Blade server system, server and CXL memory configuration method

By incorporating CXL Switch board and blade nodes into the server chassis and directly connecting using orthogonal high-density connectors, the problem of numerous cable connections in the server system is solved, reducing costs and improving maintenance reliability, and achieving flexible memory pooling topologies.

CN118689287BActive Publication Date: 2024-11-15INSPUR SUZHOU INTELLIGENT TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411184200.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-27
Publication Date
2024-11-15
Estimated Expiration
2044-08-27

AI Technical Summary

Technical Problem

In the prior art, a large number of cable connections are required between the server chassis, CXL extension BOX and CXL Switch rack BOX, resulting in high costs and insufficient reliability of equipment assembly and maintenance.

Method used

Design a blade server system to place the CXL Switch board and blade node in the server chassis housing, directly connect through orthogonal high-density connectors to reduce the number of cables, and use built-in high-density cable connectors to fix them in the chassis to avoid plugging and unplugging. Influence.

Benefits of technology

Reduces costs, improves the reliability of equipment assembly and maintenance, realizes a more flexible memory pooling topology, supports a variety of CXL memory pooling requirements, and avoids cost and operation and maintenance issues of external cables and PCIE interface cards.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118689287B_ABST
    Figure CN118689287B_ABST
Patent Text Reader

Abstract

The embodiments of the present application provide a blade server system, a server, a CXL memory configuration method, a storage medium, an electronic device, and a computer program product, wherein the blade server system comprises: a chassis housing, a plurality of blade nodes, and a computing fast link switching CXL Switch board, wherein the plurality of blade nodes and the CXL Switch board are built in the chassis housing; each of the blade nodes is directly connected to the CXL Switch board through an orthogonal high-density connector; wherein, in the case where the plurality of blade nodes include a target blade node that meets a target condition, the target blade node is further connected to the CXL Switch board through a high-density cable connector; the plurality of blade nodes are all central processing unit CPU computing nodes; or, some of the plurality of blade nodes are central processing unit CPU computing nodes, and some of the blade nodes are computing fast link CXL memory expansion nodes.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The embodiments of the present application relate to the computer field, and in particular, to a blade server system, a server, a CXL memory configuration method, a storage medium, an electronic device, and a computer program product. Background Art

[0002] CXL (Compute Express Link) technology can provide higher data throughput and lower latency, so it can meet the needs of modern computing and storage systems. In related technologies, since the server chassis, CXL expansion BOX and CXL Switch rack BOX are set up separately, a large number of cables are needed to connect the server chassis, CXL expansion BOX and CXL Switch rack BOX, which is costly and the reliability of equipment assembly and maintenance is insufficient. Summary of the invention

[0003] The embodiments of the present application provide a blade server system, a server, a CXL memory configuration method, a storage medium, an electronic device, and a computer program product, so as to at least solve the problem in the related art that since a server chassis, a CXL expansion BOX, and a CXL Switch rack-mounted BOX are separately provided, a large number of cables need to be used to connect the server chassis, the CXL expansion BOX, and the CXL Switch rack-mounted BOX, resulting in high costs and insufficient reliability of equipment assembly and maintenance.

[0004] According to an embodiment of the present application, a blade server system is provided, comprising: a chassis housing, a plurality of blade nodes and a computing fast link exchange CXL Switch board, wherein the plurality of blade nodes and the CXL Switch board are built in the chassis housing; each of the blade nodes is directly connected to the CXL Switch board through an orthogonal high-density connector; wherein, in the case where the plurality of blade nodes include a target blade node that meets a target condition, the target blade node is further connected to the CXL Switch board through a high-density cable connector; the plurality of blade nodes are central processing unit CPU computing nodes; or, some of the plurality of blade nodes are central processing unit CPU computing nodes, and some of the blade nodes are computing fast link CXL memory expansion nodes.

[0005] In an exemplary embodiment, when multiple blade nodes are the CPU computing nodes, the blade server system further includes: a memory module, wherein the memory module is arranged in the chassis shell, the memory module is connected to the cable socket cable arranged on the CXL Switch board through a high-density cable connector, and the storage method of the memory module includes at least one of the following: E3.S memory expansion module, PCIE peripheral component interconnect fast channel memory expansion card, memory expansion single board.

[0006] In an exemplary embodiment, when the plurality of blade nodes include a target blade node that meets the target condition, the target blade node is further connected to the CXL Switch board through a high-density cable connector, including: when some of the blade nodes among the plurality of blade nodes are the CPU computing nodes, and some of the blade nodes are the CXL memory extension nodes, and when first target data to be stored in the CXL memory extension node reaches a first threshold, the CXL memory extension node is further connected to the CXL Switch board through a high-density cable connector, wherein the target data is data generated by the CPU computing node to be stored in the CXL memory extension node.

[0007] In an exemplary embodiment, when the plurality of blade nodes include a target blade node that meets the target condition, the target blade node is further connected to the CXL Switch board through a high-density cable connector, including: when some of the plurality of blade nodes are the CPU computing nodes, and some of the blade nodes are the CXL memory extension nodes, and when second target data to be stored in the CXL memory extension node reaches a second threshold and the computing capacity of the CPU computing node reaches a third threshold, the CPU computing node and the CXL memory extension node are both further connected to the CXL Switch board through a high-density cable connector, and the second target data is data generated by the CPU computing node to be stored in the CXL memory extension node.

[0008] In an exemplary embodiment, when the plurality of blade nodes include a target blade node that meets the target condition, the target blade node is further connected to the CXL Switch board through a high-density cable connector, including: when all of the plurality of blade nodes are CPU computing nodes and the computing capacity of the CPU computing node reaches a fourth threshold, the CPU computing node is further connected to the CXL Switch board through a high-density cable connector.

[0009] In an exemplary embodiment, each of the blade nodes is directly connected to the CXLSwitch board via an orthogonal high-density connector, including: each of the blade nodes is provided with a female head of the orthogonal high-density connector, the CXLSwitch board is provided with a male head of the orthogonal high-density connector corresponding to each of the blade nodes, and each of the blade nodes and the CXL Switch board are directly connected by plugging the female head and the male head belonging to the same orthogonal high-density connector.

[0010] In an exemplary embodiment, the CXL Switch board is further provided with: a CXL Switch chip and a cable socket; the target blade node is further connected to the CXL Switch board via a high-density cable connector, including: the target blade node is connected to the cable socket provided on the CXL Switch board via a high-density cable connector.

[0011] In an exemplary embodiment, when some of the blade nodes among the plurality of blade nodes are the CXL memory expansion nodes, the memory expansion method of the CXL memory expansion node includes at least one of the following: E3.S memory expansion module, PCIE peripheral component interconnect fast channel memory expansion card, memory expansion board.

[0012] According to one embodiment of the present application, a blade server is also provided, which runs the above-mentioned blade server system, and the blade server includes: a power module, a fan module, and a centralized management board; wherein the centralized management board is used to interact and communicate with the management chip of each of the blade nodes, the centralized management board leads out a centralized management interface, and the power module leads out a power input interface, each of the blade nodes is provided with a first hot-swap handle, and each of the power modules is provided with a second hot-swap handle, a plurality of the blade nodes are arranged side by side on the front side of the chassis shell, a plurality of the power modules are arranged side by side below the plurality of the blade nodes, and a plurality of the fan modules, the power input interface, and the centralized management interface are arranged on the rear side of the chassis shell.

[0013] According to an embodiment of the present application, a CXL memory configuration method is also provided, which is applied to the aforementioned blade server system, including: some blade nodes among the multiple blade nodes are the CPU computing nodes, and some blade nodes are the CXL memory expansion nodes. When the CXL Switch board receives the data to be stored sent by the CPU computing node through the CXL protocol, the CXL Switch board routes the data to be stored to the CXL memory expansion nodes in sequence for storage according to the priority of each CPU computing node that sends the data to be stored.

[0014] In an exemplary embodiment, the method further includes: when the data generated by the CPU computing node to be stored in the CXL memory expansion node includes multiple types of data, storing them in sequence in the following order: operating system data, predetermined application data, security data, system performance monitoring data, user data application logs, backup data, test and development data.

[0015] In an exemplary embodiment, before routing the data to be stored to the CXL memory expansion node in sequence according to the priority of each CPU computing node that sends the data to be stored, the method further includes: determining a target memory requirement of the target CPU computing node based on the data to be stored generated by the target CPU computing node within a predetermined period of time, wherein the target CPU computing node is any one of the multiple CPU computing nodes that generate the data to be stored; acquiring and analyzing a target remaining storage space of each of the multiple memories included in the CXL memory expansion node; and determining a target memory for storing the data to be stored generated by the target CPU computing node based on the target memory requirement and the target remaining storage space.

[0016] According to another embodiment of the present application, a computer-readable storage medium is provided, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when running.

[0017] According to another embodiment of the present application, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0018] According to another embodiment of the present application, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in any of the above method embodiments are implemented.

[0019] Through the present application, since the CXL Switch board and the blade nodes are placed in the server chassis shell, and multiple blade nodes are directly connected to the CXL Switch board through orthogonal high-density connectors, it is possible to solve the problem in the related art that due to the separate arrangement of the server chassis, CXL expansion BOX and CXL Switch rack BOX, a large number of cables need to be used to connect the server chassis, CXL expansion BOX and CXL Switch rack BOX, resulting in high costs and insufficient reliability of equipment assembly and maintenance, thereby achieving the effect of reducing costs and improving the reliability of equipment assembly and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 The structure of the blade server system according to the embodiment of the present application is Figure 1 ;

[0021] Figure 2 is a schematic diagram of a blade server system according to an embodiment of the present application Figure 1 ;

[0022] Figure 3 The structure of the blade server system according to the embodiment of the present application is Figure 2 ;

[0023] Figure 4 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 1 ;

[0024] Figure 5 The structure of the blade server system according to the embodiment of the present application is Figure 3 ;

[0025] Figure 6 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 2 ;

[0026] Figure 7 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 3 ;

[0027] Figure 8 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 4 ;

[0028] Fig. 9 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 5 ;

[0029] Fig.10 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 6 ;

[0030] Fig.11 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 7 ;

[0031] Fig.12 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 8 ;

[0032] Fig.13 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 9 ;

[0033] Fig.14 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 10 ;

[0034] Fig.15 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 10 one;

[0035] Fig.16 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 10 two;

[0036] Fig.17 The structure of the blade server according to the embodiment of the present application is Figure 1 ;

[0037] Fig.18 The structure of the blade server according to the embodiment of the present application is Figure 2 ;

[0038] Fig.19 is a hardware structure block diagram of a server device according to a CXL memory configuration method according to an embodiment of the present application;

[0039] Fig. 20 is a flowchart of a CXL memory configuration method according to an embodiment of the present application. DETAILED DESCRIPTION

[0040] The embodiments of the present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0041] Figure 1 The structure of the blade server system according to the embodiment of the present application is Figure 1 , Figure 2 is a schematic diagram of a blade server system according to an embodiment of the present application Figure 1 ,like Figure 1 As shown, it includes: a chassis shell, multiple blade nodes and a computing fast link exchange CXL Switch board, wherein the multiple blade nodes and the CXL Switch board are built in the chassis shell; each of the blade nodes is directly connected to the CXL Switch board through an orthogonal high-density connector; wherein, in the case where the multiple blade nodes include a target blade node that meets the target condition, the target blade node is also connected to the CXL Switch board through a high-density cable connector; the multiple blade nodes are all central processing unit CPU computing nodes; or, some of the multiple blade nodes are central processing unit CPU computing nodes, and some of the blade nodes are computing fast link CXL memory expansion nodes.

[0042] In the above embodiment, the blade node includes, but is not limited to, a card-type server node that supports inserting a blade server in a plug-in form. The CXL Switch board has a CXL Switch chip, and the CXL Switch chip supports 256 CXL / PCIE (Peripheral Component Interconnect Express) physical channels, of which 128 CXL / PCIE physical channels are directly connected to the 8 blade nodes through the 8 orthogonal high-density connectors on the Switch board, and the other 128 CXL / PCIE physical channels on the CXL Switch board are connected to the 16 ×8 cable sockets on the CXL Switch board. The CXL Switch board can be connected to the blade node or other components through cables based on the cable sockets. Each of the blade nodes has 1 to 2 CPUs, and supports 1 ×16 at the rear of the blade node. The orthogonal high-density connector of CXL / PCIE bandwidth is plugged and connected to the CXLSwitch board, wherein each blade node includes but is not limited to a high-density cable connector with an unequal number of bandwidths of 0-112 paths. Through the above structure, different cable combinations form different topologies, thereby meeting the CXL memory pooling requirements of different customers. At the same time, the cables and the CXL Switch board are fixed inside the chassis without affecting the plugging and unplugging of the nodes.

[0043] In the above-mentioned embodiment of the present application, since the CXL Switch board and the blade nodes are placed in the server chassis shell, and the multiple blade nodes are directly connected to the CXL Switch board through the orthogonal high-density connectors, it is possible to solve the problem in the related art that due to the separate arrangement of the server chassis, CXL expansion BOX and CXL Switch rack BOX, a large number of cables need to be used to connect the server chassis, CXL expansion BOX and CXL Switch rack BOX, resulting in high costs and insufficient reliability of equipment assembly and maintenance, thereby achieving the effect of reducing costs and improving the reliability of equipment assembly and maintenance.

[0044] In an optional embodiment, when multiple blade nodes are the CPU computing nodes, the blade server system further includes: a memory module, wherein the memory module is arranged in the chassis shell, the memory module is connected to the cable socket cable arranged on the CXL Switch board through a high-density cable connector, and the storage method of the memory module includes at least one of the following: E3.S memory expansion module, PCIE peripheral component interconnect fast channel memory expansion card, memory expansion single board.

[0045] In the above embodiments, for example, Figure 3 The structure of the blade server system according to the embodiment of the present application is Figure 2 , Figure 4 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 1 ,like Figure 3 As shown, all the eight blade nodes are CPU computing nodes. Each CPU computing node is connected to the CXL Switch board using a set of orthogonal high-density connectors to transmit ×16 CXL / PCIE signals. The rear window of the blade server chassis supports a maximum of 16 E3.SCXL memory expansion modules. Each E3.S module is connected to a set of ×8 high-speed cable sockets on the CXL Switch board via cables. The memory modules are arranged side by side on the rear side of the chassis shell, and the multiple fan modules are arranged side by side above and below the memory modules.

[0046] In an optional embodiment, when the multiple blade nodes include a target blade node that meets the target condition, the target blade node is further connected to the CXL Switch board through a high-density cable connector, including: when some of the multiple blade nodes are the CPU computing nodes, and some of the blade nodes are the CXL memory extension nodes, and the first target data to be stored in the CXL memory extension node reaches a first threshold, the CXL memory extension node is further connected to the CXL Switch board through a high-density cable connector, wherein the target data is data generated by the CPU computing node to be stored in the CXL memory extension node.

[0047] In the above embodiments, for example, Figure 5 The structure of the blade server system according to the embodiment of the present application is Figure 3 ,like Figure 5 As shown, exemplarily, among the 8 blade nodes, 7 blade nodes are the CPU computing nodes, and 1 blade node is the CXL memory expansion node. The CXL memory expansion node is connected to the CXL Switch board with 8 ×16 CXL / PCIE signals, wherein 1 group of ×16 signals is connected using the orthogonal high-density connector, and the other 7 groups of ×16 signals are connected to the built-in 7 groups of high-density cable connectors using high-density connectors, and the high-density cable connectors are connected to 14 ×8 cable sockets on the CXL Switch board through cables inside the chassis.

[0048] In the above-mentioned embodiment of the present application, since the built-in high-density cable connector is fixed inside the chassis and is not pulled out along with the blade node, the impact on the plugging and unplugging maintenance of the blade node is avoided.

[0049] Figure 6 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 2 ,like Figure 6 As shown, the memory expansion method of the CXL memory expansion node includes: the E3.S memory expansion module, the PCIE memory expansion card, and the memory expansion board. Figure 7 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 3 ,like Figure 7 As shown, when the memory expansion mode of the CXL memory expansion node is the memory expansion board, illustratively, 16 CXL Memory controllers and 32 memory bars are supported. Figure 8 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 4 ,like Figure 8 As shown, when the memory expansion mode of the CXL memory expansion node is an E3.S memory expansion module, illustratively, 16 E3.S memory expansion modules are supported. Fig. 9 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 5 ,like Fig. 9 As shown, when the memory expansion mode of the CXL memory expansion node is the PCIE memory expansion card, exemplarily, 8 4DIMM (Dual In-line Memory Module) PCIE memory expansion cards and 4 8DIMM PCIE memory expansion cards are supported.

[0050] In an optional embodiment, when the multiple blade nodes include a target blade node that meets the target condition, the target blade node is further connected to the CXL Switch board through a high-density cable connector, including: when some of the multiple blade nodes are the CPU computing nodes, and some of the blade nodes are the CXL memory extension nodes, and the second target data to be stored in the CXL memory extension node reaches a second threshold and the computing capacity of the CPU computing node reaches a third threshold, the CPU computing node and the CXL memory extension node are both further connected to the CXL Switch board through a high-density cable connector, and the second target data is data generated by the CPU computing node to be stored in the CXL memory extension node.

[0051] In the above embodiment, exemplarily, among the 8 blade nodes, the blade nodes are connected to the CXLSwitch board with the same number of PCIE channels, supporting any number of the CPU computing nodes and the CXL memory expansion nodes, and each of the blade nodes is connected to the CXL Switch board with 2 ×16 CXL / PCIE signals, wherein 1 group of ×16 signals is connected using the orthogonal high-density connector, and another group of ×16 signals is connected on the blade node through a built-in group of high-density cable connectors, and the high-density cable connectors are connected to 2 ×8 cable sockets on the CXL Switch board through cables inside the chassis.

[0052] In the above-mentioned embodiment of the present application, since the built-in high-density cable connector is fixed inside the chassis and is not pulled out along with the blade node, the impact on the plugging and unplugging maintenance of the blade node is avoided.

[0053] Fig.10 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 6 ,like Fig.10 As shown, the memory expansion method of the CXL memory expansion node includes: the E3.S memory expansion module, the PCIE memory expansion card, and the memory expansion board. Fig.11 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 7 ,like Fig.11 As shown, when the memory expansion mode of the CXL memory expansion node is the memory expansion board, illustratively, 4 CXL Memory controllers and 8 memory bars are supported. Fig.12 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 8 ,like Fig.12 As shown, when the memory expansion mode of the CXL memory expansion node is an E3.S memory expansion module, illustratively, 4 E3.S memory expansion modules are supported. Fig.13 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 9 ,like Fig.13 As shown, when the memory expansion mode of the CXL memory expansion node is the PCIE memory expansion card, illustratively, two 4DIMM PCIE memory expansion cards and one 8DIMM PCIE memory expansion card are supported.

[0054] In the above embodiment, by adopting a balanced connection method, the user can arbitrarily change the number of computing nodes and the number of CXL memory expansion nodes, and support larger capacity CXL memory. At the same time, the CXL memory expansion node supports the use of a secondary switch to expand more memory capacity, thereby increasing the flexibility of memory settings while increasing memory capacity.

[0055] In an alternative embodiment, Fig.14 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 10 , Fig.15 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 10 one, Fig.16 The storage method of the blade server system according to the embodiment of the present application is shown in FIG. Figure 10 Second, add a CXL Switch chip to the CXL memory expansion node to expand the Fig.14 The 16 CXL Memory controllers and 32 memory modules shown in the figure are as follows: Fig.15 The 16 E3.S memory expansion modules shown in the figure are Fig.16 As shown, eight 4DIMM PCIE AIC memory expansion cards, or four 8DIMM PCIE AIC memory expansion cards.

[0056] In an optional embodiment, when the multiple blade nodes include a target blade node that meets the target condition, the target blade node is further connected to the CXL Switch board through a high-density cable connector, including: when the multiple blade nodes are all CPU computing nodes and the computing capacity of the CPU computing node reaches a fourth threshold, the CPU computing node is further connected to the CXL Switch board through a high-density cable connector.

[0057] In an optional embodiment, each of the blade nodes is directly connected to the CXLSwitch board via an orthogonal high-density connector, including: each of the blade nodes is provided with a female head of the orthogonal high-density connector, the CXLSwitch board is provided with a male head of the orthogonal high-density connector corresponding to each of the blade nodes, and each of the blade nodes and the CXL Switch board are directly connected by plugging the female head and the male head belonging to the same orthogonal high-density connector.

[0058] In an optional embodiment, the CXL Switch board is further provided with: a CXL Switch chip and a cable socket; the target blade node is further connected to the CXL Switch board via a high-density cable connector, including: the target blade node is connected to the cable socket provided on the CXL Switch board via a high-density cable connector.

[0059] In an optional embodiment, when some of the blade nodes among the multiple blade nodes are the CXL memory expansion nodes, the memory expansion method of the CXL memory expansion node includes at least one of the following: E3.S memory expansion module, PCIE peripheral component interconnect fast channel memory expansion card, memory expansion board.

[0060] Through the above embodiments, since a built-in CXL Switch switching board is designed in the blade server to support CXL shared memory blade nodes, the purpose of sharing pooled memory between multiple computing nodes in the blade server is achieved, so that the memory pooling solution between multiple hosts can be completed inside the blade server, avoiding the generation of external cables. Compared with the cabinet-level memory pooling solution in the related art, the cost of external cables and PCIE interface cards is reduced, and the operation and maintenance problems and reliability problems of external cables are avoided. In addition, through different connection combinations of built-in high-density connectors and cable connections, a variety of built-in CXL memory pooling topologies can be flexibly supported, and different CXL memory pooling requirements can be flexibly adapted.

[0061] Fig.17 The structure of the blade server according to the embodiment of the present application is Figure 1 , Fig.18 The structure of the blade server according to the embodiment of the present application is Figure 2 ,like Fig.17 , Fig.18 As shown, the blade server includes: a power module, a fan module, and a centralized management board; wherein the centralized management board is used to interact and communicate with the management chip of each of the blade nodes, the centralized management board leads out a centralized management interface, and the power module leads out a power input interface, each of the blade nodes is provided with a first hot-swap handle, and each of the power modules is provided with a second hot-swap handle, a plurality of the blade nodes are arranged side by side on the front side of the chassis shell, a plurality of the power modules are arranged side by side below the plurality of the blade nodes, and a plurality of the fan modules, the power input interface, and the centralized management interface are arranged on the rear side of the chassis shell.

[0062] In the above embodiment, exemplarily, the blade server includes but is not limited to: a height of 7U, 8 blade nodes are arranged vertically and side by side in the upper 6U, and 6 power modules are arranged side by side in the lower 1U, wherein the height of each blade node is 6U, and the width is not limited to 1.25U, each blade node supports a fast hot-swap function, each power module supports a fast hot-swap function, the power module inside the chassis of the blade server includes but is not limited to supplying power to the blade node and other components in the blade server through a power board, a busbar, a power cord, etc., the fan module is used to dissipate heat for the blade server, the CXL Switch board is arranged at the rear side of the blade node inside the blade server, and is designed with a CXL Switch board, and the CXL Switch board is directly connected to the blade node through the orthogonal high-density connector.

[0063] In the above embodiment, the centralized management board is used to perform the following operations: add or delete nodes, update node configurations and monitor node health; provide security maintenance functions, such as access control, encryption and monitoring system logs to detect potential malicious activities; monitor system performance, such as network traffic, CPU and memory usage, and identify bottlenecks that may affect system performance.

[0064] The method embodiments provided in the embodiments of the present application can be executed in a server device or a similar computing device. It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0065] Take running on a server device as an example, Fig.19 is a hardware structure block diagram of a server device according to the CXL memory configuration method of an embodiment of the present application. Fig.19 As shown, the server device may include one or more ( Fig.19 Only one is shown in the figure) a processor 1902 (the processor 1902 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 1904 for storing data, wherein the above-mentioned server device may also include a transmission device 1906 for communication functions and an input and output device 1908. A person skilled in the art can understand that Fig.19 The structure shown is only for illustration and does not limit the structure of the above server device. Fig.19 More or fewer components as shown, or with Fig.19 Different configurations are shown.

[0066] The memory 1904 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the CXL memory configuration method in the embodiment of the present application. The processor 1902 executes various functional applications and data processing by running the computer program stored in the memory 1904, that is, to implement the above method. The memory 1904 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 1904 may further include a memory remotely arranged relative to the processor 1902, and these remote memories may be connected to the server device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0067] The transmission device 1906 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the server device. In one example, the transmission device 1906 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 1906 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0068] In this embodiment, a CXL memory configuration method is provided, which can be applied to the blade server system described in any of the above embodiments. Fig. 20 is a flowchart of a CXL memory configuration method according to an embodiment of the present application, such as Fig. 20 As shown, the process includes the following steps:

[0069] Step S2002: When some of the blade nodes among the multiple blade nodes are the CPU computing nodes and some of the blade nodes are the CXL memory expansion nodes, the CXL Switch board, upon receiving the data to be stored sent by the CPU computing nodes through the CXL protocol, routes the data to be stored to the CXL memory expansion nodes in sequence for storage according to the priority of each of the CPU computing nodes that sends the data to be stored.

[0070] In the above embodiment, the method for determining the priority of the CPU computing node includes but is not limited to: Performance index: Determine the priority according to the performance index of the CPU computing node (for example, the number of cores, clock frequency, cache size, etc.), and the CPU node with higher performance has a higher priority. Load balancing: Dynamically adjust the priority according to the current load of the CPU computing node, and the node with lower load has a higher priority. Resource allocation: Determine the priority according to the corresponding available resources of the CPU computing node (for example, memory, disk space, etc.), and the node with sufficient resources has a higher priority. Fault tolerance: Determine the priority according to the fault tolerance capability of the CPU computing node. Data access frequency: Determine the priority according to the frequency of access to the data generated by the CPU computing node. High-frequency access data should have a higher priority so that it can be retrieved from the storage system more quickly. Data importance: Determine the priority according to the importance of the data generated by the CPU computing node. For critical business data, set a higher storage priority to ensure fast access and high availability of data. Quality of service (QoS) requirements: Assign storage priority according to the quality of service requirements of the application or service of the CPU computing node. For example, assign a higher priority to services that require low latency and high throughput. Data size: Determine the priority based on the size of the data generated by the CPU computing node. For example, the larger the amount of data generated by the CPU computing node, the higher the priority of the CPU computing node. Data type: Determine the priority of the CPU computing node based on the type of data generated by the CPU computing node. For example, transactional data may require a higher priority than non-transactional data.

[0071] The execution subject of the above steps may be a server, a terminal, etc., but is not limited thereto.

[0072] In an optional embodiment, the method further includes: when the data generated by the CPU computing node to be stored in the CXL memory expansion node includes multiple types of data, storing them in sequence in the following order: operating system data, predetermined application data, security data, system performance monitoring data, user data application logs, backup data, test and development data.

[0073] In the above embodiments, the operating system data includes, but is not limited to: operating system logs, system configuration, driver information, etc. The predetermined application data includes, but is not limited to: application configuration files, application logs, data generated by applications, etc. The security data includes, but is not limited to: security event logs, security policies, user access rights, etc. The system performance monitoring data includes, but is not limited to: CPU usage, memory usage, disk space usage, network traffic, etc. The user data application log includes, but is not limited to: user operation logs, application operation logs, etc. The backup data includes, but is not limited to: regularly backed up system files, application files, databases, etc. The test and development data includes, but is not limited to: test cases, code generated during the development process, logs during the development process, etc.

[0074] In an optional embodiment, before routing the data to be stored to the CXL memory expansion node in sequence according to the priority of each CPU computing node that sends the data to be stored, the method further includes: determining a target memory requirement of the target CPU computing node based on the data to be stored generated by the target CPU computing node within a predetermined period of time, wherein the target CPU computing node is any one of the multiple CPU computing nodes that generate the data to be stored; acquiring and analyzing a target remaining storage space of each of the multiple memories included in the CXL memory expansion node; and determining a target memory for storing the data to be stored generated by the target CPU computing node based on the target memory requirement and the target remaining storage space.

[0075] In the above embodiment, the target memory requirement of the target CPU computing node is determined based on the data to be stored generated by the target CPU computing node within a predetermined time period, including but not limited to: performing data collection: collecting all the data to be stored generated by the target CPU computing node within a predetermined time period, including input data, intermediate results, output data, etc.; performing data classification: classifying the collected data according to type, size, importance, etc.; performing data volume estimation: estimating the magnitude of data in each category; considering redundancy and backup factors: when determining the memory requirement, considering the redundancy and backup requirements of the data, for example, when multiple copies of the data need to be backed up, increase the corresponding memory requirement; performing data growth prediction: considering the trend of data changes over time, counting the data to be stored generated by the target CPU computing node within a predetermined time period, thereby predicting the data changes in future time periods, and adjusting the memory requirement; considering the safety margin: in order to cope with unexpected situations, a certain proportion of the safety margin is added on the basis of the calculated memory requirement to ensure the stable operation of the system.

[0076] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0077] In this embodiment, a blade server system is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "unit" can implement a combination of software and / or hardware of a predetermined function. Although the units described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0078] In an embodiment of the present application, the blade server system includes: a routing unit, which is used to route the data to be stored to the CXL memory expansion node in sequence for storage according to the priority of each CPU computing node that sends the data to be stored, when the CXLSwitch board receives the data to be stored sent by the CPU computing node through the CXL protocol.

[0079] In an optional embodiment, the blade server system further includes: a storage unit for storing, in the following order, when the data generated by the CPU computing node to be stored in the CXL memory expansion node includes multiple types of data: operating system data, predetermined application data, security data, system performance monitoring data, user data application log, backup data, test and development data.

[0080] In an optional embodiment, the blade server system further includes: a first determination unit, for determining a target memory requirement of the target CPU computing node based on the data to be stored generated by the target CPU computing node within a predetermined period of time before routing the data to be stored to the CXL memory expansion node in sequence according to the priority of each CPU computing node that sends the data to be stored, wherein the target CPU computing node is any one of the multiple CPU computing nodes that generate the data to be stored; an acquisition unit, for acquiring and analyzing a target remaining storage space of each of the multiple memories included in the CXL memory expansion node; the first determination unit, for determining a target memory for storing the data to be stored generated by the target CPU computing node based on the target memory requirement and the target remaining storage space.

[0081] It should be noted that the above-mentioned units can be implemented by software or hardware. For the latter, it can be implemented in the following ways, but not limited to this: the above-mentioned units are all located in the same processor; or the above-mentioned units are located in different processors in any combination.

[0082] An embodiment of the present application further provides a computer-readable storage medium, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when running.

[0083] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media that can store computer programs, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk.

[0084] An embodiment of the present application further provides an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0085] In an exemplary embodiment, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0086] An embodiment of the present application further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps in any one of the above method embodiments are implemented.

[0087] For specific examples in this embodiment, reference may be made to the examples described in the above embodiments and exemplary implementation modes, and this embodiment will not be described in detail herein.

[0088] Obviously, those skilled in the art should understand that the above modules or steps of the present application can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order from that herein, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present application is not limited to any specific combination of hardware and software.

[0089] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. A blade server system, characterized in that: include: A chassis housing, a plurality of blade nodes and a computing fast link switch CXL Switch board, wherein the plurality of blade nodes and the CXL Switch board are built in the chassis housing; Each of the blade nodes is directly connected to the CXL Switch board via an orthogonal high-density connector; Wherein, in the case where the plurality of blade nodes include a target blade node that meets the target condition, the target blade node is further connected to the CXL Switch board via a high-density cable connector; The plurality of blade nodes are all central processing unit CPU computing nodes; or, some of the plurality of blade nodes are central processing unit CPU computing nodes, and some of the blade nodes are computing fast link CXL memory expansion nodes; Wherein, in the case where the plurality of blade nodes include a target blade node that meets the target condition, the target blade node is further connected to the CXL Switch board through a high-density cable connector, including: In the case where some of the multiple blade nodes are the CPU computing nodes, and some of the blade nodes are the CXL memory expansion nodes, and the second target data to be stored in the CXL memory expansion node reaches a second threshold and the computing capacity of the CPU computing node reaches a third threshold, the CPU computing node and the CXL memory expansion node are both connected to the CXL Switch board through a high-density cable connector, and the second target data is data generated by the CPU computing node to be stored in the CXL memory expansion node.

2. The blade server system according to claim 1, characterized in that: In the case where a plurality of the blade nodes are the CPU computing nodes, the blade server system further includes: a memory module, wherein the memory module is arranged in the chassis shell, the memory module is connected to a cable socket cable arranged on the CXL Switch board through a high-density cable connector, and the storage method of the memory module includes at least one of the following: an E3.S memory expansion module, a PCIE peripheral component interconnect fast channel memory expansion card, and a memory expansion single board.

3. The blade server system according to claim 1, characterized in that: In the case where the plurality of blade nodes include a target blade node that meets the target condition, the target blade node is further connected to the CXL Switch board through a high-density cable connector, including: When some of the blade nodes among the multiple blade nodes are the CPU computing nodes, and some of the blade nodes are the CXL memory expansion nodes, and first target data to be stored in the CXL memory expansion nodes reaches a first threshold, the CXL memory expansion nodes are further connected to the CXL Switch board through a high-density cable connector, wherein the target data is data generated by the CPU computing nodes to be stored in the CXL memory expansion nodes.

4. The blade server system according to claim 1, characterized in that: In the case where the plurality of blade nodes include a target blade node that meets the target condition, the target blade node is further connected to the CXL Switch board through a high-density cable connector, including: When all of the plurality of blade nodes are the CPU computing nodes and the computing capacity of the CPU computing nodes reaches a fourth threshold, the CPU computing nodes are further connected to the CXL Switch board through a high-density cable connector.

5. The blade server system according to claim 1, characterized in that: Each of the blade nodes is directly connected to the CXL Switch board via an orthogonal high-density connector, including: Each of the blade nodes is provided with a female head of the orthogonal high-density connector, and the CXL Switch board is provided with a male head of the orthogonal high-density connector corresponding to each of the blade nodes. Each of the blade nodes and the CXLSwitch board are directly connected by plugging the female head and the male head belonging to the same orthogonal high-density connector.

6. The blade server system according to claim 5, characterized in that: The CXL Switch board is also provided with: CXL Switch chip and cable socket; The target blade node is also connected to the CXL Switch board via a high-density cable connector, including: The target blade node is connected to a cable socket cable provided on the CXL Switch board through a high-density cable connector.

7. The blade server system according to claim 1, characterized in that: In the case where some of the blade nodes among the multiple blade nodes are the CXL memory expansion nodes, the memory expansion method of the CXL memory expansion node includes at least one of the following: E3.S memory expansion module, PCIE peripheral component interconnect fast channel memory expansion card, memory expansion board.

8. A blade server, characterized in that: A blade server system running any one of claims 1 to 7, wherein the blade server comprises: Power module, fan module, centralized management board; Among them, the centralized management board is used to interact and communicate with the management chip of each blade node, the centralized management board leads out a centralized management interface, the power module leads out a power input interface, each blade node is provided with a first hot-swap handle, each power module is provided with a second hot-swap handle, multiple blade nodes are arranged side by side on the front side of the chassis shell, multiple power modules are arranged side by side below multiple blade nodes, and multiple fan modules, the power input interface, and the centralized management interface are arranged on the rear side of the chassis shell.

9. A CXL memory configuration method, characterized in that: The blade server system applied to any one of claims 1 to 7 comprises: Some of the blade nodes among the multiple blade nodes are the CPU computing nodes, and some of the blade nodes are the CXL memory expansion nodes. When the CXL Switch board receives the data to be stored sent by the CPU computing node through the CXL protocol, it routes the data to be stored to the CXL memory expansion node in sequence for storage according to the priority of each CPU computing node that sends the data to be stored.

10. The CXL memory configuration method according to claim 9, characterized in that: The method further comprises: In the case where the data generated by the CPU computing node to be stored in the CXL memory expansion node includes multiple types of data, the data are stored in the following order: operating system data, predetermined application data, security data, system performance monitoring data, user data application logs, backup data, test and development data.

11. The CXL memory configuration method according to claim 9, characterized in that: Before routing the data to be stored to the CXL memory expansion node in sequence according to the priority of each of the CPU computing nodes that sends the data to be stored, the method further includes: Determine the target memory requirement of the target CPU computing node according to the data to be stored generated by the target CPU computing node within a predetermined period of time, wherein the target CPU computing node is any one of the plurality of CPU computing nodes generating the data to be stored; Acquire and analyze target remaining storage space of each of a plurality of memories included in the CXL memory expansion node; According to the target memory requirement and the target remaining storage space, a target memory for storing the to-be-stored data generated by the target CPU computing node is determined.

12. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program, when executed by a processor, implements the steps of the CXL memory configuration method described in any one of claims 9 to 11.

13. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the CXL memory configuration method described in any one of claims 9 to 11 are implemented.

14. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the CXL memory configuration method described in any one of claims 9 to 11 are implemented.