application;

[0048] Figure 8 A schematic diagram of the structure of a BCU management system provided in this application;

[0049] Fig. 9 A schematic diagram of the structure of a single board management device provided in this application;

[0050] Fig.10 A schematic diagram of the structure of a computing device provided in this application. DETAILED DESCRIPTION

[0051] The technical threshold for the development of traditional server motherboards is high. In addition to the central processing unit (CPU), it also includes bus fan-out, power fan-out, maintenance management and other functions. The CPU-related circuits on these motherboards are all from the reference design provided by the CPU manufacturer, and the reference designs provided by different CPU manufacturers are completely different, which requires a lot of resources and time to develop and design the motherboard. In order to meet the needs of rapid updates and replacements of computing products such as servers, whole machine manufacturers need to invest more energy in differentiated innovation, but they can only focus on low-level hardware specification competition. This can not only fail to meet the diverse scenarios and computing power needs of customers, but also force whole machine manufacturers to fall into inefficient homogeneous competition. With the trend of computing power diversity, more processor manufacturers have emerged and launched more processor products with different architectures, and the iteration speed of various processors has also increased rapidly. At the same time, the power consumption of processors has continued to increase, and the heat dissipation technology of traditional servers cannot meet the needs. In addition, in order to improve system performance, the industry has also introduced new media types (for example, Intel launched 3D Xpoint, a new non-volatile medium, etc.) and forms, and the above new media types and forms also require new architecture support and adaptation. In order to develop servers that adapt to the above-mentioned technology trends, machine manufacturers need to invest a huge amount of development work, but due to the differences between different products, the design of the same motherboard or machine cannot be reused. Therefore, the entire industry has put forward higher requirements for cross-architecture shared components, cross-generation evolution, shortened time to market (TTM), and reduced total cost of operation (TCO). The further development of the industry requires the construction of a more open and standardized server architecture, improved development efficiency, increased component reuse, and more flexibility and differentiation.

[0052] This application proposes an innovative peer-to-peer interconnection architecture (also referred to as a new server architecture or new architecture). In this architecture, the traditional motherboard is first split into a basic board (Basic Computing Unit, BCU) and an expansion board (Extension Unit, EXU), and the basic board is used in conjunction with the expansion board to support the specifications and forms of the motherboard required for different scenarios. The same computing device may include a basic board and an expansion board, the same computing device may also include multiple basic boards and an expansion board, and the same computing device may also include a basic board and multiple expansion boards. The basic board includes a CPU, a double data rate (double data rate, DDR) and related power supplies, providing general computing capabilities and peripheral storage, input / output (input / output, IO), acceleration and other expansion interfaces. The basic board supports Optionally, the base board supports heterogeneous processors, that is, the base board can support different types of processors, for example, the base board supports CPU, and any processor or any combination thereof, such as application-specific integrated circuit (ASIC), programmable logic device (PLD), complex programmable logical device (CPLD), field-programmable gate array (FPGA), generic array logic (GAL), system on chip (SoC), software-defined infrastructure (SDI) chip, artificial intelligence (AI) chip, etc.

[0053] Furthermore, according to business requirements and hardware attributes, the present application embodiment provides at least 6 different forms of basic boards, respectively for different computing performance and memory configurations. For the convenience of description, these 6 basic boards are referred to as A1, A2, B1, B2, C1, and C2. In addition, in this embodiment, "P" is used to represent the number of processors, P is an integer greater than 0, and "DPC" represents dual in-line memory module per channel (DIMM Per Channel). For example, the A1 form factor base board supports one processor, with one DIMM inserted in each channel (abbreviated as 1P1DPC); the A2 form factor base board supports one processor, with one or two DIMMs inserted in each channel (abbreviated as 1P2DPC or 1P1DPC); the B1 form factor base board supports two processors, with one DIMM inserted in each channel (abbreviated as 2P1DPC), or one processor, with one or two DIMMs inserted in each channel (abbreviated as 1P2DPC or 1P1DPC); the B2 form factor base board supports two processors, with one or two DIMMs inserted in each channel (abbreviated as 2P2DPC or 2P1D The C1 form factor base board supports four processors, one DIMM per channel (referred to as 4P1DPC), or two processors, one or two DIMMs per channel (referred to as 2P2DPC or 2P1DPC); the C2 form factor base board supports four processors, one or two DIMMs per channel (referred to as 4P2DPC or 4P1DPC), or two processors, one or two DIMMs per channel (referred to as 2P2DPC or 2P1DPC). With the development of technology, the CPU package size, memory channels and number of DIMMs may change, but the standard size and mounting holes of the motherboard will remain unchanged, so as to ensure that the base board can be compatible and evolved across generations and series when it is updated. For example: the B2 form factor base board supports 2P2DPC (2P32DIMM) when there are currently 8 channels of DDR per CPU. After the number of CPU memory channels is increased to 12, 2P2DPC (2P48DIMM) will not be possible. Then, the B2 form factor can support 2P1DPC (2P24DIMM), while 2P2DPC (2P48DIMM) can be achieved with other forms such as C1, because the mounting hole positions and base board dimensions are standard, and can be directly replaced and installed.

[0054] The expansion board includes a baseboard management controller (BMC) chip, which is a management extension of the base board. As the management center of the entire system, it provides management functions such as equipment, security, energy efficiency, and reliability. Among them, BMC can also be called a baseboard management controller. Optionally, the expansion board can also include a management system and a bridge chip (for example, the platform controller hub (PCH) of the Intel system).

[0055] In the new architecture, the basic board is connected to the components through high-speed buses such as PCIe, Compute Express Link (CXL), or unified bus (UB or Ubus), and is connected to the expansion board through a management interface. In the specific implementation, the specific connection methods between the above-mentioned basic board and components, and between the basic board and the expansion board include: a soft connection method using a cable to achieve the above-mentioned connection, or a hard connection method using a connector to achieve the above-mentioned connection. Furthermore, the component is a general term for a class of electronic devices or electronic devices. Among them, the components include storage components (STorage Unit, STU), IO components (Input Output Unit, IOU), acceleration components (ACceleration Unit, ACU), memory expansion components (Memory Expansion Unit, MEU), heat dissipation components, computing components, management components, etc. according to their functions. The basic board supports Kunpeng, The expansion board provides management functions and power supply for the basic board and various expansion components. With the support of the expansion board, there are various options for power supply and heat sink.

[0056] Among them, storage components include hard disk backplane, expansion board (Expander), PCIe switch (switch), etc., which are used to expand system storage and support multiple media and forms such as hard disk drive (HDD) / solid-state drive (SSD) / non-volatile high-speed transmission bus (Non-Volatile Memory express, NVMe) / storage class memory (Storage Class Memory, SCM) and so on.

[0057] IO components include Riser and other components to expand the system IO and support PCIe standard cards and Open Compute Project (OCP) cards.

[0058] Acceleration components include risers, carrier boards, accelerator card interconnection switches, etc., providing system acceleration component expansion and interconnection functions.

[0059] Memory expansion components include carrier boards, memory expansion chips, dual in-line memory modules (DIMMs), SCM media, etc., which provide the system with the function of expanding memory bandwidth and memory capacity.

[0060] The heat dissipation component is used to dissipate heat from a computing device or hardware in the computing device, including a combination of several heat dissipation methods such as air cooling, liquid cooling, or a combination of the two. It should be understood that the structure, type, and number of the heat dissipation components do not constitute a limitation on the technical solution to be protected by this application.

[0061] Computing components, such as central processing unit (CPU), memory, etc., provide general computing capabilities.

[0062] Management components, baseboard management controllers, etc. provide device management devices.

[0063] It is worth noting that a basic board or an expansion board including a processor, a memory, and a baseboard management controller can also be considered as a component.

[0064] On the other hand, in the traditional server architecture, due to the evolution of power supply, number of memory channels, number of IOs, rate, etc., the socket of the processor (for example, CPU) can generally only be compatible with each generation (two small upgrades of Tick / Tock), and it is difficult to be compatible across generations. The motherboard provided in this application can be set up with external interfaces in a standardized manner, and various external expansions can be performed in a soft connection manner such as cables, which can shield the differences caused by the power supply related to the processor, the interconnection between different processors and components, and the components. The changes in components such as memory are only contained inside the motherboard, realizing the function of cross-generation compatibility of the motherboard. In this way, for each manufacturer, when the processor is updated, the supporting whole machine, components, etc. can not be replaced, so the supporting components have a longer life cycle. For customers, without changing the chassis and increasing the workload of hardware development, the latest components can be replaced at any time, and the latest computing power in the industry can be used as quickly as possible. For the whole machine manufacturer, after the cross-generation upgrade and cross-series evolution of the new server architecture is realized, the upgrade of the processor or the replacement of different processor manufacturers only needs to simply replace the basic board, which subverts the original development model and derives a new industrial model.

[0065] In addition to providing a new server architecture, this embodiment also implements hardware standardization, including standardization of basic boards and component interfaces, in order to support diverse computing power and diverse devices.

[0066] The standardization of the basic board includes the standardization of size, installation hole position, interface electrical characteristics, management interface protocol and parameters, etc. Table 1 is an example of a basic board interface description table provided in this application.

[0067] Table 1

[0068]

[0069]

[0070] Among them, the power supply adopts a unified 12V input, and the basic board is converted into various required power supplies through DC / DC. Taking into account the future evolution of I / O and the differentiation of different CPUs, this embodiment defines a Flexible I / O interface based on UBC and UBCDD connectors to replace the original PCIe interface. The Flexible I / O interface can be flexibly configured into PCIE / HCCS / SAS / SATA / Ethernet and other interfaces according to needs. The BCU management interface mainly includes common low-speed maintenance interfaces, such as I2C, UART, JTAG and other interfaces, which are compatible with the management of common processor platforms.

[0071] Standardization of the internal component interfaces of computing systems: Components include expansion boards, power supply components, heat dissipation components, storage components, IO components, acceleration components, memory components, etc. The electrical interfaces, management interfaces and parameters of the components are standardized without defining and constraining the physical size, installation, location, etc. of the components. This will provide a broad space for innovation and support differentiation and flexible expansion. In addition to power and high-speed signals, the other low-speed management interfaces of the components are defined as shown in Table 2 below:

[0072] Table 2

[0073]

[0074] In addition to the interface between EXU and BCU, other interfaces are connected to various components through EXU. It is worth noting that this embodiment only defines the functions of these interfaces, and does not limit the specific pin layout (PINMAP) method. Any implementation method that can achieve this function is within the protection scope of this embodiment.

[0075] It is worth noting that the contents of Tables 1 to 2 above are only examples provided to assist in explaining the technical solution of the present application. In specific implementations, the new server architecture, the interface of the basic board, and the low-speed interface of the functional component may respectively include more or less content.

[0076] In addition, the present application also provides a method for implementing intelligent management software, which implements management object templates according to the standardized requirements of computing devices. After the server is powered on, the management software automatically detects components through a standard management bus and obtains the self-description information of the components, and then creates management object instances according to the management object templates, thereby realizing adaptive management of the management software and realizing intelligent management software, and supporting automatic discovery and automatic adaptation of components.

[0077] For ease of description, the following embodiments are described using a computing device as a server as an example. The solution provided in this application is also applicable to other computing devices such as edge servers and personal computers (PCs).

[0078] For servers, since servers need to carry a large number of services and perform a large amount of data calculations, this requires that more components be deployed in the server, more processors need to be deployed on the basic board, larger memory, and more hard disks need to be connected. The working status of the processor, memory, hard disk, and various components on the basic board as the main components of the server determines the operating status of the server. In order to ensure that the main components of the server can work normally, the server is also equipped with temperature sensors (to measure the temperature of the device), voltage sensors (to measure the working voltage of the device), different types of power supplies (to provide voltages of different volts), fans (to cool the device) and other devices. An important part of the out-of-band management of the server is the monitoring and management of the working environment of the main components in the server to ensure that the main components in the server can work in a suitable working environment, such as the temperature is within the working temperature range of the device, the voltage meets the working voltage of the device, the power supply of different types of power supplies is normal, and the fan operates normally.

[0079] Due to the different structures of different server boards, there is no unified interface to the outside to achieve out-of-band management, resulting in the baseboard management controller to achieve out-of-band management of any server board requires a lot of adaptation work, and the flexibility is poor. To this end, the embodiment of the present application provides a single board management system, method, device and equipment. In the embodiment of the present application, the baseboard management controller can be connected to the server board through a management bus through a unified interface. That is, only one management bus is needed between the baseboard management controller and the server board to achieve connection. The baseboard management controller can obtain the management information required for managing the server board from the memory deployed on the server board through the management bus, and can also interact with the device manager on the server board through the management bus to manage the server board. The baseboard management controller can obtain the working information of the device connected to the device manager on the server board, thereby achieving out-of-band management. In this way, the baseboard management controller can achieve out-of-band management for different server boards without having to perform a lot of adaptation work, simplifying the entire process of the baseboard management controller to achieve out-of-band management.

[0080] like Figure 1 As shown, it is a structural diagram of a single board management system provided in an embodiment of the present application. The single board management system can be deployed in a server, and the single board management system includes a server single board 100 and a baseboard management controller 200.

[0081] It should be noted that the server single board 100 can be the basic board in the new architecture described above, or any component. It can also be the mainboard in a traditional server. The embodiment of the present application does not limit the number of server single boards 100, which can be one or more. When a single board management system includes multiple server single boards 100, the multiple server single boards 100 can be single boards of the same type. For example, the multiple server single boards 100 are all basic boards, and the multiple server single boards 100 can also be single boards of different types, such as the multiple server single boards 100 include a basic board, an IO component and a storage component.

[0082] In an embodiment of the present application, the baseboard management controller 200 is connected to the server single board 100 through a management bus 300. The management bus 300 can be an I2C (Inter-Integrated Circuit) bus or a serial peripheral interface (SPI) bus. The management bus 300 can also be other types of buses. The management bus 300 can be understood as a root management bus 300 for single board management, and the root management bus 300 can be used as a root management link. The baseboard management controller 200 can obtain management information and working information of devices on the server single board 100 through the root management link to achieve management of the server single board 100.

[0083] The server board 100 is equipped with server devices, including but not limited to: processor, memory, temperature sensor, analog to digital converter (ADC), power interface, peripheral component interconnect express (PCIe) slot, hard disk interface, fan, power supply, etc. Different types of server boards 100 may have different types and quantities of devices deployed on the server boards 100.

[0084] It should be noted that the embodiment of the present application does not limit the way in which the components are deployed on the server board 100. For example, the server components can be directly soldered on the server board 100. For another example, the server components can be connected to the server board 100 through an interface (such as a high-speed interface such as UBC). In actual applications, some components (such as IO components, storage components, etc.) can be connected to the base board through an interface. In this case, these components can also be considered as components deployed on the server board 100.

[0085] In order to implement out-of-band management of the server board 100, a device manager 120 and a memory 110 are also deployed on the server board 100. The baseboard management controller 200 is connected to the device manager 120 and the memory 110 through the management bus 300 respectively.

[0086] The device manager 120 can be connected to some or all of the devices on the server board 100. In the embodiment of the present application, the devices on the server board 100 include two categories, one category is the devices that establish a connection with the baseboard management controller 200 through the device manager 120. For the convenience of explanation, this type of device is called the first category of devices. The other category is the devices that are directly connected to the baseboard management controller 200 through the management bus 300. For the convenience of explanation, this type of device is called the second category of devices. For any server board 100, the devices involved in out-of-band management (the devices involved in out-of-band management refer to the devices that affect the working environment of the main devices on the server board 100) can all belong to the first category of devices, that is, the devices on the server board 100 are all connected to the device manager 120. Figure 1 The figure is drawn based on the example that all the devices on the server board 100 belong to the first type of devices. The devices involved in out-of-band management may also include the first type of devices and the second type of devices. The second type of devices are not connected to the device manager 120, but can be connected to the baseboard management controller 200 through the management bus 300. The connection method between this type of server board 100 and the baseboard management controller 200 can be referred to in the subsequent Figure 2 See the relevant instructions in .

[0087] The device manager 120 can interact with the first type of device to obtain the working information of the first type of device. For example, the device manager 120 can obtain the temperature of the temperature sensor, the voltage value of the voltage sensor, whether the power interface is connected to a power supply, the voltage provided by the power supply (the power supply voltage is obtained by connecting to an ADC, and the ADC converts analog signals such as voltage into data signals), whether a component of the PCIe interface (such as an accelerator card, etc.) is inserted into the PCIe slot, whether the hard disk interface is connected to a hard disk, whether the fan is running, etc.

[0088] The memory 110 stores the management information of the server board 100, which is necessary information for the baseboard management controller 200 to implement out-of-band management. The management information of the server board 100 will be described below.

[0089] The baseboard management controller 200 connects the device manager 120 and the memory 110 via the management bus 300. The baseboard management controller 200 can obtain the management information of the server board 100 from the memory 110, and understand the attributes of the service board, the device attributes, and the topology information of the server board 100. The baseboard management controller 200 can also obtain the working information of the first type of device by interacting with the device manager 120. The server board 100 is managed based on the management information (and the working information of the first type of device).

[0090] like Figure 2As shown, it is a structural schematic diagram of another single board management system provided by an embodiment of the present application. The single board management system can be deployed in a server, and the single board management system includes a server single board 100 and a baseboard management controller 200. The baseboard management controller 200 and the server single board 100 only need to be connected through one management bus 300, and the baseboard management controller 200 is connected to the device manager 120, the memory 110, and the second type of devices through one management bus 300. The device manager 120 is connected to the first type of devices on the server single board 100. For the description of the baseboard management control, the device manager 120, the memory 110, the management bus 300, the first type of devices and the second type of devices, please refer to the above content, the difference is Figure 1 The single board management system, Figure 2 In the single board management system shown, the devices on the server single board 100 may also be directly connected to the baseboard management controller 200 via the management bus 300 and directly managed by the baseboard management controller 200 .

[0091] In this single board management system, only one management bus 300 is needed to connect the baseboard management controller 200 and the server single board 100. The connection method between the baseboard management controller 200 and the server single board 100 is simple, and the baseboard management controller 200 is also applicable to different server single boards 100.

[0092] The following is a description of the various components in the board management system:

[0093] (1) Memory 110.

[0094] In the embodiment of the present application, the memory 110 is used to store the management information of the server board 100 required for out-of-band management. The embodiment of the present application does not limit the type of the memory 110. The memory 110 can be an electrically erasable programmable read only memory (EEPROM) or other non-volatile memory. The memory 110 is a field replaceable unit description (FRUD), and the management information required to manage the server board 100 is stored in the FRUD. The management information includes the attribute information of the server board 100, the information of the device to be managed, the topology information, the alarm information, etc.

[0095] The management information includes attribute information of the server board 100 , topology information of the server board 100 , and attribute information of devices.

[0096] Among them, the attribute information of the server board 100 is used to describe the hardware information of the server board 100, and the attribute information of the server board includes but is not limited to: board type, board identification (ID), printed circuit board (PCB) version number of the board, and bill of material (BOM) version number of the board.

[0097] After acquiring the attribute information of the server board 100 , the baseboard management controller 200 can understand the basic information of the server board 100 .

[0098] The topology information of the server board 100 describes the connection relationship of the devices on the server board 100 . The topology information of the server board 100 may include an in-band management topology and a management bus topology.

[0099] The in-band management topology can also be called the service bus topology. The in-band management topology describes the topology information of the service plane of the server board 100, that is, the connection relationship between the devices (processor, hard disk, memory) that carry the server service in the server board 100, including but not limited to: the connection management of the devices on the basic board, the connection relationship between the basic board and the components, the connection management between the components, etc. The in-band management topology includes but is not limited to: component signals, processor information (such as port number, type, quantity, bit width, etc.), memory information (such as port number, type, quantity, bit width, etc.), hard disk information (such as hard disk interface, type, quantity, bit width, etc.), and the connection method between the processor, memory, and hard disk. All information related to the devices on the service plane can be recorded in the in-band management topology.

[0100] The management bus topology can also be called the out-of-band management topology. The management bus topology describes the topological information of the devices involved in the out-of-band management of the server board 100. That is, the connection relationship between the devices involved in the out-of-band management in the server board 100 (temperature sensor, voltage sensor, ADC, power supply, fan). The management bus topology includes but is not limited to: information about the devices hanging under the management bus 300 (such as the device manager 120 or the second type of device), and information about the devices connected to the device manager 120 (that is, the first type of device). All information about devices involved in out-of-band management can be recorded in the in-band management topology.

[0101] The baseboard management controller 200 obtains the topology information of the server board 100 and can understand the connection relationship of the devices on the server board 100. Based on the topology information of the server board 100, it can determine which device's working environment is described by the working information of the first type of device subsequently read through the device manager 120 (such as temperature, voltage, whether the power supply is working, etc.) and the working information of the second type of device directly read from the second type of device, and then determine whether the working environment of the device meets the requirements or whether the device is faulty, and whether an alarm needs to be issued.

[0102] The device attribute information includes the attribute information of the first type of device and the attribute information of the second type of device. In terms of device type, the device includes chips (such as processor chips, etc.), connectors, buses, and slots (slots refer to slots where input / output devices are inserted, such as PCIe slots, hard disk slots, etc.).

[0103] The information included in the management information can be found in Table 3. It should be noted that the above description and Table 1 only show part of the information in the management information. The embodiment of the present application does not limit the division method and information content of the management information. All information required for out-of-band management can be used as management information and stored in the memory 110.

[0104] Table 3

[0105]

[0106]

[0107]

[0108] In the embodiment of the present application, the management information required for out-of-band management is stored in the memory 110, and the address of the memory 110 can be a preset address. When the baseboard management controller 200 is connected to the memory 110 through the management bus 300, it can interact with the memory 110 through the address and read the management information from the memory 110 to implement subsequent out-of-band management. The baseboard management controller 200 can obtain the management information relatively simply and quickly, simplifying the process of out-of-band management.

[0109] In the embodiment of the present application, the server board 100 The out-of-band management interfaces are unified into a root management bus. A fixed address memory 110 (such as EEPROM) can be attached to the physical link as FRUD, and the server is described in FRUD. The management information of the single board 100, the baseboard management controller 200 can automatically load the management information of the single board by reading the information in the FRUD Configuration.

[0110] (2) Device manager 120.

[0111] The device manager 120 may also be referred to as a satellite management center (SMC). After the SMC collects the working information of the first type of devices on the single board, it reports it to the baseboard management controller 200 through the root management bus interface. The reporting method may be a command word method, and one type of working information may correspond to one command word.

[0112] On the server board 100, the server board 100 uses SMC As the board-level management center, collect The working information of the first type of device, such as sensor information, alarm information, upgrade requirements of the processing server board 100 and single Management requirements for other devices on the board. SMC The root management bus interface uses command words to communicate with the baseboard management controller 200 communication.

[0113] In the embodiment of the present application, the baseboard management controller 200 does not need to be connected to each device that needs to be managed out-of-band, but obtains the working information of these devices through the device manager 120, and then determines the working environment of the main devices in the server. The baseboard management controller 200 only needs to be connected to the device manager 120, which can greatly simplify the connection method between the baseboard management controller 200 and the server single board 100, and realize the intelligent management of the out-of-band management of the server single board 100. This connection method is also suitable for different server single boards 100.

[0114] The embodiment of the present application does not limit the specific structure of the device manager 120. For example, the device manager 120 may be a complex programmable logic device (CPLD) or a microcontroller unit (MCU). After the device manager 120 collects the working information of each device connected to it, it can report the collected information to the baseboard management controller 200 through the management bus 300.

[0115] The embodiment of the present application does not limit the interaction mode between the device manager 120 and the baseboard management controller 200. For example, the device manager 120 and the baseboard management controller 200 can interact in the form of command words. One type of work information corresponds to one command word. The format of the command word can be shared by different server boards 100, so that the baseboard management controller 200 can interact with the device managers 120 on different server boards 100 in the same way, reducing unnecessary adaptation work.

[0116] The following is a design method for command words. Figure 3The command word format defined between the device manager 120 and the baseboard management controller 200 mainly includes two parts, one part is an operation code (OP code) and a device parameter (parameter). The embodiment of the present application does not limit the specific size of the command word. In a possible implementation, the commander can occupy 4 bytes (i.e., 32 bits). Among them, the device parameter can occupy 1 byte and the operation code can occupy 3 bytes.

[0117] The operation code is used to describe the need to operate the device. In the embodiment of the present application, the operation may include reading the working information of the device and issuing a command to the device (issuing a command to the device can be understood as writing information to the device). The device parameter is used to indicate the device that needs to be operated. The device parameter can be the number or identification of the device.

[0118] The opcode consists of four fields: function field, command field, read count field ( Figure 3 MS is used to represent this field), and the read-write identification field ( Figure 3 RW is used to represent this field).

[0119] The function field is used to indicate the server board 100 to which the command word is directed. When the board management system includes multiple boards, the function field cannot be defaulted. When there is only one board in the board management system, the content of the function field can be set to a default value or a null value. The function field can occupy 6 bits.

[0120] When there are different types of server boards 100, different numbers may be used to indicate the different types of server boards 100. Figure 3 1 can indicate an expansion component (an expansion component refers to a component used to add an interface or slot in a server). 2 indicates a storage component (a storage component refers to a component used to connect a hard disk and implement data storage in a server). 3 indicates a base board. 4 indicates a memory expansion component (a memory expansion component refers to a component that performs memory functions in a server). 0 is used to represent a general command, that is, the command word is for all server boards 100.

[0121] The command field is used to describe the type of operation, such as indicating which type of operating information to read (such as temperature, voltage, whether the power supply is normal, fault or alarm information). The command field needs to be defined in advance to distinguish different operations. The command field can occupy 16 bits.

[0122] The read count field is used to distinguish whether this operation is multiple reads or a single read, that is, to indicate whether the working information of multiple devices is read at one time or the working information of one device is read at one time. For example, when this field is 0, it indicates multiple reads, and when it is 1, it indicates a single read. The read count field can occupy 1 bit.

[0123] The read / write identification field is used to distinguish whether the operation is a read operation or a write operation. For example, when the field is 0, it indicates that the operation is a read operation, and when it is 1, it indicates that the operation is a write operation. The read / write identification field can occupy 1 bit.

[0124] When the BMC 200 needs to read the operating parameters of the device, the interaction process between the device manager 120 and the BMC 200 includes: the BMC 200 initiates a read request to the device manager 120 , and the device manager 120 feeds back a read response to the BMC 200 .

[0125] like Figure 4A FIG. 1 is a schematic diagram of a format of a read request provided in an embodiment of the present application. Figure 4B A schematic diagram of the format of a read response request provided in an embodiment of the present application. Figure 4A and Figure 4B The first line is the name of each field, and the second line is the number of bits occupied by each field.

[0126] When the baseboard management controller 200 needs to write information to the device, that is, when the baseboard management controller 200 sends a command to the device (such as controlling the device to start, stop, or upgrade), the interaction process between the device manager 120 and the baseboard management controller 200 includes: the baseboard management controller 200 initiates a write request to the device manager 120, and the write request carries the command (such as a control command) or data (upgrade file) to be written.

[0127] like Figure 4C Shown is a schematic diagram of the format of a write request provided in an embodiment of the present application. Figure 4C The first line is the name of each field, and the second line is the number of bits occupied by each field.

[0128] Figure 4A to Figure 4C The meaning of each field can be found in Table 4.

[0129] Table 4

[0130]

[0131]

[0132] It should be noted that the above Figure 4A to Figure 4CThe fields in the description are only examples. In actual applications, when designing the fields in the read request, write request, and read response, the fields may be increased or decreased according to actual needs.

[0133] In the embodiment of the present application, in addition to exchanging the working information of the first-category device between the device manager 120 and the baseboard management controller 200, the baseboard management controller 200 can also send control commands to the first-category device through interaction with the device manager 120 to control the working state of the first-category device. For example, the control command can control one or several first-category devices to stop working or start working. The control command can be carried as data in the following manner: Figure 4C When the device manager 120 receives the write request, it can identify the control command therein and control the corresponding first-category device according to the control command, such as controlling the first-category device to stop working or start working.

[0134] The baseboard management controller 200 can also send an upgrade command to the first type of device through interaction with the device manager 120 to instruct the first type of device to upgrade. The upgrade file required for the first type of device upgrade can be carried as data in such a manner as Figure 4C When the device manager 120 receives the write request, it can identify the upgrade file therein, send the upgrade file to the corresponding first-type device, and instruct the first-type device to upgrade.

[0135] The baseboard management controller 200 may also directly instruct the device manager 120 to upgrade. The upgrade file required for the device manager 120 to upgrade may be carried as data in a file such as Figure 4C When the device manager 120 receives the write request, it can identify the upgrade file therein and use the upgrade file to perform the upgrade.

[0136] 3) Baseboard management controller 200.

[0137] From the above description about the memory 110 and the device manager 120 , it can be known that the basic management controller can read management information from the memory 110 through the management bus 300 , and can also implement out-of-band management of the first type of devices through interaction with the device manager 120 .

[0138] If there are devices (i.e., second-type devices) on the server board 100 that cannot be managed through the SMC, they can be directly installed on The baseboard management controller 200 directly accesses the root management bus, and is described in FRUD. Baseboard Management Controller 200 Automatically load the Management features of this class of devices.

[0139] In an embodiment of the present application, the existence of a second type of device on the server board 100 is allowed. The second type of device can be directly connected to the baseboard management controller 200 through the management bus 300. The baseboard management controller 200 can directly interact with the second type of device through the management bus 300 to obtain the working information of the second type of device and implement out-of-band management of the second type of device.

[0140] The baseboard management controller 200 can determine the second type of device deployed on the server board 100 according to the management information, that is, obtain information about the second type of device directly connected to the management bus 300. The baseboard management controller 200 can preload the management driver for the second type of device based on the management information (the management driver refers to the software program required to manage the second type of device) to achieve management of the second type of device.

[0141] In the embodiment of the present application, the baseboard management controller 200 can be deployed on a single board to form a BMC management single board (that is, the expansion board mentioned above). The BMC management single board can be used as the management center of the server to implement out-of-band management of the server. The appearance of the BMC management single board can be as follows Figure 5A The BMC management board provides external management interfaces, including a debug serial port, a unit identification (UID) indicator, a management network port, a video graphics array (VGA) interface, and a universal serial bus (USB) interface. The BMC management board provides external management interfaces, see Figure 5B .

[0142] Table 5 shows the definition and description of the management interface functions provided by the BMC management board.

[0143] Table 5

[0144]

[0145]

[0146] The BMC management board provides the management interface required for board management through the 4C+ connector, including an out-of-band management bus interface. If the management bus is an I2C bus, the out-of-band management bus interface is the I2C interface.

[0147] The BMC management board can also provide other management interfaces, and the embodiments of the present application do not limit the types of the other management interfaces. Other management interfaces include some or all of the following: Joint Test Action Group (JTAG) interface, SPI interface, network controller sideband interface (NCSI), platform environment control interface (PECI) debug serial port, UID button indicator, management network port, VGA interface. The types of other management interfaces are only examples, and the embodiments of the present application do not limit the number and type of other management interfaces.

[0148] The BMC management board also provides the low pin count (LPC) interface, USB interface, and PECI interface required for in-band management. The BMC management board also deploys the power supply, clock circuit, and stray signal circuit required for the baseboard management control itself. The management interface pin definitions provided by the BMC management board are shown in Table 6 below:

[0149] Table 6

[0150]

[0151]

[0152]

[0153]

[0154] Among them, Power / GND indicates a power signal or a ground signal, USB3 means supporting USB3.0. Input indicates signal input, and output indicates signal output. VGA refers to a VGA signal. The VGA signal in the above table includes three signals, namely red, green, and blue signals. HCSL refers to a high-speed current steering logic level. The signal definitions are only exemplary. In actual use, different signal definitions can also be set according to actual needs. The following takes the structure of the single-board management system to which three different types of server single boards 100 belong as an example to illustrate the single-board management system provided in an embodiment of the present application.

[0155] The first type, the server board 100 is a basic computer unit (BCU).

[0156] like Fig. 6AAs shown, a single board management system provided in an embodiment of the present application can be used to implement out-of-band management of a computing processing unit. The BMC is connected to the EEPROM and CPLD of the BCU respectively through an I2C bus. Among them, the EEPROM is used to implement the function of the memory 110 in the above embodiment, in which the management information of the computing processing unit, such as the attribute information of the computing processing unit, is stored. The CPLD is used to implement the function of the SMC in the above embodiment, such as implementing management control of the device, processing upgrade commands or control commands, etc. The CPLD connects devices such as ADC, temperature sensor, clock circuit, flash memory, etc. Fig. 6A In the example, the CPLD can obtain some signals through some devices on the first conversion chip server board 100. Fig. 6A The CPLD can obtain three signals through the first conversion chip server board 100, and the three signals include a power OK (PG) signal (used to indicate whether the power is connected or not connected), a present signal (such as the present signal can be used to indicate whether there is a device connected to the connector), and a fault signal.

[0157] The power OK signal is used to indicate whether the power is connected or not. The present signal can be used to indicate whether a device is connected to the connector. The fault signal can be used to indicate whether a device is faulty, such as a CPU or a power controller. For example, the CPU can be directly connected to the first conversion chip (e.g., a 9555 chip) through a low-speed signal line to provide a CPU alarm signal, which indicates that an error has occurred in the CPU. The first conversion chip is used to increase the number of connected devices.

[0158] CPLD can obtain ADC working information (ADC working information is the digital signal converted by ADC into voltage signal), temperature, CPU alarm signal, power supply information, etc. CPLD can also load the frequency of clock circuit and flash memory upgrade function.

[0159] The second conversion chip (e.g., 9545 chip) can provide multiple I2C interfaces. Multiple voltage regulator controllers on the computing processing unit are directly connected to the I2C bus after being expanded by the second conversion chip. The topology information in the computing processing unit in the EEPROM describes the connection relationship of the voltage regulator controller directly connected to the I2C bus. The voltage regulator controller is used to supply power to the CPU.

[0160] BMC can directly manage the voltage regulator power controller. CPLD interacts with BMC through I2C bus based on command words, transmits the working information of the device connected to CPLD, and can also accept the control of BMC to upgrade and load some devices. BMC can also upgrade CPLD through I2C bus.

[0161] The second type is that the server board 100 is an IO component (input output unit, IOU).

[0162] like Figure 6B As shown, a single board management system provided by an embodiment of the present application can be used to implement out-of-band management for an IO expansion unit. The BMC is connected to the EEPROM and MCU of the IOU respectively through an I2C bus.

[0163] Among them, the EEPROM is used to implement the function of the memory 110 in the above embodiment, in which the management information of the IOU is stored, and the MCU is used to implement the function of the SMC in the above embodiment, such as realizing the management and control of the device, processing the upgrade command or the control command, etc. The MCU is connected to the temperature sensor power supply, power supply, PCIe slot and other devices. The MCU can obtain the working information such as the temperature, the PG signal of the first conversion chip and the in-place signal (the in-place signal can indicate whether the connector has a device inserted).

[0164] The MCU interacts with the BMC through the I2C bus based on command words, and transmits the working information of the device connected to the MCU. The BMC implements the upgrade function for the MCU through the I2C bus. The PCIe card inserted into the PCIe slot is directly connected to the I2C bus through the second conversion chip. The topology information in the IOU in the EEPROM describes the connection relationship between the PCIe card and the I2C bus. The BMC can directly manage the PCIe card.

[0165] The third type is that the server board 100 is a storage unit (STU).

[0166] like Figure 6CAs shown, a single board management system provided by an embodiment of the present application can be used to implement out-of-band management for a storage expansion unit. The BMC is connected to the EEPROM and CPLD of the BCU respectively through an I2C bus. Among them, the EEPROM is used to implement the function of the memory 110 in the above embodiment, in which the management information of the computing processing unit, such as the attribute information of the computing processing unit, is stored. The CPLD is used to implement the function of the SMC in the above embodiment, such as implementing the management and control of the device, processing upgrade commands or control commands, etc. The CPLD is connected to devices such as temperature sensors, ADCs, and hard disks. The CPLD can obtain voltage, temperature, whether the hard disk is connected, and can also obtain working information such as PG signals, in-position signals, and CPU alarm signals through the fifth conversion chip. The CPLD can also implement the management function of the hard disk and obtain the working information of each hard disk through the sixth conversion chip. The CPLD interacts with the BMC through the I2C bus based on the command word method to transmit the working information of the device connected to the CPLD. The BMC can obtain the working information of each hard disk of the single board through the command word method. The BMC can also implement the upgrade function of the CPLD through the I2C bus.

[0167] Based on the single board management system provided above, the single board management method provided in the embodiment of the present application is described below. Figure 7 , the method comprises the following steps:

[0168] Step 701 : After being started, the baseboard management controller 200 scans the memory 110 at a preset address under the management bus 300 through the management bus 300 .

[0169] After the server is powered on, the baseboard management controller 200 is started, and the baseboard management controller 200 can find the memory 110 with a preset address from the devices connected to the management bus 300 through the management bus 300 .

[0170] Step 702: After scanning the memory 110, the baseboard management controller 200 reads the management information of the server board 100 from the memory 110 through the management bus 300. The information included in the management information can refer to the description of the above content. By reading the management information, the baseboard management controller 200 can understand the hardware information of the server board 100, the topology information of the server board 100, and the attribute information of the device of the server board 100.

[0171] Step 703: After the server board 100 is powered on, the device manager 120 on the server board 100 collects the working information of the first type of devices.

[0172] After the server board 100 is powered on, the device manager 120 can interact with the first type of device to which the device manager 120 is connected to obtain working information of the first type of device, such as the temperature detected by the temperature sensor, the voltage detected by the ADC, the power supply OK information from the voltage regulating power supply controller, and the device fault information (such as CPU alarm information).

[0173] Step 704: the baseboard management controller 200 obtains the working information of the first type of device from the device manager 120. If the server board 100 includes the second type of device, the baseboard management controller 200 may also obtain the working information of the second type of device from the second type of device via the management bus 300.

[0174] In step 704, the baseboard management controller 200 can collect the working information of the first type of components of the server through the device manager 120, and can also obtain the working information of the second type of components through direct interaction. The baseboard management controller 200 does not need to be connected to each component on the server board 100, and the method for the baseboard management controller 200 to obtain the working information of the server's components is relatively simple.

[0175] Step 705: the baseboard management controller 200 manages the server board 100 based on the management information and the acquired working information of the components (such as the working information of the first type of components and the working information of the second type of components).

[0176] The baseboard management controller 200 can understand the connection management of the devices on the server board 100 based on the management information, and can determine the working environment of some main devices on the server board 100 (such as temperature, voltage, whether it is powered on, whether it is faulty, etc.) based on the working information of the devices. Based on this, the baseboard management controller 200 can determine whether to control the devices on the server board 100, such as turning on the fan, restarting the power supply, etc. The baseboard management controller 200 can send a control command to the device manager 120 to control the first type of devices. The baseboard management controller 200 can also directly send a control command to the second type of devices through the management bus 300 to control the second type of devices. For the method of sending the control command, please refer to the above content, which will not be repeated here.

[0177] In addition to controlling the devices, the baseboard management controller 200 can also upgrade the devices. For example, the baseboard management controller 200 can send an upgrade command to the device manager 120 to upgrade the first type of devices. The baseboard management controller 200 can send an upgrade command to the device manager 120 to upgrade the first type of devices. The baseboard management controller 200 can also send an upgrade command to the device manager 120 to upgrade the device manager 120. The baseboard management controller 200 can also directly send an upgrade command to the second type of devices through the management bus 300 to upgrade the second type of devices. For the method of sending the upgrade command, please refer to the above content, which will not be repeated here.

[0178] The baseboard management controller 200 can also determine whether to issue an alarm to the user to prompt the user that a device fails or the temperature is high, or there is a power supply error, etc. In this way, the baseboard management controller 200 can manage the server board 100 to ensure that the server board 100 can work normally, or the user can understand the status of the server board 100 in a timely manner.

[0179] like Figure 8 As shown, a management system of a BCU module provided in an embodiment of the present application is provided. The management system of the BCU module is used to ensure the management characteristics of the BCU module.

[0180] The management features of the BCU module include the management interface provided by the BCU module to the outside, and the management features of the management module to the BCU module.

[0181] The low-speed signal on the high-speed connector of the BCU module contains the management signal, which can be used for out-of-band management of the riser card of the BCU module. The advantage of this design is that the low-speed management signal line can be omitted on the riser card.

[0182] The management module of the BCU module is divided into out-of-band management and in-band management. The Tianchi management architecture recommends terminating the independent management features on the BCU module directly on the BCU module. For example, the frequency synthesizer configuration on the BCU module is loaded directly on the BCU module and does not require separate management by the management module.

[0183] like Figure 8 In the example, the BMC on the management module provides an intelligent platform management bus (IPMB) interface to connect to the CPU of the BCU module as an intelligent platform management interface (IPMI) bus channel;

[0184] The BMC on the management module provides an LPC interface to connect to the CPU of the BCU module as a BT bus channel;

[0185] The BMC on the management module provides an I2C interface to connect to the CPLD and FRUD of the BCU module. The BMC implements basic out-of-band management of the BCU module through this I2C, including reading information in the FRUD and accessing the BCU module CPLD register as an SMC bus channel.

[0186] The CPLD chip on the management module provides two hisport interfaces to connect to the CPLD of the BCU module. One hisport0 is used as the logical register interaction channel between the BCU module and the management module, and the other is used as the hisport overI2C interface for the BCU module to expand the external management interface.

[0187] The CPLD chip on the BCU module provides multiple I2C interfaces for reading and configuring the information of the ADC chip, clock frequency synthesizer chip, and temperature sensing chip of the BCU module. That is, the CPLD on the BCU module reads basic information such as temperature and voltage, and reports it to the BMC chip through a unified SMC interface, thereby terminating the independent management feature inside the module.

[0188] The CPLD chip on the BCU module provides multiple I2C interfaces to connect to the UBC high-speed connector as a management channel for external expansion modules. These external management I2Cs are derived from the hisport over I2C feature provided by the management module. This management channel can connect to out-of-band management devices such as FRU chips and temperature sensors on the Riser to achieve out-of-band management features for components.

[0189] Based on the same inventive concept as the method embodiment, the embodiment of the present application also provides a single board management device, which is used to perform the above-mentioned Figure 7 The method executed by the baseboard management controller in the method embodiment shown in the figure, the relevant features can be found in the above method embodiment, and will not be repeated here. Fig. 9 As shown, the single board management device 900 includes an acquisition unit 901 and a management unit 902 .

[0190] The acquisition unit 901 is used to acquire management information from the memory via the management bus.

[0191] The management unit 902 is used to manage the computing device board by interacting with the device manager through the management bus based on the management information.

[0192] In a possible implementation manner, the computing board includes a first type of device, the device management device is connected to the first type of device, and the acquisition unit 901 can acquire the working information of the first type of device from the device manager through the management bus.

[0193] In a possible implementation, the computing board includes a second type of device, and the second type of device is connected to the baseboard management controller via a management bus. The acquisition unit 901 can acquire the working information of the second type of device from the second type of device via the management bus.

[0194] In a possible implementation manner, the management information includes part or all of the following: attribute information of a computing device board, topology information of a computing device board, attribute information of first-category devices, and attribute information of second-category devices.

[0195] In a possible implementation, when the management unit 902 interacts with the device manager through the management bus, the interaction may be based on command words.

[0196] In a possible implementation manner, the apparatus further includes an upgrade unit 903. The upgrade unit 903 may transfer an upgrade file of the first type of device to the device manager, instructing the first type of device to be upgraded. Alternatively, the upgrade file of the device manager may be transferred to the device manager, instructing the device manager to be upgraded.

[0197] In a possible implementation manner, the management bus is an I2C bus or an SPI bus.

[0198] It should be noted that the division of units in the embodiments of the present application is schematic and is only a logical function division. There may be other division methods in actual implementation. Each functional unit in the embodiments of the present application may be integrated into a processing unit, or each unit may exist physically separately, or two or more units may be integrated into a module. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional modules.

[0199] This application also provides Fig.10 The computing device 1000 shown in FIG. 1 includes a computer board and a baseboard management controller 1500 . The computer board may include a bus 1100 , a processor 1200 , a communication interface 1300 , and a memory 1400 . The processor 1200 , the memory 1400 , and the communication interface 1300 communicate with each other via the bus 1100 .

[0200] Among them, the processor 1200 can be a central processing unit (CPU), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), an artificial intelligence (AI) chip, a system on chip (SoC) or a complex programmable logic device (CPLD), a graphics processing unit (GPU), etc.

[0201] The memory 1400 may include a volatile memory, such as a random access memory (RAM). The memory 1400 may also include a non-volatile memory, such as a read-only memory (ROM), a flash memory, a HDD or a SSD. The memory 1400 may also include the memory 110 mentioned above, that is, management information may be stored therein. The memory 1400 may also store software modules required for other running processes such as an operating system. The operating system may be LINUX. TM ,UNIX TM ,WINDOWS TM wait.

[0202] The baseboard management controller 1500 includes a processor 1510 and a memory 1520. The memory 1520 stores computer program codes. The processor 1510 executes the computer program codes to perform the aforementioned Figure 7 The baseboard management controller 1500 may also only include a processor 1510, on which a computer program code is burned, and the processor 1510 may execute the aforementioned Figure 7 The method described.

[0203] The descriptions of the processes corresponding to the above-mentioned figures have different emphases. For parts that are not described in detail in a certain process, please refer to the relevant descriptions of other processes.

[0204] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments may be implemented in the form of a computer program product. The computer program product includes computer program instructions, which, when loaded and executed on a computer, generate all or part of the embodiments according to the present invention. Figure 7 The process or function described.

[0205] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid state drive (SSD).

[0206] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A computing device single board, characterized in that: include A memory, wherein the memory records management information of the computing device board; A device manager, wherein the device manager manages the devices deployed on the computing device board; wherein the memory and the device manager are connected to a baseboard management controller BMC via a management bus, and provide the management information to the BMC so that the BMC manages the computing device board.

2. The computing device board according to claim 1, characterized in that: The memory is an electrically erasable programmable read-only memory (EEPROM).

3. The computing device board according to claim 1 or 2, characterized in that: The managed information includes a field replaceable unit description FRUD.

4. The computing device board according to any one of claims 1 to 3, wherein the management information includes part or all of the following: The attribute information of the computing device board, the topology information of the computing device board, the attribute information of the first type of devices, and the attribute information of the second type of devices.

5. The computing device board according to any one of claims 1 to 4, characterized in that: The device manager is a complex programmable logic device (CPLD) or a microcontroller unit (MCU).

6. The computing device board according to any one of claims 1-5, wherein the devices on the computing device board include first-category devices, the first-category devices are connected to the device manager, and the working information of the first-category devices is provided to the BMC through the device manager.

7. According to any one of claims 1-6, the computing device board on the computing device board includes second-category devices, the second-category devices are connected to the BMC via the management bus, and the working information of the second-category devices is provided to the BMC via the management bus. 8 . The computing device board according to claim 1 , wherein the device manager interacts with the BMC based on command words.

9. The computing device board according to any one of claims 1 to 8, characterized in that: The device manager receives the upgrade file sent by the BMC and upgrades the device on the computing device board.

10. The computing device board according to any one of claims 1 to 9, wherein the management bus is an internal integrated circuit I2C bus or a serial peripheral interface SPI bus.

11. A computing device, characterized in that: The computing device comprises a computing device board as claimed in any one of claims 1 to 10.

Images