Server nodes and servers
By setting up a dedicated external channel and adapter board on the server node's chassis, a pluggable connection between the RAID card and the motherboard is achieved, solving the problem of cumbersome component replacement in existing technologies and improving maintenance efficiency.
Patent Information
- Application Number
- CN202522431910.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Utility models(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-14
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2035-11-14
AI Technical Summary
The replacement or maintenance of components in existing server nodes requires cumbersome procedures, resulting in low maintenance efficiency.
A dedicated external channel is set on the server node chassis, through which the RAID card and motherboard connector can be directly plugged in and out, avoiding disassembly of the chassis. The RAID card and storage components are pluggable and hot-swapped using an adapter board and a backup power supply, and hot-swapping is performed through a signal receiver.
It simplifies the connection and disconnection process between the RAID card and the motherboard, improves the maintenance efficiency of server nodes, and reduces operational complexity and time costs.
Smart Images

Figure CN224987708U_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of server technology, and in particular to a server node and a server. Background Technology
[0002] As the core carrier of computing, storage, and network services, the reliability and maintainability of servers directly affect the operational efficiency and business continuity of the entire data center. The operational efficiency of their hardware has become one of the key factors affecting the total cost of ownership and service quality of a data center.
[0003] However, the components within a single server node are all encapsulated in a chassis or similar enclosure. When these internal components need to be replaced due to upgrades, expansions, or malfunctions, maintenance personnel have to go through cumbersome procedures to disassemble them, reducing maintenance efficiency. Utility Model Content
[0004] The purpose of this application is to provide a server node and a server to improve the maintenance efficiency of the server node.
[0005] To achieve the above objectives, embodiments of this application provide the following technical solutions: In a first aspect, embodiments of this application provide a server node, which includes a chassis and a motherboard, storage components, and a RAID card housed within a receiving cavity. The chassis has a receiving cavity and a first opening communicating with the receiving cavity. The motherboard has a first connector, and the RAID card has a second connector and a third connector. The RAID card enters and exits the receiving cavity through the first opening and is pluggably connected to the first connector of the motherboard through the second connector. The third connector is located on the side of the RAID card opposite to the second connector. When the RAID card is plugged into the motherboard, the third connector is exposed in the first opening and is used to connect to the storage components.
[0006] In this embodiment, the server node's enclosure has a first opening that provides a dedicated external channel for the RAID card, allowing maintenance personnel to directly plug and unplug it from outside the enclosure, thus eliminating the cumbersome step of opening the enclosure. The RAID card is pluggably connected to the first connector on the motherboard via its second connector. The RAID card can be removed without touching or damaging the motherboard or its surrounding structure, enabling the RAID card to disconnect from the motherboard without damaging the motherboard structure. The third connector is located on the side of the RAID card facing away from the motherboard, providing a physical connection interface for the storage components. The third connector's exposure to the first opening facilitates operation by maintenance personnel.
[0007] Therefore, maintenance personnel do not need to disassemble the chassis when maintaining the RAID card connected to the motherboard. They can connect and disconnect the RAID card from the motherboard through the first opening, thereby improving the maintenance efficiency of the server node, solving the problem of low disassembly and assembly efficiency of internal components of the server node, and improving the maintenance efficiency of the server node.
[0008] In one possible implementation, the server node in this embodiment further includes an adapter board disposed in the receiving cavity. The adapter board is connected to the storage component, and a third connector is used for pluggable connection with the adapter board. Since the adapter board and the third connector are pluggable, when maintenance personnel disassemble the RAID card, they only need to disconnect this connection component from the third connector to separate the RAID card from the storage component.
[0009] Another possible implementation is that the enclosure has a second opening communicating with the receiving cavity, the adapter board has a fourth connector exposed in the second opening, and the fourth connector is pluggably connected to the third connector via a first connecting cable. By setting the second opening, exposing the fourth connector in the second opening, and using an external pluggable first connecting cable for connection, the connection or disconnection of the RAID card and the adapter board can be done outside the server enclosure.
[0010] In another possible implementation, in this embodiment, the server node includes a backup power supply housed within a cavity. The backup power supply is connected to an adapter board and a motherboard. The adapter board has a fifth connector exposed in a second opening, and the RAID card has a sixth connector exposed in a first opening. The fifth and sixth connectors are detachably connected via a second connecting cable. The backup power supply can provide temporary power to the RAID card when the connection between the second and first connectors is disconnected, ensuring that signals not yet written to storage components in the RAID card's cache can be transmitted. Simultaneously, the backup power supply is integrated into the adapter board, connecting to the RAID card via the adapter board, facilitating the circuit layout of the server node.
[0011] Another possible implementation is to have the first and second openings facing the same direction. Maintenance personnel can then perform plugging and unplugging operations on the RAID card and all its external cables from a fixed location, eliminating the need to navigate to different sides of the server. This simplifies the process and reduces unnecessary movement and repositioning time during maintenance.
[0012] In another possible implementation, the server node in this embodiment further includes a signal receiver connected to the RAID card and exposed through a first opening. When the signal receiver is in a first operating state, the motherboard sends a first control signal to the RAID card and / or a second control signal to the backup power supply. The RAID card responds to the first control signal and transmits cached information to the storage components via a third connector and an adapter board. The backup power supply responds to the second control signal and transmits a power supply signal to the RAID card via a fifth connector and a sixth connector. After receiving a plug-in / plug-out signal, the motherboard sends the first control signal to the RAID card to release the storage information between them. The motherboard also sends the second control signal to the backup power supply to enable the backup power supply to start supplying power to the RAID card. The RAID card can then continue to transmit temporary cached information to the storage components. Maintenance personnel can send instructions to the motherboard via the signal receiver to perform hot-plug operations on the RAID card.
[0013] Another possible implementation is that the storage components and RAID card are located on one side of the motherboard. When maintenance operations are required (such as simultaneously checking or replacing the RAID card or its associated storage components), maintenance personnel do not need to go to different sides of the server or adjust the device's position, making it convenient for maintenance personnel to perform operations on the same side.
[0014] Another possible implementation is that the motherboard is located between the RAID card and the storage components, forming two independent operating surfaces on opposite walls of the enclosure. When maintenance of both the RAID card and the storage components is required simultaneously, maintenance personnel can operate synchronously from opposite sides of the enclosure without obstructing each other or causing tool interference.
[0015] Another possible implementation involves a storage component comprising a hard drive backplane and at least one hard drive mounted on the backplane. A third connector is used to connect to the hard drive backplane, thus freeing the RAID card from the limitation of its own physical interfaces in terms of the number of hard drives it can manage and drive. By connecting the hard drive backplane to the adapter board, the storage capacity and expansion potential of a single server node are increased without replacing the RAID card.
[0016] Another possible implementation is that the insertion and removal directions of the first and second connectors are consistent with the direction in which the RAID card enters and exits the receiving cavity through the first opening. When the maintenance personnel remove or install the RAID card, they can directly connect and disconnect the first and second connectors. Thus, the maintenance personnel can directly operate the RAID card from the first opening to drive the insertion and removal of the second connector and the first connector without having to go deep into the receiving cavity.
[0017] Another possible implementation is that one of the first and second connectors is an OCP gold finger interface, and the other of the first and second connectors is an OCP slot corresponding to the OCP gold finger interface. Using the open computing OCP standard interface allows the RAID card to function as a standardized, interchangeable component, improving the versatility of the RAID card.
[0018] Secondly, embodiments of this application also provide a server, which includes a server node in any of the above possible implementations, a connected computing node and a storage node, wherein the computing node and / or storage node includes the server node, so as to facilitate operation and maintenance within a single server node.
[0019] Another possible implementation is that the server includes multiple server nodes arranged along a first direction, and the enclosure includes a surrounding panel arranged around the first direction. A first opening is provided in the surrounding panel, and the first opening of each server node can remain exposed. Maintenance personnel can directly disassemble and install internal components from the side of each server node without first removing or disassembling its adjacent server nodes.
[0020] The technical effects of any implementation method provided in the second aspect above can be found in the technical effects of any implementation method provided in the first aspect above, and will not be repeated here. Attached Figure Description
[0021] Figure 1 A three-dimensional structural diagram of the server provided in an embodiment of this application; Figure 2 for Figure 1 A schematic diagram of a single server node in the system; Figure 3 for Figure 2 One of the schematic diagrams of the internal component connection structure of a single server node in a computer. Figure 4 for Figure 2 The second schematic diagram of the internal component connection structure of a single server node in the diagram; Figure 5 for Figure 2 The third schematic diagram of the internal component connection structure of a single server node in the diagram.
[0022] Figure label: 1-Server; 100-Server Node; 101-Receiving Cavity; 102-Enclosure; 1021-Enclosure Panel; 1021a-First Opening; 1022b-Second Opening; 103-Motherboard; 1031-Central Processing Unit; 1032-First Circuit Board; 104-First Connector; 105-RAID Card; 1051-Second Circuit Board; 1052-Processing Chip; 106-Second Connector; 107-Third Connector; 108-Fourth Connector; 109-Fifth Connector; 110-Sixth Connector; 113-Adapter Board; 1131-Third Circuit Board; 114-Signal Control Line; 115-Backup Power Supply; 116-Storage Component; 1161-Hard Drive Backplane; 1162-Hard Drive; 117-Stop Plate; 118-Signal Receiver; 119-First Connecting Cable; 120-Second Connecting Cable; 121-Backup Power Supply Connector. Detailed Implementation
[0023] The technical solutions of the embodiments of this application will be described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.
[0024] Hereinafter, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature.
[0025] Furthermore, in the embodiments of this application, directional terms such as "upper," "lower," "horizontal," and "vertical" are defined relative to the orientation of the components shown in the accompanying drawings. It should be understood that these directional terms are relative concepts, used for relative description and clarification, and can change accordingly depending on the orientation of the components in the accompanying drawings.
[0026] In the embodiments of this application, unless otherwise explicitly specified and limited, the term "connection" should be interpreted broadly. For example, "connection" can be a fixed connection, an electrical connection, a detachable connection, or an integral part; it can be a direct connection or an indirect connection through an intermediate medium.
[0027] In some embodiments, see Figure 1 This application provides a server 1 to perform functions such as calculation, data processing, information storage and transmission.
[0028] Server 1 includes multiple interconnected server nodes 100. Server 1 processes data through multiple server nodes 100, wherein the multiple server nodes 100 can be composed of server nodes 100 with different functional types.
[0029] For example, the server node 100 mentioned in the embodiments of this application may be a computing node that undertakes data processing tasks.
[0030] As another example, the server node 100 mentioned in this application embodiment may be a storage node responsible for data storage.
[0031] It should be noted that the server nodes 100 mentioned in the embodiments of this application can be all compute nodes or all storage nodes. Of course, multiple server nodes 100 can be partially compute nodes or partially storage nodes. This application does not limit this.
[0032] It should be understood that the number of server nodes 100 is not specifically limited in the embodiments of this application. The term "multiple" as used herein should be interpreted as two or more, and should not be construed as an upper limit or exact value limitation on the number of server nodes 100 in the embodiments of this application.
[0033] When multiple server nodes 100 are all compute nodes, the connected compute nodes can provide the computing power of server 1. When multiple server nodes 100 include connected compute nodes and storage nodes, the connected compute nodes and storage nodes can provide integrated computing and storage capabilities for server 1.
[0034] It should be understood that the embodiments of this application do not limit the specific connection method and topology between compute nodes and storage nodes. The connection methods of the embodiments of this application include, but are not limited to, one storage node connecting to one compute node, one storage node connecting to multiple compute nodes, or multiple storage nodes connected to multiple compute nodes in any topology. The above specific examples should not be construed as limiting the scope of protection of the embodiments of this application.
[0035] In some embodiments, see Figure 1 and combined Figure 2 In this embodiment, server 1 includes multiple server nodes 100 arranged along a first direction. For ease of explanation, a coordinate system is established in this embodiment. Please refer to [link to relevant documentation] for further details. Figure 1 This application establishes a coordinate system for the three-dimensional structure diagram of server 1. Figure 1 The X direction shown is the depth direction of server 1. Figure 1 The Y direction shown is the width direction of server 1. Figure 1 The Z direction shown is the height direction of server 1.
[0036] For example, the first direction is the height direction of server 1, that is... Figure 1 The Z direction is shown in the diagram.
[0037] The following embodiments of this application describe a single server node 100, which can be a storage node or a computing node.
[0038] In some embodiments, see Figure 2 and combined Figure 3 The server node 100 includes a motherboard 103, on which a central processing unit 1031 (CPU) is provided to process the data of the server node 100.
[0039] For example, the computing node includes a motherboard 103 and a central processing unit 1031 for performing main data processing and business logic operations, and is responsible for running the operating system and applications to provide computing power to the outside world.
[0040] In another example, the storage node includes a motherboard 103 to manage the data within the storage node and control tasks such as storage input or output requests from the compute node, thereby fulfilling the function of the storage node in providing data storage services.
[0041] In some embodiments, see Figure 2 and combined Figure 3 In this embodiment of the application, the server node 100 also includes a RAID card 105 (Redundant Array of Independent Disks card). The RAID card 105 is connected to the motherboard 103 to perform the work instructed by the motherboard 103 in order to meet the storage management needs of the server node 100.
[0042] In some embodiments, see Figure 2 and combined Figure 3 The motherboard 103 is provided with a first connector 104, and the RAID card 105 is provided with a second connector 106. In this embodiment of the application, data transmission between the motherboard 103 and the RAID card 105 can be realized through the first connector 104 and the second connector 106.
[0043] For example, the first connector 104 and the second connector 106 can be PCIe connectors. The first connector 104 is a PCIe slot on the motherboard, and the second connector 106 is a PCIe gold finger on the RAID card 105. Of course, the first connector 104 is a PCIe gold finger on the motherboard, and the second connector 106 is a PCIe slot on the RAID card 105. This application does not limit this.
[0044] As another example, the first connector 104 and the second connector 106 can also be MCIO connectors. The first connector 104 connects to the motherboard 103 via an MCIO cable, and the second connector 106 connects to the RAID card 105 via an MCIO cable.
[0045] The first connector 104 and the second connector 106 can also be other types of connectors, which are not limited in this application. It is understood that the first connector 104 and the second connector 106 form a communication channel through pins on their corresponding gold fingers to transmit information. The number of pins on the gold fingers can specify the protocol between them and the type of information transmitted.
[0046] It should be understood that the embodiments of this application can meet the data transmission requirements between the RAID card 105 and the motherboard 103 under different circumstances by selecting different first connectors 104 and second connectors 106.
[0047] It should be noted that the configuration of the first connector 104 on the motherboard 103 in this embodiment is not limited. The two can be connected via a circuit board, a cable, or other electrical connection methods. Similarly, the configuration of the second connector 106 on the RAID card 105 in this embodiment is not limited. The two can be connected via a circuit board, a cable, or other electrical connection methods. For ease of explanation, this embodiment uses an example where the motherboard 103 has a first circuit board 1032, a central processing unit 1031 is mounted on the first circuit board 1032, and the first connector 104 and the central processing unit 1031 are electrically connected via the first circuit board 1032.
[0048] Similarly, this application embodiment does not limit the connection method between the second connector 106 and the RAID card 105. The two can be connected by a circuit board, a cable, or other electrical connection methods. For ease of explanation, this application embodiment takes as an example that the RAID card 105 is provided with a second circuit board 1051, and the second connector 106 and the RAID card 105 are electrically connected through the second circuit board 1051.
[0049] For example, a processing chip 1052 is integrated on the second circuit board 1051. The processing chip 1052 is used to perform data operations, signal conversion or protocol processing, etc., corresponding to the PCIe expansion function, so as to enhance the specific application capabilities of the module.
[0050] In some embodiments, see Figure 2 and combined Figure 3In this embodiment of the application, the server node 100 also includes a housing 102, which can serve as the physical support and protection structure for the entire server node 100.
[0051] It should be understood that the housing 102 in the embodiments of this application can be a closed or semi-closed hollow shell, or an open frame structure. This application does not limit it in this way.
[0052] In some embodiments, see Figure 2 and combined Figure 3 In this embodiment of the application, the housing 102 has a receiving cavity 101, and the motherboard 103 is housed in the receiving cavity 101. That is, the central processing unit 1031, the first circuit board 1032 and the first connector 104 are housed in the receiving cavity 101.
[0053] For example, the central processing unit 1031 is installed on the preset pad area of the first circuit board 1032 by spot welding, and the first connector 104 is fixed to the corresponding interface area of the first circuit board 1032 by welding (such as reflow soldering). The first connector 104 and the central processing unit 1031 are electrically connected through the traces on the first circuit board 1032.
[0054] In some embodiments, see Figure 2 and combined Figure 3 When the RAID card 105 is connected to the motherboard 103 and is working, the RAID card 105 is housed in the receiving cavity 101, and the second connector 106 provided on the RAID card 105 is housed in the receiving cavity 101 and connected to the first connector 104. The RAID card 105 and the second connector 106 are connected through the second circuit board 1051.
[0055] For example, the RAID card 105 is installed on the preset pad area of the second circuit board 1051 by spot soldering, and the second connector 106 is fixed to the corresponding interface area of the second circuit board 1051 by soldering (such as reflow soldering). The second connector 106 and the central processing unit 1031 are electrically connected through the traces on the second circuit board 1051.
[0056] In some embodiments, see Figure 2 and combined Figure 3 The enclosure 102 has a first opening 1021a that communicates with the receiving cavity 101. The RAID card 105 enters and exits the receiving cavity 101 through the first opening 1021a, and is pluggably connected to the first connector 104 of the motherboard 103 through the second connector 106.
[0057] It should be noted that the RAID card 105 enters and exits the receiving cavity 101 through the first opening 1021a means that the plane perpendicular to the direction of the first opening 1021a pointing towards the first connector 104 is the first reference plane, that is... Figure 1 In a plane perpendicular to the X-direction, the orthographic projection of the RAID card 105 onto the first reference plane lies within the projection of the first opening 1021a onto the first reference plane. The outline of the RAID card 105 on this first reference plane is completely surrounded by the outline of the first opening 1021a on the first reference plane, thereby allowing the structure formed by the RAID card 105 to be removed from the first opening 1021a. Furthermore, along the installation and removal path of the RAID card 105 via the first opening 1021a, there are no parts that may cause mechanical interference between the RAID card 105 and the internal structure or other fixed structures of the enclosure 102.
[0058] For example, the first connector 104 and the second connector 106 can be pluggable via cable connection, providing greater wiring freedom for the internal layout of the housing 102.
[0059] As another example, the first connector 104 and the second connector 106 are directly connected by a mating connection to achieve pluggability, thereby improving signal transmission quality and saving internal space.
[0060] It should be understood that the embodiments of this application do not limit the specific connection implementation between the first connector 104 and the second connector 106, and any appropriate method that can achieve pluggable connection between the two should be included within the protection scope of this application.
[0061] This application does not limit the separation method of the first connector 104 and the second connector 106. In some embodiments, the first connector 104 and the second connector 106 adopt a passive separation design. When the maintenance personnel remove the RAID card 105 from the first opening 1021a, the second connector 106 automatically disconnects from the first connector 104 as the RAID card 105 moves. In other embodiments, the first connector 104 and the second connector 106 are designed to allow active separation. That is, after the RAID card is removed, the second connector 106 remains connected to the first connector 104, and the maintenance personnel can manually disconnect the connection according to the actual situation. This application does not limit the specific separation structure of the first connector 104 and the second connector 106. Any appropriate method that can achieve effective separation of the two should be included within the protection scope of this application.
[0062] In some embodiments, see Figure 2 and combined Figure 3This application embodiment is illustrated by the example of the second connector 106 of the RAID card 105 automatically disconnecting from the first connector 104 as the RAID card 105 moves. The RAID card 105 is located on the side of the first connector 104 facing the first opening 1021a, and the second connector 106 is located between the first opening 1021a and the first connector 104. The RAID card 105 and the first connector 104 can be pluggably connected in a first direction via the second connector 106. The RAID card 105 and the second connector 106 can enter and exit the receiving cavity 101 in the first direction through the first opening 1021a, so that the second connector 106 and the first connector 107 are pluggable.
[0063] In some embodiments, the first opening 1021a is disposed opposite to the first connector 104, the direction of the first opening 1021a pointing to the first connector 104 is the first direction, and the insertion and removal direction of the second connector 106 and the first connector 104 is parallel to the direction from the first opening 1021a to the first connector 104.
[0064] For example, the direction in which the first opening 1021a points to the first connector 104 is... Figure 1 The X direction shown is the depth direction of server 1.
[0065] It should be noted that the above statement that "the insertion and removal direction of the second connector 106 and the first connector 104 is parallel to the direction from the first opening 1021a to the first connector 104" means that the insertion and removal direction of the second connector 106 and the first connector 104 is basically consistent with the direction from the first opening 1021a to the first connector 104. There may be slight angular deviations or local slight displacements due to manufacturing process errors or wiring space limitations, but there is no obvious intersection or large directional deviation.
[0066] For example, the front panel of server node 100 has a first opening 1021a.
[0067] As another example, the top cover of server node 100 has a first opening 1021a.
[0068] When installation or maintenance is required, the second connector 106 is plugged into and unplugged from the first connector 104 along a straight path. The direction of this plugging and unplugging action is the same as the direction from the external first opening 1021a directly to the first connector 104 inside the enclosure 102. This allows maintenance personnel to operate the RAID card 105 and its connector from the outside of the enclosure 102 along a single, straight direction through the first opening 1021a, thereby simplifying maintenance operations.
[0069] In some embodiments, see Figure 2 and combined Figure 3 The enclosure 102 includes a surrounding panel 1021 arranged around a first direction, and a first opening 1021a is provided in the surrounding panel 1021. The first opening 1021a of each server node 100 can remain exposed, and maintenance personnel can directly remove and install internal components from the side of each server node 100 without first removing or disassembling its adjacent server nodes 100.
[0070] For example, the enclosure 1021 may be the front panel of the housing 102 arranged along the second direction.
[0071] It should be noted that the "front panel" in this embodiment can be understood as the "maintenance side panel" in actual operation—when multiple server nodes 100 are deployed closely side by side in a rack, the front panel of each node faces the maintenance channel (i.e., the "front" side of all nodes maintains a uniform orientation), and the first opening 1021a is set on this panel. At this time, maintenance personnel standing in the maintenance channel can directly disassemble and install internal components through the first opening 1021a of the front panel of each node without moving adjacent nodes.
[0072] Maintenance personnel can access the interior of the receiving cavity 101 through the first opening 1021a without disassembling the chassis, eliminating the need for complete disassembly during maintenance. Simultaneously, the second connector 106 is used for plugging and unplugging with the first connector 104, and the plugging and unplugging direction is consistent with the direction of the first opening 1021a pointing towards the first connector 104. Maintenance personnel can apply force to the RAID card 105 in a single direction through the first opening 1021a, causing the RAID card 105 to move the second connector 106, thereby enabling the plugging and unplugging of the second connector 106 and the first connector 104, thus allowing the separation or connection of the RAID card 105 to the central processing unit 1031.
[0073] Furthermore, the RAID card 105 is located on the side of the first connector 104 facing the first opening 1021a, and the second connector 106 is located between the first opening 1021a and the first connector 104. Maintenance personnel can directly operate the RAID card 105 from the first opening 1021a to insert and remove the second connector 106 from the first connector 104 without needing to penetrate the receiving cavity 101. The projections of the second connector 106 and the RAID card 105 onto the first reference plane are located within the projection of the first opening 1021a onto the first reference plane, ensuring that the dimensions of the RAID card 105 and the second connector 106 allow maintenance personnel to remove them from the first opening 1021a.
[0074] It should be noted that maintenance personnel can choose the operating method according to the actual situation. For example, they can directly use their hands to grasp the corresponding gripping part of the RAID card 105 at the first opening 1021a to insert or remove it. As another example, when the module is located deep or is inconvenient to grip, a tool can be used to reach into the receiving cavity 101 to assist in removing it.
[0075] In some embodiments, see Figure 2 and combined Figure 3 The second connector 106 is a gold finger mounted on the edge of the second circuit board 1051. Correspondingly, the first connector 104 is a slot fixed on the first circuit board 1032 and matching the gold finger, with its socket facing the second connector 106 to achieve mating.
[0076] In some embodiments, see Figure 2 and combined Figure 3 In this embodiment of the application, the server node 100 further includes a third connector 107. The third connector 107 is connected to the second circuit board 1051 of the RAID card 105 and is connected to the second connector 106 through the second circuit board 1051. The third connector 107 is located on the side opposite to the second connector 106 of the RAID card 105. When the RAID card 105 is plugged into the motherboard 103, the third connector 107 is exposed to the first opening 1021a. The third connector 107 is used to connect to the storage component 116.
[0077] It should be noted that in this embodiment, the third connector 107 being exposed at the first opening 1021a means that the mating surface of the third connector 107 is not physically obstructed at the first opening 1021a, allowing connection to external devices. Regardless of whether the mating surface of the third connector 107 is suspended inside the receiving cavity 101, flush with the plane of the first opening 1021a, or protrudes from the housing 102, as long as the above-mentioned unobstructed condition is met, it falls within the scope defined in this embodiment.
[0078] Thus, the third connector 107 can serve as the interface between the RAID card 105 and the storage component 116 located in the housing cavity 101, and at the same time, it forms a path with the second connector 106 through the RAID card 105 to ensure that the signal or power transmission between it and the central processing unit 1031 is uninterrupted.
[0079] Of course, the third connector 107 can also serve as an interface between the RAID card 105 and the external hard drive to meet the expansion needs of the server node 100.
[0080] The third connector 107 is exposed to the first opening 1021a, allowing it to directly interface with external components without disassembling the enclosure 102, providing a physical basis for the RAID card 105 to connect to other components via the third connector 107 to perform additional actions.
[0081] When removing or installing the RAID card 105, maintenance personnel only need to insert their fingers into the receiving cavity 101 through the first opening 1021a to connect it to the central processing unit 1031 (i.e., operate the second connector 106), while all external connections are made through the exposed third connector 107, reducing the operational difficulty for maintenance personnel.
[0082] Similarly, this application does not limit the specific type of the third connector 107. The third connector 107 may be an RJ45 interface or fiber optic interface for network communication, a SATA or SAS interface for storage expansion, a USB or Type-C interface for general data transmission, or a DC power supply interface for power supply. Any interface known to those skilled in the art that can achieve the required electrical connection function is applicable to the embodiments of this application.
[0083] For example, the third connector 107 is connected to the data connector via the second circuit board 1051.
[0084] In some embodiments, see Figure 2 and combined Figure 3 The projection of the third connector 107 on the first reference surface falls within the projection of the first opening 1021a on the first reference surface, so that the third connector 107 can move with the RAID card 105.
[0085] In some embodiments, see Figure 2 and combined Figure 3 In this embodiment of the application, the storage component 116 is housed within the receiving cavity 101. The storage component 116 includes a hard disk backplane 1161 and at least one hard disk 1162 disposed on the hard disk backplane 1161. A third connector 107 is used to connect to the hard disk backplane 1161. The hard disk backplane 1161 provides a unified power management and data exchange interface for multiple hard disks 1162, forming a centralized storage module.
[0086] When replacing the RAID card 105, maintenance personnel only need to disconnect the single connection between the third connector 107 and the hard drive backplane 1161 to separate the RAID card 105 from the entire storage system. This eliminates the need to operate each hard drive 1162 individually, simplifying the maintenance process.
[0087] In some embodiments, see Figure 2 and combined Figure 3In this embodiment of the application, the server node 100 further includes a first connecting cable 119, a fourth connector 108, and an adapter board 113. The adapter board 113 is connected to the storage component 116. Specifically, the adapter board 113 is connected to the hard disk backplane 1162. At least a portion of the fourth connector 108 and the adapter board 113 are housed within the receiving cavity 101. The adapter board 113 is connected between the fourth connector 108 and the storage component 116. The adapter board 113 can connect the first connecting cable 119 between the fourth connector 108 and the third connector 107. The first connecting cable 119 and the third connector 107 are pluggable.
[0088] For example, the third connector 107 and the fourth connector 108 may be SlimSAS connectors to enable high-speed transmission of storage signals received by the RAID card 105.
[0089] For example, the adapter board 113 is a floating connector. One end of the adapter board 113 is connected to a fourth connector 108, and the other end of the adapter board 113 is provided with multiple interfaces for connection to the storage component 116. Each hard drive in the storage component 116 can be connected to one interface.
[0090] In another exemplary embodiment, the adapter plate 113 includes a third circuit board 1131 disposed within the receiving cavity 101, and a fourth connector 108 connected to the third circuit board 1131 disposed within the adapter plate 113 in the receiving cavity 101. That is, the fourth connector 108 and the third circuit board 1131 are mounted to a predetermined position on the third circuit board 1131 by soldering or using a plug-in connector, and signal transmission is completed using the wiring of the third circuit board 1131. Specifically, the third connector 107 has an interface facing the first opening 1021a, and the end of the first connecting cable 119 has a pluggable connector adapted to the third connector 107. During connection, the connector of the first connecting cable 119 is aligned with the interface of the third connector 107 to complete the connection. During separation, the first connecting cable 119 is separated from the third connector 107 by applying external force to the first connecting cable 119.
[0091] Therefore, the RAID card 105 can be separated from the entire storage unit 116 through only one cable interface. Furthermore, maintenance personnel do not need to touch or operate the fourth connector 108 and the adapter board 113 when replacing or repairing the RAID card 105, simplifying the maintenance process.
[0092] It should be noted that the present application does not impose specific limitations on the arrangement of the connection wiring between the adapter board 113 and the hard disk backplane 1161 in the physical space. The wiring route can be flexibly selected. Any path design that can achieve the above functions, whether along the outside, inside or mixed path of the enclosure 102, falls within the scope of protection of the present application.
[0093] In some other embodiments, the housing 102 is also provided with a second opening 1022b communicating with the receiving cavity 101, and the fourth connector 108 is exposed in the second opening 1022b. Since the fourth connector 108 and the third connector 108 are pluggable and detachable through the first connecting cable 119, the fourth connector 108 being exposed in the second opening 1022b makes it easy for maintenance personnel to access the fourth connector 108 for operation.
[0094] It should be noted that, in the embodiments of this application, "the fourth connector 108 is exposed to the second opening 1022b" means that there is no physical obstruction between the fourth connector 108 and the external operating environment, and maintenance personnel can directly observe and operate the fourth connector 108 through the second opening 1022b without removing or avoiding any structural components. For example, the fourth connector 108 is completely housed within the receiving cavity 101. In another example, the fourth connector 108 is partially housed within the receiving cavity 101, and partially extends out of the receiving cavity 101 through the second opening 1022b.
[0095] It should be understood that the embodiments of this application do not specifically limit the fit between the size of the second opening 1022b and the external dimensions of the adapter plate 113. In some alternative embodiments, the size of the second opening 1022b is configured to allow the adapter plate 113 as a whole to enter and exit the receiving cavity 101 through the second opening 1022b. The adapter plate 113 itself also becomes an independently maintainable unit that can be installed and removed through the second opening 1022b, improving the flexibility and modularity of server node 100 maintenance.
[0096] In some embodiments, the first opening 1022a and the second opening 1022b face the same direction.
[0097] The phrase "the first opening 1022a and the second opening 1022b face the same direction" means that the walls of the box 102 containing these two openings are parallel or substantially coplanar, so that the normal directions of the openings are the same or approximately the same.
[0098] For example, the first opening 1022a and the second opening 1022b are both provided on the front panel of the housing 102 arranged along the second direction.
[0099] Maintenance personnel can operate all related components (RAID card 105 and connecting cables) from the same location without having to go to different sides of server node 100.
[0100] It should be noted that, in the embodiments of this application, the first opening 1022a and the second opening 1022b can be two independent openings, or they can be merged into a continuous large opening region. This application does not limit this.
[0101] In other embodiments, the first opening 1022a and the second opening 1022b may also be respectively provided on different walls of the enclosure 102. Maintenance personnel can approach and operate from different directions.
[0102] In some embodiments, the adapter plate 113 is provided with a third circuit board 1131 housed in the receiving cavity 101, and the storage component 116 is connected to the adapter plate 113 via the third circuit board 1131.
[0103] For example, storage component 116 can be connected to third circuit board 1131 via SlimSAS (Slim Small Form Factor Serial Attached SCSI) connector, thereby enabling high-speed transmission of storage signals.
[0104] In some embodiments, see Figure 2 and combined Figure 3 The adapter board 113 may also include a chip for processing signals, which is connected to the third circuit board 1131 and is responsive to signals from the RAID card 105 and the storage unit 116.
[0105] In some embodiments, see Figure 2 and combined Figure 3 In this embodiment of the application, the server node 100 further includes a backup power supply 115, a fifth connector 109, and a sixth connector 110. The backup power supply 115 is housed in a receiving cavity 101 and is connected to an adapter board 113 and a central processing unit 1031. The fifth connector 109 is connected to the adapter board 113 and is connected to the backup power supply 115 through the adapter board 113. The sixth connector 110 is connected to a RAID card 105 and is connected to a second connector 106 through the RAID card 105. The sixth connector 110 is exposed in a first opening 1021a.
[0106] For example, the embodiments of this application also include a second connecting cable 120, which is connected between the fifth connector 109 and the sixth connector 110, and the sixth connector 110 and the fifth connector 109 are detachably connected via the second connecting cable 120.
[0107] For example, the backup power supply 115 can be a supercapacitor.
[0108] When an unexpected disconnection occurs between the RAID card 105 and the central processing unit 1031 (such as a line fault or loose interface) or when a plugging or unplugging operation is performed, the central processing unit 1031 can trigger the backup power supply 115 to start, and continuously supply power to the RAID card 105 through the adapter board 113, ensuring that the cached information in the RAID card 105 can be stably transmitted to the storage component 116, and avoiding data loss.
[0109] The sixth connector 110 is exposed to the first opening 1021a, which allows maintenance personnel to directly plug and unplug the connecting parts of the sixth connector 110, thereby facilitating the plug and unplugging of the RAID card 105.
[0110] It should be noted that in this embodiment, the sixth connector 110 being exposed at the first opening 1021a means that the mating surface of the sixth connector 110 is not physically obstructed at the first opening 1021a, allowing connection to external devices. Regardless of whether the mating surface of the sixth connector 110 is suspended inside the receiving cavity 101, flush with the plane of the first opening 1021a, or protrudes from the housing 102, as long as the above-mentioned unobstructed condition is met, it falls within the scope defined in this embodiment.
[0111] For example, please see Figure 3 and combined Figure 4 The fifth connector 109 and the sixth connector 110 can also be SlimSAS connectors.
[0112] In some embodiments of this application, the adapter board 113 is connected to a backup power connector 121, which is used to connect to a backup power supply 115 so that the backup power supply 115 can supply power to the RAID card 105.
[0113] In some embodiments, see Figure 2 and combined Figure 3 In this embodiment of the application, the server node 100 further includes a signal receiver 118, which is connected to the RAID card 105 and can be in multiple operating states. The signal receiver 118 can be connected to the enclosure 102 and connected to the RAID card 105 via a cable. Alternatively, it can be separated from the enclosure 102 and remotely connected to the RAID card 105, allowing maintenance personnel to remotely input commands to the RAID card 105 through the signal receiver 118.
[0114] In other embodiments, the signal receiver 118 is connected to the motherboard 105 to directly input control signals to the motherboard 105.
[0115] For ease of explanation, this application embodiment takes the example of a signal receiver 118 being connected to a RAID card 105 via a cable for illustrative purposes. The signal receiver 118 is exposed in a first opening 1021a, and the RAID card 105 responds to the insertion / removal indication of the signal receiver 118.
[0116] For example, the multiple operating states include a first operating state, which is receiving a hot-plug instruction for the RAID card 105. That is, the maintenance personnel transmit a signal for hot-plugging to the central processing unit through the signal receiver 118.
[0117] For example, the multiple operating states include a second operating state, which indicates that the RAID card 105 is pluggable.
[0118] As another example, the multiple operation states include a third operation state, which is a completed plugging / unplugging state.
[0119] In some embodiments, see Figure 2 and combined Figure 3 When the signal receiver 118 is in the first operating state, the motherboard 103 sends a first control signal to the RAID card 105, and the RAID card 105 responds to the first control signal and transmits cache information to the storage unit 116 through the third connector 107.
[0120] Specifically, when the signal receiver 118 is in the first operating state, the maintenance personnel send a request to insert or remove the RAID card 105 through the signal receiver 118. The signal receiver 118 transmits the command at this time through the traces on the second circuit board 1051. The RAID card 105 transmits the command information through the sixth connector 110, the second connecting cable 120 and the fifth connector 109, and transmits the command information to the adapter board 113 through the traces on the third circuit board 1131. The adapter board 113 transmits the command information to the central processing unit 1031 on the motherboard 103. The central processing unit 1031 sends a first control signal to the RAID card 105 through the first connector 104 and the second connector 106.
[0121] RAID card 105 is used to respond to the first control signal and transmit cached information to storage component 116 through third connector 107 and connection component. Specifically, after receiving the first control signal, RAID card 105 cooperates with central processing unit 1031 to release the electrical signal connection between first connector 104 and second connector 106, thereby releasing the cached information between first connector 104 and second connector 106 and preparing for physical insertion and removal. At this time, RAID card 105 can drive second connector 106 to be inserted or removed.
[0122] For example, a first signal interface is connected to the third circuit board 1131, and a second signal interface is connected to the first circuit board 1032. The first signal interface and the second signal interface are connected through a signal control line 114 so that the adapter board 113 and the main board 103 can transmit signals through the signal control line 114.
[0123] In other embodiments, please refer to Figure 2 and combined Figure 3 When the signal receiver 118 is in the first operating state, the motherboard 103 is also used to send a second control signal to the backup power supply 115, and the backup power supply 115 responds to the second control signal to supply power to the RAID card 105.
[0124] For example, the backup power supply 115 is used to transmit power supply signals to the RAID card 105 in response to the second control signal, via the fifth connector 109 and the sixth connector 110.
[0125] It should be noted that the motherboard 103 sending the first control signal to the RAID card 105 and the motherboard 103 sending the second control signal to the backup power supply 115 can be performed simultaneously or sequentially, and this application does not limit this.
[0126] In this way, when the connection between the first connector 104 and the second connector 106 is broken, the backup power supply 115 starts to supply power to the RAID card 105. Specifically, the backup power supply 115 transmits the power supply signal to the adapter board 113 through the backup power connector 121 and the traces on the third circuit board 1131. The power supply signal is transmitted to the RAID card 105 through the fifth connector 109 and the sixth connector 110. The RAID card 105 can still continue to work, so that the temporary storage information in the RAID card 105 that could not be transmitted to the storage component 116 is transmitted to the storage component 116 during the time when the maintenance personnel plug and unplug the device.
[0127] In some embodiments, the third connector 107 and the fourth connector 108 are used to transmit a first signal at a first rate, and the sixth connector 110 and the fifth connector 109 are used to transmit a second signal at a second rate, wherein the first rate is greater than the second rate; the second signal includes a signal generated by the RAID card 105 in response to a plug-in / plug-out instruction.
[0128] The first signal includes a high-speed storage signal transmitted from the RAID card 105 to the storage component 116. The second signal includes a command signal transmitted from the RAID card 105 indicating that maintenance personnel need to remove the RAID card 105. The second signal may also include a power supply signal transmitted from the backup power supply 115 to the RAID card 105.
[0129] For example, the signal receiver 118 may be implemented in various forms, including but not limited to: for example, a combination of a signal indicator light and a signal button mounted on the second circuit board 1051, whereby maintenance personnel press the signal button and interpret the status information of the indicator light to determine the operation of the RAID card 105; or it may be a control command originating from the maintenance personnel's operating terminal. This application will selectively describe one of the aforementioned first implementation forms by way of example, but this should not be construed as limiting the scope of protection of this application.
[0130] For ease of explanation, this application provides an exemplary integration scheme for a RAID card 105, a second circuit board 1051, and various connectors (this scheme should not be considered a limitation of this application). Specifically, the RAID card 105 is mounted on the second circuit board 1051; the second connector 106 is a gold finger located on the edge of the circuit board to provide a convenient plug-in interface. A third connector 107 and a fourth connector 108 are arranged on the second circuit board 1051 at a position opposite to the second connector 106, along the direction from the first opening 1021a towards the first connector 104, both exposed through the first opening 1021a.
[0131] In some embodiments, see Figure 2 and combined Figure 3 The second circuit board 1051 is also connected to a stop plate 117, which is located on the side of the second circuit board 1051 opposite to the second connector 106. The stop plate 117 has a second opening 1022b for exposing the third connector 107 and the fourth connector 108. Furthermore, the stop plate 117 is detachably connected to the outer wall of the housing 102.
[0132] For example, the stop plate 117 is detachably connected to the outer wall of the housing 102 by screws.
[0133] Another example is that the stop plate 117 is magnetically connected to the outer wall of the housing 102.
[0134] It should be noted that when maintenance personnel need to disassemble or install the RAID card 105, they can first disassemble or install the second connecting cable 120 and the first connecting cable 119, and then disassemble or connect the stop plate 117 to the outer wall of the enclosure 102. Alternatively, they can first disassemble or connect the stop plate 117 to the outer wall of the enclosure 102, and then disassemble or install the second connecting cable 120 and the first connecting cable 119. This application does not limit the specific implementation of this method.
[0135] Based on the above, this application provides an exemplary description of a pluggable process for a RAID card 105.
[0136] When maintenance personnel need to insert or remove the RAID card 105, they first press the signal button to generate a removal command. This command is transmitted to the RAID card 105, which responds and uploads the command to the adapter board 113 via the fifth connector 109 and the sixth connector 110. The adapter board 113 forwards the command to the central processing unit 1031 via control signal lines. A command is sent to the backup power supply 115, which then starts up and supplies power to the adapter board via the adapter board and the fourth connector 108 connected to it. This is to maintain power to the adapter board's cache, ensuring that temporarily stored data can be safely transferred to the storage unit 116 via the third connector 107, thereby preventing data loss. After completing the above safety preparation process, maintenance personnel can remove the stop plate 117 and remove the entire module integrating the RAID card 105 (including the second connector 106) from the enclosure 102. Finally, the first connecting cable 119 and the second connecting cable 120 are disconnected to perform maintenance on the RAID card 105. Of course, maintenance personnel can disconnect the first connecting cable 119 and the second connecting cable 120 before removing the stop plate 117. This embodiment of the application does not limit this.
[0137] Based on the above operating procedures, maintenance personnel can replace, upgrade, or repair the RAID card 105 without shutting down the entire server node 100. In other words, this embodiment also meets the hot-swappable operation requirements of the RAID card 105.
[0138] In some embodiments, see Figure 2 and combined Figure 3 The signal receiver 118 includes the aforementioned signal button and signal indicator light. Both the signal button and the signal indicator light are connected to the main board 103 via the adapter board 113. When the maintenance personnel press the signal button, the signal receiver 118 enters its first operating state.
[0139] For example, the multiple operating states include a second operating state, in which the RAID card 105 is displayed as pluggable or removable, and the signal indicator light displays the corresponding visual state. The motherboard 103 is used to send a third control signal to the signal indicator light, at which time the signal indicator light responds to the third control signal, changing from a standby state to a flashing state, corresponding to "system preparation".
[0140] As another example, the multiple operation states include a third operation state, which is the completed plugging / unplugging state. At this time, the signal indicator light displays the corresponding visual state. The motherboard 103 is used to send a fourth control signal to the signal indicator light, and the signal indicator light turns to a constant on state, which corresponds to "safe to unplug", thereby guiding the maintenance personnel to perform the correct and safe subsequent operations.
[0141] During the aforementioned signal transmission process, this embodiment of the application can utilize the existing backup power supply 115 to synchronously transmit control commands from the motherboard 103 to the signal receiver 118. Specifically, the motherboard 103 transmits the second or third control signal to the adapter board 113 via the signal control line 114. The adapter board 113 then transmits the second or third control signal to the indicator light via the fifth connector 109 and the sixth connector 110, causing it to change to the corresponding state. Alternatively, the motherboard 103 can directly transmit the second or third control signal to the signal receiver 118 via the signal control line 114, the third circuit board 1131, the fifth connector 109, and the sixth connector 110, causing it to change to the corresponding state.
[0142] Therefore, there is no need to set up an additional independent control signal pin between the first connector 104 and the second connector 106. By using path multiplexing, the signaling interaction necessary to trigger the hot-plug operation is achieved.
[0143] It is understood that maintenance personnel can install the RAID card 105 in the reverse order of the above disassembly sequence. Of course, maintenance personnel can also customize the order. This application embodiment does not limit this.
[0144] In some embodiments, see Figure 2 and combined Figure 3 The storage component 116 and the RAID card 105 are located on one side of the motherboard 103.
[0145] Therefore, when maintenance operations are required (such as simultaneously inspecting or replacing the RAID card 105 or its associated storage components 116), maintenance personnel do not need to go around to different sides of the server 1 or adjust the device's position, making it convenient for maintenance personnel to perform operations on the same side.
[0146] In some embodiments, see Figure 2 and combined Figure 4 The motherboard 103 is located between the RAID card 105 and the storage unit 116.
[0147] Therefore, two independent operating surfaces can be formed on the two opposite walls of the enclosure 102. When maintenance is required on both the RAID card 105 and the storage unit 116 at the same time, maintenance personnel can operate synchronously on both sides of the enclosure 102 without obstructing each other or causing tool interference.
[0148] In some embodiments, see Figure 2 and combined Figure 5 One of the first connector 104 and the second connector 106 is an OCP gold finger interface, and the other of the first connector 104 and the second connector 106 is an OCP slot corresponding to the OCP gold finger interface.
[0149] By adopting the Open Computing (OCP) standard interface, the RAID 105 can be used as a standardized and interchangeable component, thus improving its versatility.
[0150] For example, the first connector 104 is an OCP standard slot (such as the 4C connector defined by OCP NIC 3.0) on the motherboard of server 1, and the second connector 106 is a gold finger integrated on the OCP NIC 3.0 accelerator card, carrying the RAID card 105.
[0151] Meanwhile, this application solves the problem of limited number of standard interface pins by multiplexing the control signal of the signal receiver 118 with the backup power supply 115 signal. Thus, without changing the OCP physical specification, the server node 100 of this application embodiment can realize hot-swapping of the RAID card 105.
[0152] The above description is merely a specific implementation of the embodiments of this application, but the protection scope of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the protection scope of this application. Therefore, the protection scope of this application should be determined by the protection scope of the claims.
Claims
1. A server node, characterized by include: A housing having a receiving cavity, the housing having a first opening communicating with the receiving cavity; The motherboard, storage components, and RAID card are housed within the receiving cavity. The motherboard has a first connector, and the RAID card has a second connector and a third connector. The RAID card enters and exits the receiving cavity through the first opening and is pluggably connected to the first connector of the motherboard through the second connector. The third connector is located on the side of the RAID card opposite to the second connector. When the RAID card is plugged into the motherboard, the third connector is exposed to the first opening. The third connector is used to connect to the storage component.
2. The server node of claim 1, wherein, Also includes: An adapter plate is disposed in the receiving cavity, the adapter plate is connected to the storage component, and the third connector is used for pluggable connection with the adapter plate.
3. The server node according to claim 2, characterized in that, The housing is provided with a second opening communicating with the receiving cavity, and the adapter plate is provided with a fourth connector. The fourth connector is exposed in the second opening, and the fourth connector and the third connector are pluggably connected via a first connecting cable.
4. The server node according to claim 3, characterized in that, Also includes: A backup power supply is housed within the receiving cavity. The backup power supply is connected to the adapter board and the motherboard respectively. The adapter board is provided with a fifth connector, which is exposed in the second opening. The RAID card is provided with a sixth connector, which is exposed in the first opening. The fifth connector and the sixth connector are detachably connected via a second connecting cable.
5. The server node according to claim 4, characterized in that, The first opening and the second opening are oriented in the same direction.
6. The server node according to claim 5, characterized in that, Also includes: A signal receiver, connected to the RAID card, is exposed through the first opening; When the signal receiver is in the first operating state, the motherboard is used to send a first control signal to the RAID card and / or a second control signal to the backup power supply; The RAID card is used to respond to the first control signal and transmit cache information to the storage component through the third connector and the adapter board; The backup power supply is used in response to the second control signal and to transmit power supply signals to the RAID card through the fifth connector and the sixth connector.
7. The server node according to any one of claims 1-6, characterized in that, The storage component and the RAID card are located on one side of the motherboard; or, The motherboard is located between the RAID card and the storage component.
8. The server node according to any one of claims 1-6, characterized in that, The storage component includes a hard disk backplane and at least one hard disk disposed on the hard disk backplane, and the third connector is used to connect to the hard disk backplane.
9. A server, characterized in that, include: At least one server node as described in any one of claims 1-8; Connected compute nodes and storage nodes, wherein the compute nodes and / or the storage nodes include the server node.
10. The server according to claim 9, characterized in that, The server includes a plurality of server nodes arranged along a first direction, and the enclosure includes a surrounding panel arranged around the first direction, with the first opening located in the surrounding panel.
