A server

Abstract

The present invention relates to the technical field of computer equipment design, and discloses a server, including: a chassis, a computing node, a management board, a bus expansion card, a power board, a power module, a fan board, a mechanical hard disk and a hard disk backplane; the bus expansion card and the hard disk backplane are arranged on the chassis front window of the chassis, the power module and the fan board are arranged on the chassis rear window of the chassis, the fan board is used to install a cooling component, the management board is arranged on the first side in the chassis, the computing node is connected to the management board, and the computing node is used to be pulled out from the reserved port on the second side of the chassis, and the first side and the second side are two opposite chassis sides between the chassis front window and the chassis rear window; the bus expansion card, the power board, the fan board and the hard disk backplane are all connected to the management board, the power board is used to connect the power module, the mechanical hard disk is connected to the hard disk backplane for communication, and the mechanical hard disk is used to be pulled out from the reserved port on the front window of the chassis. The present invention reduces the difficulty of operation through the side maintenance of the computing node.

Description

Technical Field

[0001] The present invention relates to the technical field of computer equipment design, and in particular to a server. Background Art

[0002] Currently, in order to ensure the uninterrupted and continuity of business, financial and cloud businesses are focusing on reliability design when selecting servers. Even if problems arise, they need to support system backup and hot maintenance of faulty equipment or boards to reduce service interruptions and stagnation, ensure the continuous operation of the business, and improve customer satisfaction. Among them, the thermal maintenance of server computing nodes is very important. At present, the thermal maintenance of computing nodes mainly includes two solutions: front window maintenance and rear window maintenance. The front window maintenance is to maintain the computing node on the front window of the server chassis, but the mechanical hard disk and PCIE (Peripheral Component Interconnect Express) bus expansion card are often placed on the front window of the server chassis. To achieve front maintenance in this solution, two computing nodes need to be plugged into a structural fixture built into the rear window. The PCIE signals of the two computing nodes need to be wired from the structural fixture to the bus expansion card on the front window, resulting in an excessively long PCIE signal routing, which cannot meet the PCIE 6.0 routing length requirements; rear window maintenance refers to the maintenance of computing nodes on the rear window of the server chassis. Often, the rear window of the chassis needs to be installed with a cold plate for chassis cooling. When the computing node is plugged in and out of the rear window, it will interfere with the inlet and outlet water pipes designed for the cold plate, resulting in inconvenient maintenance. Therefore, the current server chassis structure makes the thermal maintenance of server computing nodes very inconvenient, and improving the operability of thermal maintenance of server computing nodes is a problem that needs to be solved. Summary of the invention

[0003] In view of this, the present invention provides a server to solve the problem of low operability of hot maintenance of server computing nodes.

[0004] In a first aspect, the present invention provides a server, comprising: a chassis, a computing node, a management board, a bus expansion card, a power board, a power module, a fan board, a mechanical hard disk and a hard disk backplane; the bus expansion card and the hard disk backplane are arranged on the chassis front window of the chassis, the power module and the fan board are arranged on the chassis rear window of the chassis, the fan board is used to install a cooling component, the management board is arranged on a first side surface inside the chassis, the computing node is connected to the management board, and the computing node is used to be pulled out from a reserved opening on a second side surface of the chassis, the first side surface and the second side surface are two opposite chassis sides between the chassis front window and the chassis rear window; the bus expansion card, the power board, the fan board and the hard disk backplane are all connected to the management board, the power board is used to connect the power module, the mechanical hard disk is communicatively connected to the hard disk backplane, and the mechanical hard disk is used to be pulled out from the reserved opening on the chassis front window.

[0005] According to the above technical means, the present invention sets the management board for realizing the interconnection of boards on one side of the chassis, the front window of the chassis is the bus expansion card and the hard disk backplane, and the rear window of the chassis is the power module and the fan board, ensuring that the bus expansion card and the management board are close to each other, and the connection between the two can meet the wiring length requirements of PCIE 6.0. Afterwards, an outlet is reserved on the other side of the chassis, and the computing node is installed inside the chassis, connected to the management board, and unplugged through the reserved outlet on the other side of the chassis, thereby achieving the effect of hot maintenance of the computing node on the side of the chassis. An outlet is reserved on the front window of the chassis, and the mechanical hard disk is used to be unplugged from the reserved outlet on the front window of the chassis for hot maintenance. Through this scheme, the hot maintenance directions of the mechanical hard disk and the computer point are different, and they will not interfere with each other. Moreover, the computing node does not need to be unplugged from the rear window of the chassis for hot maintenance from the side, and will not interfere with the cooling component, which improves the operability of hot maintenance of the computing node and improves its hot maintenance efficiency.

[0006] In an optional embodiment, a computing node drawer and a hard disk drawer are provided in the chassis, the computing node drawer is used to be pulled out from the second side of the chassis, and the hard disk drawer is used to be pulled out from the front window of the chassis, the computing node is arranged in the computing node drawer, and when the computing node drawer is pushed into the chassis, the computing node is plugged into the management board through a first high-density connector; the fan board and the hard disk backplane are connected to the management board through lines, the mechanical hard disk is placed in the hard disk drawer, and the mechanical hard disk is connected to the hard disk backplane through lines; when the bus expansion card is pushed into the chassis from the front window of the chassis, it is plugged into the management board through a second high-density connector, and the power board is plugged into the management board through a third high-density connector.

[0007] According to the above technical means, through the design of the present invention, a hot maintenance solution that can achieve a blind plug effect is provided. When the computing node drawer is pushed into the second side, the computing node drawer is flush with the second side, the first high-density connector on the computing node and the first high-density connector on the management board are aligned, and the computing node is accurately plugged into the management board in the direction toward the first side through the first high-density connector. There is no need for the user to open the chassis or observe whether the computing node is plugged in place, as long as the computing node drawer is pushed in. Similarly, when the bus expansion card is pushed into the chassis from the front window of the chassis, it is accurately plugged into the management board through the second high-density connector. Through this solution, the interconnection of the four boards, namely the computing node, the bus expansion card, the management board, and the power board, is designed as a hard connection between the boards based on the high-density connector, realizing blind plug hot maintenance of the node and the PCIE device, without opening the chassis cover or plugging and unplugging the corresponding cables. The node or PCIE device is operated with power on during normal operation of the server, thereby avoiding the situation where the downtime affects the business interruption, and further improving the convenience of plugging and unplugging the computing node and the bus expansion card during hot maintenance.

[0008] In an optional embodiment, the hard disk backplane includes a plurality of hard disk connectors, the hard disk connectors are used to connect a plurality of extended solid-state hard disks, and the hard disk backplane is provided with a first ventilation hole penetrating the hard disk backplane.

[0009] According to the above technical means, a hard disk backplane is provided to expand the capacity of the solid state hard disk, and a first ventilation hole penetrating the hard disk backplane is arranged on the hard disk backplane to facilitate ventilation of the front window and the rear window of the chassis, thereby facilitating the heat dissipation of the computing nodes inside the chassis.

[0010] In an optional implementation, the computing nodes include two, wherein a first computing node and a second computing node are stacked one above the other, and the hard disk drawer is placed at a lower layer of the computing nodes.

[0011] According to the above technical means, the stacking method of computing nodes and hard disk drawers can adapt to the limitation of 19-inch chassis width of the new generation server platform.

[0012] In an optional implementation, the computing node includes a processor and multiple memories, and the first high-density connector provided on the computing node includes two bus interfaces, a power interface, and a low-speed control signal interface.

[0013] In an optional embodiment, the side of the management board facing the first side is the back side, and the side facing the second side is the front side, and the first high-density connector, the third high-density connector, the fan board interface and the hard disk backplane interface are arranged on the front side of the management board; the second high-density connector, the baseboard management controller and the power-on management module are arranged on the back side of the management board.

[0014] According to the above technical means, the management board layout structure provided by the embodiment of the present invention can make full use of the front and back space of the management board, which is convenient for the layout of various components. In addition, for the scenario where the front window of the chassis needs to be plugged in both the hard disk backplane and the bus expansion card, the second high-density connector and the hard disk backplane interface are respectively arranged on the front and back sides of the management board, which reduces the interference when the hard disk backplane and the bus expansion card are plugged in, is more conducive to the plug-in and unplug operation of the server during hot maintenance, and reduces the risk of damage to board components.

[0015] In an optional embodiment, the cooling component is a fan component, the fan board includes a fourth high-density connector, the fan board interface and a fan management module, the fan component is plugged into the fan board through the fourth high-density connector, and the fan management module is used to control the fan component.

[0016] In an optional embodiment, the third high-density connector is disposed on the back of the power board, the power connector of the power module is disposed on the front of the power board, and a second ventilation hole penetrating the power board is disposed on the power board.

[0017] According to the above technical means, the embodiment of the present invention respectively arranges the power connector and the third high-density connector of the power module on two different surfaces of the power board, one side is used for board-to-board plugging with the management board, and the other side is used to connect the lines of the power module. This can reduce the interference between the power board and the management board when plugging them together, and is more conducive to the disassembly operation of the power board. At the same time, a second ventilation hole that runs through the power board is arranged on the power board, which is more conducive to improving the overall heat dissipation effect of the chassis.

[0018] In an optional embodiment, the bus expansion card includes a first expansion interface, a second expansion interface, a third expansion interface and a second high-density connector, the first expansion interface is used to simultaneously receive bus signals of the first computing node and the second computing node, the second expansion interface is used to receive the bus signal of the first computing node, and the third expansion interface is used to receive the bus signal of the second computing node.

[0019] In an optional embodiment, the first extension interface has an X16 bandwidth specification, wherein the X8 bandwidth is used to receive the bus signal of the first computing node, and the X8 bandwidth is used to receive the bus signal of the second computing node, and the second extension interface and the third extension interface have an X8 bandwidth specification.

[0020] In an optional embodiment, the bus expansion card includes a first expansion interface, a second expansion interface and a third expansion interface, and the first expansion interface, the second expansion interface and the third expansion interface are used to select to receive the bus signal of the first computing node, or to receive the bus signal of the second computing node according to the control signal of the management board.

[0021] In an optional embodiment, the bus expansion card includes a first expansion interface, a second expansion interface and a third expansion interface, and the first expansion interface, the second expansion interface and the third expansion interface are all used to receive the bus signal of the first computing node, or are all used to receive the bus signal of the second computing node.

[0022] According to the above technical means, the present invention provides three layout schemes for bus expansion cards. The first scheme allocates independent high data bandwidth expansion interfaces to two computing nodes respectively, and also adopts a fusion expansion interface. By expanding one device, it can provide device capabilities for two computing nodes at the same time, saving the production cost of the number of expansion interfaces while improving the expansion capability of the server. The second scheme uses the baseboard management controller to control the three expansion interfaces through chip select signals to choose whether to connect to the first computing node or the second computing node. Thereby taking into account a variety of data communication needs, switching between fast signal needs and simple and vulgar signal needs. The third scheme is for business scenarios with low performance requirements. Customers will consider using only one computing node, thereby providing a variety of expansion interfaces with different bandwidths for one computing node, taking into account a variety of data communication needs, and switching between fast signal needs and simple and vulgar signal needs.

[0023] In an optional embodiment, the hard disk drawer and the node drawer are separated by a tray, the slide of the first computing node drawer is set on the tray, the slide of the second computing node drawer is set on the first computing node drawer, the first computing node is placed in the first computing node drawer, and the second computing node is placed in the second computing node drawer.

[0024] According to the above technical means, a slide is designed for the drawer of the second computing node drawer above the first computing node drawer. Therefore, when installing and maintaining the second computing node drawer, the first computing node drawer slides in along the slide above the first computing node drawer. This installation scheme can also significantly save the space occupied by the drawer slide and reduce the difficulty of designing the drawer slide on the side wall of the chassis.

[0025] In an optional implementation, the computing node is fixed in the node drawer by screws.

[0026] In an optional embodiment, a power-assisting handle is provided at a portion of the node drawer exposed outside the chassis. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0028] Figure 1 is a schematic diagram of the structure of a server according to an embodiment of the present invention;

[0029] Figure 2 is a perspective structural diagram of a server according to an embodiment of the present invention;

[0030] Figure 3 is another structural schematic diagram of a server according to an embodiment of the present invention;

[0031] Figure 4 is a front structural schematic diagram of a hard disk backplane according to an embodiment of the present invention;

[0032] Figure 5 is a schematic diagram of the back structure of a hard disk backplane according to an embodiment of the present invention;

[0033] Figure 6 is a schematic structural diagram of a fan plate according to an embodiment of the present invention;

[0034] Figure 7 is a front view of a front window of a chassis according to an embodiment of the present invention;

[0035] Figure 8 is a front view of a second side surface according to an embodiment of the present invention;

[0036] Fig. 9 is a perspective top view of a chassis according to an embodiment of the present invention;

[0037] Fig.10 is a front view of a rear window of a chassis according to an embodiment of the present invention;

[0038] Fig.11 is a schematic diagram of a computing node according to an embodiment of the present invention;

[0039] Fig.12 is a schematic diagram of the front structure of a management board according to an embodiment of the present invention;

[0040] Fig.13 is a schematic diagram of the back structure of a management board according to an embodiment of the present invention;

[0041] Fig.14 is a schematic structural diagram of the front side of a power board according to an embodiment of the present invention;

[0042] Fig.15 is a schematic structural diagram of the back side of a power board according to an embodiment of the present invention;

[0043] Fig.16 is a schematic diagram of the structure of a bus expansion card according to an embodiment of the present invention;

[0044] Reference numerals:

[0045] Chassis front window A, chassis rear window B, first side C, second side D, chassis 1, computing node 2, first computing node 2a, second computing node 2b, management board 3, bus expansion card 4, power board 5, power module 6, fan board 7, mechanical hard disk 8, hard disk backplane 9, computing node drawer 10, first computing node drawer 10a, second computing node drawer 10b, hard disk drawer 11, power handle 12, hard disk connector 13, first ventilation hole 14, PCIE connector 15, power supply terminal 16, fan assembly 1 7, first fan board interface 18, management chip module 19, fan board power supply interface 20, first high-density connector 21, second high-density connector 22, third high-density connector 23, processor 24, memory 25, bus interface 26, power interface 27, low-speed control signal interface 28, second fan board interface 29, hard disk backplane interface 30, baseboard management controller 31, power-on management module 32, power connector 33, second ventilation hole 34, first expansion interface 35, second expansion interface 36, third expansion interface 37. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0047] According to an embodiment of the present invention, a server is provided, such as Figure 1 and Figure 2 As shown, the server includes: a chassis 1, a computing node 2, a management board 3, a bus expansion card 4, a power board 5, a power module 6, a fan board 7, a mechanical hard disk 8 and a hard disk backplane 9.

[0048] Among them, the chassis 1 refers to the external shell of the server, which is used to accommodate various boards and components required by the server. The computing node 2 is a board device including a processor 24 and a memory 25, which is used for the server's calculation and logical judgment functions, and is the brain of the server. The bus expansion card 4 is a board for expanding external devices, such as a PCIE bus expansion card, which is used to expand the connection of graphics cards, network cards or other PCIE devices for the server. The power board 5 refers to a board for connecting the power module 6 and processing the power signal, thereby transmitting low-voltage direct current suitable for the server to the management board 3, and the management board 3 can power other connected boards and devices. The power module 6 refers to the chassis power supply. The fan board 7 is used to connect the cooling component, which includes but is not limited to the fan component and the water cooling component. The fan board 7 is provided with a management chip module 19, which is used to cooperate with the management board 3 to regulate the working efficiency of the cooling component. The mechanical hard disk 8 refers to the server hard disk for storing server data, and the hard disk backplane 9 is a circuit board for connecting the mechanical hard disk 8 to the management board 3. The hard disk backplane 9 can also support the installation of the NVME hard disk connector 13 of the solid-state hard disk. The management board 3 functions as a board interconnection, and is used to interconnect the computing node 2, the bus expansion card 4, the power board 5, the fan board 7, and the hard disk backplane 9, which is equivalent to the mainboard in the computer device.

[0049] See also Figure 1 , is a schematic diagram of a server chassis provided in an embodiment of the present invention. The chassis front window A usually refers to the side facing the user. The chassis front window often has an operation panel, which is provided with various indicator lights, expansion interfaces (such as USB interfaces), power buttons, restart buttons, etc. The chassis rear window B usually refers to the side without an operation panel, which cannot be operated. The chassis heat dissipation components and power supply components are usually arranged on the chassis rear window.

[0050] If the customer has designs for disaster recovery and data migration, it is often necessary to deploy more spare machines. In addition, there will be data loss and inaccuracy in the data transfer process during data migration, which will affect customer business and directly affect the user's sensory experience. In more serious cases, it will lead to data loss for the user. Based on this, it becomes urgent and important to perform hot maintenance on the computing node 2 and the bus expansion card 4 in the server. In order to achieve hot maintenance on the computing node 2 and the bus expansion card 4, the computing node and the bus expansion card can be operated and maintained with power on without shutting down the computer. Hot maintenance refers to the operation and maintenance of the server with power on when it is in normal startup state. There is no need to power off and then replace the corresponding components. In order to achieve hot maintenance, each board and component in the server chassis 1 is designed with an independent power supply, that is, it can be powered independently, and the removal of one component does not affect the continued operation of other components. In order to minimize the impact of node failures on customers. Assuming that the number of computing nodes is not unique, each computing node 2 and bus expansion card 4 are designed to be hot-swappable while powered. That is, when any computing node 2 in the server is abnormal or fails, the failed computing node 2 can be operated and maintained while the system has power, without the need to power off and shut down other normal computing nodes 2. This can maximize the guarantee of continuous support for the business and minimize the additional business migration actions before the customer shuts down all the computing nodes 2.

[0051] In order to improve the operability of hot maintenance, the present invention provides an improved server structure, such as Figure 2 As shown, in this structure, the bus expansion card 4 and the hard disk backplane 9 are arranged on the front window A of the chassis 1, so that the user can plug in the expansion network card, expansion graphics card, expansion solid state hard disk, etc. The power module 6 and the fan board 7 are arranged on the rear window B of the chassis, and the management board 3 is arranged on the first side C in the chassis. The first side C is a side of the chassis between the front window A and the rear window B of the chassis. The computing node 2 is connected to the management board 3. The connection method includes but is not limited to direct connection and wired connection. On the opposite side of the first side C, that is, the second side D, there is a reserved port. The computing node is used to be unplugged from the reserved port on the second side D of the chassis, as shown in FIG. Figure 3 As shown, it is a schematic diagram of pulling out the computing node from the second side of the chassis and disassembling it, so as to facilitate thermal maintenance of the computing node 2 on the side of the chassis. The bus expansion card 4, the power board 5, the fan board 7 and the hard disk backplane 9 are all connected to the management board 3, and the connection methods include but are not limited to direct plug-in connection and wired connection. Among them, a reserved opening is also provided on the front window A of the chassis, which is used to pull out the mechanical hard disk 8 from the reserved opening of the front window A of the chassis, so as to perform thermal maintenance on the mechanical hard disk 8.

[0052] According to the above technical means, the present invention ensures that the bus expansion card 4 and the management board 3 are close to each other, and the connection between the two can meet the wiring length requirements of PCIE 6.0. The computing node is unplugged through the reserved outlet on the side of the chassis, thereby achieving the effect of thermal maintenance of the computing node 2 on the side of the chassis. The mechanical hard disk 8 is used to be unplugged from the reserved outlet on the front window A of the chassis for thermal maintenance. The thermal maintenance directions of the mechanical hard disk 8 and the computing node 2 are different, and they will not interfere with each other. The computing node 2 does not need to be unplugged from the rear window B of the chassis for thermal maintenance from the side, and will not interfere with the cooling component, which improves the operability of thermal maintenance of the computing node 2 and improves its thermal maintenance efficiency.

[0053] In some optional implementations, a computing node drawer 10 and a hard disk drawer 11 are provided in the chassis. Figure 1 and Figure 3 As shown, the computing node drawer 10 is used to be pulled out from the second side D in the chassis, and the hard disk drawer 11 is used to be pulled out from the front window A of the chassis, wherein the mechanical hard disk 8 of the server is placed in the hard disk drawer 11, and the computing node 2 is placed in the computing node drawer 10. The portion of the computing node drawer 10 exposed outside the chassis is provided with a power-assisting handle 12 for easy gripping by the user. In addition, the power-assisting handle 12 is embedded in the side wall of the chassis and is flush with the side wall of the chassis. The purpose of this design is that, considering that a cabinet can hold multiple chassis 1, the chassis 1 needs to be pulled out of the cabinet in the direction of the front window A of the chassis. The embedded power-assisting handle 12 can prevent the chassis 1 from colliding with the cabinet, and prevent the raised power-assisting handle 12 from affecting the forward and backward sliding of the chassis 1.

[0054] Among them, the management board is provided with a first high-density connector 21, a second high-density connector 22 and a third high-density connector 23. When the computing node drawer 10 is pushed into the chassis, the computing node 2 is plugged into the management board 3 through the first high-density connector 21, and when the bus expansion card 4 is pushed into the chassis 1 from the front window A of the chassis, it is plugged into the management board 3 through the second high-density connector 22, and the power board 5 is plugged into the management board 3 through the third high-density connector 23. A high-density connector is a terminal for interconnection, which can be set on a circuit board. The high-density connector is divided into a male head and a female head. A male head is set on one board and a female head is set on the other board, so as to realize board-to-board plug-in and realize signal docking and intercommunication.

[0055] The fan board 7 and the hard disk backplane 9 are connected to the management board 3 through lines, the mechanical hard disk 8 is placed in the hard disk drawer 11, and the mechanical hard disk 8 is connected to the hard disk backplane 9 through lines. The hard disk backplane 9 and the mechanical hard disk 8 adopt the same line connection method as the related technology. Since the components for thermal maintenance of the hard disk backplane 9 are mainly solid-state hard disks, the solid-state hard disk can be directly removed from the hard disk backplane, so there is no need to improve the wired connection method of the hard disk backplane 9 management board 3.

[0056] Through the design of the present invention, a thermal maintenance solution that can achieve a blind plug-in effect is provided. When the computing node drawer 10 is pushed into the second side D, the computing node drawer 10 is flush with the second side D, the first high-density connector 21 on the computing node 2 is aligned with the first high-density connector 21 on the management board 3, and is accurately plugged into the management board through the first high-density connector 21 in the direction toward the first side C. The user does not need to open the chassis or observe whether the computing node 2 is plugged in place, and the computing node drawer 10 only needs to be pushed in. Similarly, when the bus expansion card 4 is pushed into the chassis 1 from the front window A of the chassis, it is accurately plugged into the management board 3 through the second high-density connector 22. Through this solution, the interconnection design of the four boards, namely the computing node 2, the bus expansion card 4, the management board 3, and the power board 5, is designed as a hard connection of board to board based on the high-density connector, thereby realizing blind plug-in hot maintenance of the computing node 2 and the PCIE device. There is no need to open the chassis cover or plug and unplug the corresponding cables. The computing node 2 or the PCIE device can be operated under power during normal operation of the server, thereby avoiding the situation where downtime affects business interruption, and further improving the convenience of plugging and unplugging the computing node 2 and the bus expansion card 4 during hot maintenance.

[0057] In addition, in some optional implementations, considering that the computing node 2 is often set in the computing node drawer 10, its board is closely attached to the bottom of the drawer, so it is difficult to disassemble the computing node 2, for example Figure 3 As shown, it is difficult to take the second computing node 2b out of the second computing node drawer 10b. In view of this problem, the present invention also sets a plurality of spring springs on the bottom surface of the computing node drawer 10, and presses the spring springs through a baffle plate that passes through the opposite side of the computing node drawer 10. The baffle plate is exposed by a preset length through the side of the computing node drawer 10. When the user needs to disassemble the computing node 2, the baffle plate is pulled out from the side of the computing node drawer 10, so that the spring spring releases the elastic force and pushes the computing node 2 to move upward, so that the board of the computing node 2 and the bottom of the computing node drawer 10 are no longer tightly fitted, which is convenient for the user to disassemble.

[0058] In some optional embodiments, Figure 4 As shown, the hard disk backplane 9 includes a plurality of hard disk connectors 13, which are used to connect a plurality of extended solid state hard disks, and the solid state hard disks can be hot-swapped. In addition, a first ventilation hole 14 that penetrates the hard disk backplane is provided on the hard disk backplane 9. In order to facilitate ventilation of the front window A and the rear window B of the chassis, it is beneficial to the heat dissipation of the computing node 2 inside the chassis.

[0059] Specifically, in some practical application embodiments, the hard disk connector 13 can be configured as 8 SFF 8639 connectors disposed on the hard disk backplane 9 and exposed on one side of the front window A of the chassis, so that 8 solid-state hard disks can be expanded. Figure 5As shown, the hard disk backplane 9 includes a high-speed signal PCIE connector 15 and a power supply terminal 16 on the back of the front window of the chassis, wherein the high-speed signal PCIE connector 15 and the power supply terminal 16 are used to connect the mechanical hard disk 8 of the server and the management board 3.

[0060] In some optional implementations, considering the limitation of the width of the 19-inch chassis of the new generation server platform, the maximum width of each computing node 2 cannot exceed 210mm. Each processor of the new platform is equipped with twenty-four memory sticks. The node width of 210mm cannot accommodate twenty-four memory sticks, so two computing nodes cannot be placed left and right. The technical solution provided by the present invention includes two computing nodes, and the chassis adopts a chassis with a height of 3U. The upper 2U is used to place the hard disk backplane 9, the bus expansion card 4 and the two computing nodes, wherein the first computing node 2a and the second computing node 2b are stacked up and down, and the lower 1U is used to place the hard disk drawer 11 that can accommodate 12 3.5-inch hard disks. The hard disk drawer 11 is placed in the lower layer of the computing node 2 to meet the size of the 19-inch chassis width of the new generation server platform.

[0061] In some optional embodiments, Figure 6 As shown, the cooling component is a fan component 17, and the fan board includes a fourth high-density connector, a first fan board interface 18, a management chip module 19 and a fan board power supply interface 20. The fan component 17 is plugged into the fan board 7 through the fourth high-density connector, and the fan component 17 can also be hot-swapped. The first fan board interface 18 is used to connect the communication line, and the fan board power supply interface 20 is used to connect the power supply line, so as to be connected to the management board 3 through the communication line and the power supply line. The management chip module 19 is a chip module, such as a CPLD (Complex Programmable Logic Device) chip, which is used to cooperate with the chip on the management board 3 to execute the control strategy to control the speed of the fan component 17, thereby adjusting the heat dissipation efficiency.

[0062] like Figure 7 As shown in the figure, it is the front view of the server chassis. Figure 8 As shown, it is a front view of the second side of the server chassis; Fig. 9 As shown in the figure, it is a perspective view of the chassis from a top-down angle. Fig.10 The figure shows the front view of the rear window of the chassis.

[0063] The upper 2U inside the chassis 1 are two 8*2.5cm hard disk backplanes 9 from the front window to the rear window; a bus expansion card 4 with 3 expansion slots, which can expand PCIE devices such as network cards, storage array cards, graphics cards, smart network cards, etc.; in the middle are two computing nodes 2; the management board 3 realizes the interconnection between boards and cards, and is used to manage and expand the power supply function; further down is the power board 5, and the rear window is the fan board 7 and two power modules 6, and the fan board 7 is installed with a fan assembly 17. For example, the fan board on the rear window B of the chassis can be installed with four 8056-type fans, which are placed in the center, with air flowing in from the front and out from the rear, providing air cooling for the entire chassis. For example, two power modules 6 (AC power supplies) can be stacked on the right side of the rear window of the chassis, and the two power modules 6 are connected to the power board 5 through the CRPS (Common Redundant Power Supply) power connector. The lower 1U of the rear window of the chassis is a baffle. The lower 1U inside the chassis is a hard disk drawer 11 for a 12*3.5cm mechanical hard disk.

[0064] In some optional embodiments, the hard disk drawer 11 and the node drawer 10 are separated by a tray, that is, there is an isolation tray between the lower 1U and the upper 2U, wherein the slide of the first computing node drawer 10a is set on the tray, and the slide of the second computing node drawer 10b is set on the first computing node drawer, the first computing node 2a is placed in the first computing node drawer 10a, and the second computing node 2b is placed in the second computing node drawer 10b, and each computing node 2 is fixed in the node drawer 10 by screws.

[0065] Specifically, during installation, the computing node 2 is inserted from the second side D of the chassis, and the first computing node drawer 10a is installed by sliding it in through the slide on the tray between the lower 1U and the upper 2U. A slide is designed for the second computing node drawer 10b above the first computing node drawer 10a. Therefore, when installing and maintaining the second computing node drawer 10b, the second computing node drawer 10b is slid in through the slide above the first computing node drawer 10a. When the first computing node drawer 10a and the second computing node drawer 10b are inserted, the first high-density connector 21 on each node is docked with the management board 3. After the first computing node drawer 10a and the second computing node drawer 10b are fully pushed in, they are plugged into place with the high-density connector of the management board. At this time, the computing node drawer 10 is flush with the side wall of the chassis, and the hand-tightened screwdriver side walls on both sides of the node drawer are locked to complete the installation of the node. This installation scheme can also significantly save the space occupied by the drawer slides, reducing the difficulty of designing drawer slides on the side walls of the chassis.

[0066] When it is necessary to perform operation and maintenance on a computing node, the user first shuts down the computing node that needs operation and maintenance, and the other computing node can continue to work. The two computing nodes work independently and are controlled separately. After the computing node to be operated and maintained is shut down, the server chassis can be pulled out of the cabinet toward the front window of the chassis. After pulling it until the computing node drawer is visible, on the right side of the chassis, the first computing node 2a or the second computing node 2b can be pulled out through the power handle 12 designed on the outside of the computing node drawer 10. After pulling out the computing node 2, the user can perform further operation and maintenance operations on the computing node 2.

[0067] In some optional implementations, the computing node includes a processor 24 and multiple memories 25 , and the first high-density connector 21 provided on the computing node 2 includes two bus interfaces 26 , a power interface 27 and a low-speed control signal interface 28 .

[0068] like Fig.11 As shown, it is a structural schematic diagram of a computing node provided by an embodiment of the present invention. The first computing node 2a and the second computing node 2b have the same structure. Taking a computing node as an example, the computing node includes a processor 24 and multiple memories 25, wherein the processor 24 includes but is not limited to a central processing unit, a network processor or a combination thereof. Among them, the processor may further include a hardware chip. The above-mentioned hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The above-mentioned programmable logic device may be a complex programmable logic device, a field programmable gate array, a general array logic or any combination thereof. The memory 25 is a computer storage device for temporarily storing programs and data. It is a very important component in the server and is responsible for storing running programs and data so that the processor can quickly access and process this information. Taking a 19-inch chassis as an example, up to twenty-four memory sticks are supported, thereby improving the computing power of the computer point.

[0069] Among them, the processor 24 is in the middle of the computing node 2, and the two ends of the processor 24 are the memory 25. The first high-density connector 21 is set on a side of the computing node 2. The processor 24, the memory 25 and the first high-density connector 21 are communicatively connected on the board of the computing node 2 through a printed circuit.

[0070] In an embodiment of the present invention, the first high-density connector 21 includes two bus interfaces 26, a power interface 27 and a low-speed control signal interface 28, wherein the bus interface 26 is used to communicate with the bus expansion card 4 or the hard disk backplane 9 through the management board 3, and provide bus signals for the hard disk backplane 9 and the bus expansion card 4, so as to expand the external extended graphics card, network card, solid state hard disk and other devices to the computing node 2. In a specific embodiment, the two bus interfaces on the computing node 2 are PCIE X32 interfaces, indicating a PCIE bus interface with a bandwidth specification of X32, thereby providing high-speed signal transmission capability based on a large bandwidth, so as to facilitate the transmission of large-volume signal data. The power interface 27 is used to connect to the power interface on the management board 3, so as to take power from the management board and power the processor 24, memory 25 and voltage regulator and other components on the computing node 2. The low-speed control signal interface 28 is used to interconnect the control signals of the chips on the computing node 2 and the management board 3, so that the two can interact with each other in commands to ensure the correct execution of computing tasks and logical judgment tasks.

[0071] In an alternative embodiment, Fig.12 , which is a schematic diagram of the front structure of the management board 3 provided in the embodiment of the present invention. In the embodiment of the present invention, the side of the management board 3 facing the first side is defined as the back side, and the side facing the second side is defined as the front side. Based on this, the front side of the management board 3 in the embodiment of the present invention is provided with a first high-density connector 21, a third high-density connector 23, a second fan board interface 29 and a hard disk backplane interface 30; Fig.13 As shown, a second high-density connector 22, a baseboard management controller 31 and a power-on management module 32 are disposed on the back of the management board.

[0072] Among them, the first high-density connector 21 is divided into an upper and lower part, which are respectively used to dock the first computing node 2a and the second computing node 2b. When the node drawers 10 of the first computing node 2a and the second computing node 2b are pushed into the chassis, the first high-density connector 21 on the computing node 2 and the first high-density connector 21 on the management board 3 can be accurately docked and inserted. The third high-density connector 23 is arranged on the front of the management board 3 and close to the rear window B of the chassis, which is convenient for the board-to-board plug-in of the power board 5. The second fan board interface 29 and the hard disk backplane interface 30 are two wired connection interfaces respectively. The second fan board interface 29 is used to connect to the first fan board interface 18 on the fan board 7 through a cable, and the hard disk backplane interface 30 is used to connect to the PCIE connector 15 on the hard disk backplane through a cable. The hard disk backplane interface 30 is located on the front of the management board 3 and close to the front window A of the chassis. The second high-density connector 22 is arranged on the back of the management board 3 and close to the front window A of the chassis, and is used to plug with the bus expansion card 4. Similarly, the management board 3 is provided with a baseboard management controller 31 and a power-on management module 32. The baseboard management controller 31 (Baseboard Management Controller, BMC) plays an important role in server management. It is a small operating system or a dedicated management subsystem independent of the server computing node, which is used to manage and monitor the hardware status, operating system, health status and power consumption of the server. The BMC monitors various physical variables of the server through sensors, such as temperature, humidity, power supply voltage, fan speed, etc., and notifies the administrator when these variables exceed the set range. The administrator can take measures remotely, such as restarting the server, to ensure the stable operation of the server. In an embodiment of the present invention, the BMC can use an AST2700 chip to manage two computing nodes. The power-on management module 32 realizes the control of the power-on timing of the entire system. In an embodiment of the present invention, the power-on management module 32 can be implemented using a CPLD (Complex Programmable Logic Device) chip.

[0073] In the embodiment of the present invention, the hard disk backplane interface 30 on the management board can be implemented by two PCIE bus interfaces with a bandwidth of X32, thereby improving the transmission speed of the extended solid state hard disk data. For the second high-density connector 22 connecting the bus expansion card 4 and the management board 3, a PCIE bus interface with a bandwidth of X48 can be used to fully support a variety of bus expansion devices and improve the data transmission speed of the expansion devices.

[0074] In addition, the first high-density connector 21, the third high-density connector 23, the second fan board interface 29, the hard disk backplane interface 30, the second high-density connector 22 and the power-on management module 32 all establish communication connections with the baseboard management controller 31 by means of printed circuits. The management board layout structure provided by the embodiment of the present invention can make full use of the front and back spaces of the management board 3, which is convenient for the layout of various components. And for the scenario where the front window A of the chassis needs to be plugged in both the hard disk backplane 9 and the bus expansion card 4, the second high-density connector 22 and the hard disk backplane interface 30 are respectively arranged on both sides of the front and back of the management board 3, which reduces the interference when the hard disk backplane 9 and the bus expansion card 4 are plugged in, is more conducive to the plug-in and unplug operation of the server during hot maintenance, and reduces the risk of damage to board components.

[0075] In some optional embodiments, Fig.14 Shown and Fig.15 As shown, the third high-density connector 23 is arranged on the back of the power board 5, the front of the power board 5 is provided with a power connector 33 of the power module, and the power board 5 is provided with a second ventilation hole 34 that passes through the power board, wherein the back of the power board 5 faces the first side, and the front of the power board 5 faces the second side.

[0076] Specifically, the embodiment of the present invention arranges the power connector 33 and the third high-density connector 23 of the power module 6 on two different surfaces of the power board 5, one surface is used for board-to-board plugging with the management board 3, and the other surface is used to connect the line of the power module 6. This can reduce the interference between the power board 5 and the management board 3 when plugging and the power module 6, and is more conducive to the disassembly operation of the power board 5. At the same time, a second ventilation hole 34 that penetrates the power board is arranged on the power board 5, which is more conducive to improving the overall heat dissipation effect of the chassis.

[0077] In some optional implementations, the present invention also provides three bus expansion card configuration solutions. Fig.16 As shown, it is a schematic diagram of the structure of the bus expansion card 4, wherein the configuration of scheme 1 is as follows:

[0078] The bus expansion card 4 includes a first expansion interface 35, a second expansion interface 36, a third expansion interface 37 and a second high-density connector 22. The first expansion interface 35 is used to simultaneously receive bus signals of the first computing node 2a and the second computing node 2b, the second expansion interface 36 is used to receive bus signals of the first computing node 2a, and the third expansion interface 37 is used to receive bus signals of the second computing node 2b.

[0079] The first expansion interface 35 has an X16 bandwidth specification, wherein the X8 bandwidth is used to receive the bus signal of the first computing node 2a, and the X8 bandwidth is used to receive the bus signal of the second computing node 2b. The second expansion interface 36 and the third expansion interface 37 have an X8 bandwidth specification.

[0080] Specifically, the bus expansion card 4 provided in the embodiment of the present invention includes 3 PCIE slots, which can be used to expand PCIE devices, such as graphics cards, smart network cards, multihost network cards, disk array cards, etc. Among them, the first expansion interface 35 is used to simultaneously receive the bus signals of the first computing node 2a and the second computing node 2b. For example, the bandwidth of the first expansion interface 35 is PCIE X8 of the first computing node 2a + PCIE X8 of the second computing node 2b, that is, the X8 bandwidth is used to receive the X8 PCIE bus signal on the processor of the first computing node 2a, and the other X8 bandwidth is used to receive the X8 PCIE bus signal on the processor of the second computing node 2b. For example: the multihost network card can be expanded through the first expansion interface 35, and the X16 signal of the multihost network card can be evenly connected to 2 nodes, that is, the processors of the 2 nodes each separate an X8 PCIE line to connect to the X16 multihost network card. The embodiment of the present invention can simultaneously process the task requirements of dual-node processors through a single network card.

[0081] The second expansion interface 36 is used to receive a bus signal of a processor on the first computing node 2a. In a specific embodiment, it can receive a PCIE bus signal with an X8 bandwidth, and is used to expand an X16 PCIE device for the first computing node 2a, such as a GPU card, an intelligent network card, an ordinary high-specification network card, or other PCIE devices;

[0082] The third extension interface 37 is used to receive the bus signal of the second computing node 2b, for example, to receive the PCIE bus signal of the X8 bandwidth of the processor on the second computing node 2b, and to extend the X16 PCIE device of the second computing node 2b, such as a GPU card, an intelligent network card, an ordinary high-specification network card or other PCIE devices;

[0083] The second high-density connector 22 mainly includes a low-speed control signal interface and a power supply interface, wherein the low-speed control signal interface is used to implement the baseboard management controller on the management board. 2 The power supply interface is used to interconnect with the management board, draw power from the management board, and supply power to the PCIE devices expanded on the bus expansion card.

[0084] Through the configuration scheme of the bus expansion card provided by the embodiment of the present invention, independent high data bandwidth expansion interfaces are allocated to the two computing nodes respectively, and a first expansion interface with integrated functions is also adopted. By expanding one device, device capabilities can be provided to two computing nodes at the same time, which saves the production cost of the number of expansion interfaces and improves the expansion capability of the server.

[0085] The second solution provided by the embodiment of the present invention is configured as follows:

[0086] The bus expansion card includes a first expansion interface 35, a second expansion interface 36 and a third expansion interface 37. The first expansion interface 35, the second expansion interface 36 and the third expansion interface 37 are used to select to receive the bus signal of the first computing node 2a, or to receive the bus signal of the second computing node 2b according to the control signal of the management board 3.

[0087] Specifically, the bus expansion card 4 provided in the embodiment of the present invention includes 3 PCIE slots, which can be used to expand PCIE devices, such as graphics cards, smart network cards, multihost network cards, disk array cards, etc. Among them, the first expansion interface 35 is used to receive the bus signal of the first computing node 2a or the second computing node 2b, the second expansion interface 36 is used to receive the bus signal of the first computing node 2a or the second computing node 2b, and the third expansion interface 37 is used to receive the bus signal of the first computing node 2a or the second computing node 2b. In other words, the first expansion interface 35, the second expansion interface 36 and the third expansion interface 37 can not only expand the PCIE device for the first computing node 2a, but also expand the PCIE device for the second computing node 2b, but one expansion interface can only expand the PCIE device for one of the computing nodes 2 at the same time. Which computing node 2 is to be expanded specifically needs to be determined according to the program of the baseboard management controller 31 on the management board 3. The baseboard management controller 31 controls the three expansion interfaces through the chip selection signal to select whether to connect to the first computing node 2a or the second computing node 2b.

[0088] In a specific embodiment, the three expansion interfaces may include one X16 bandwidth PCIE bus interface and two X8 bandwidth PCIE bus interfaces, for example, the first expansion interface 35 is a PCIE X16 interface, and the second expansion interface 36 and the third expansion interface 37 are PCIE X8 interfaces, thereby taking into account a variety of data communication requirements and replacing between fast signal requirements and simple and low-level signal requirements. The present invention is only an example and is not limited thereto.

[0089] The configuration of solution 3 provided in the embodiment of the present invention is as follows:

[0090] The bus expansion card 4 includes a first expansion interface 35, a second expansion interface 36 and a third expansion interface 37, which are all used to receive bus signals of the first computing node 2a, or are all used to receive bus signals of the second computing node 2b.

[0091] Specifically, in an embodiment of the present invention, for a business scenario with low performance requirements, the customer may consider using only one computing node, for example, the customer only expands the first computing node 2a, or only expands the second computing node 2b, thereby only utilizing the performance of the first computing node 2a or only utilizing the performance of the second computing node 2b. Assuming that the user only uses the first computing node 2a, based on this situation, the PCIE signals of the three expansion interfaces on the bus expansion card 4 can all be connected to the first computing node 2a. For example, in a specific embodiment, the three expansion interfaces may include an X16 bandwidth PCIE bus interface and two X8 bandwidth PCIE bus interfaces, for example, the first expansion interface 35 is a PCIE X16 interface, the second expansion interface 36 and the third expansion interface 37 are PCIE X8 interfaces, and the three expansion interfaces are all connected to the PCIE signal of the first computing node 2a, thereby taking into account a variety of data communication requirements, and replacing between fast signal requirements and simple and vulgar signal requirements. The present invention is only taken as an example and is not limited thereto.

[0092] Through the technical solution provided by the present invention, compared with the front and back maintenance methods of the related technology, the configuration of the mechanical hard disk 8 is subject to certain restrictions during front maintenance, and during rear maintenance, it will interfere with the inlet and outlet water pipes of the liquid cooling pipeline in the computer room. In addition, when the node is maintained from the rear, the node layout is biased to the rear, and there is a risk that the PCIE device on the front window A of the chassis cannot support the PCIE 6.0 rate. The embodiment of the present invention performs pull-out maintenance on the computing node on the side wall of the server chassis. The side wall maintenance proposed in the embodiment of the present invention requires the chassis to be pulled out of the cabinet in the direction of the front window, and the chassis is maintained from the side wall after the computing node is leaked out, while the maintenance of the mechanical hard disk 8 is the maintenance of the front window A of the chassis. This method not only solves the problem of maintenance interference, but also avoids the risk that the PCIE device on the front window A of the chassis cannot support the PCIE 6.0 rate.

[0093] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A server, characterized in that: include: Chassis, computing nodes, management board, bus expansion card, power board, power module, fan board, mechanical hard disk and hard disk backplane; The bus expansion card and the hard disk backplane are arranged on the front window of the chassis, the power module and the fan board are arranged on the rear window of the chassis, the fan board is used to install a cooling component, the management board is arranged on the first side of the chassis, the computing node is connected to the management board, and the computing node is used to be powered off from the reserved port on the second side of the chassis, and the first side and the second side are two opposite sides of the chassis between the front window of the chassis and the rear window of the chassis; The bus expansion card, the power board, the fan board and the hard disk backplane are all connected to the management board, the power board is used to connect the power module, the mechanical hard disk is communicatively connected to the hard disk backplane, and the mechanical hard disk is used to be pulled out from the reserved opening on the front window of the chassis; A computing node drawer is provided in the chassis, and the computing node drawer is used to be drawn out from the second side of the chassis; The computing node is arranged in the computing node drawer. When the computing node drawer is pushed into the chassis, the computing node is plugged into the management board through the first high-density connector, and the second side surface is on the right side of the chassis; A plurality of spring springs are arranged on the bottom surface of the computing node drawer, and the spring springs are pressed by baffles penetrating opposite sides of the computing node drawer, and the baffles are exposed by a preset length through the sides of the computing node drawer; The computing nodes include two, and the bus expansion card includes a first expansion interface, which is used to simultaneously receive bus signals of the first computing node and the second computing node. The first expansion interface has an X16 bandwidth specification, wherein the X8 bandwidth is used to receive the bus signal of the first computing node, and the X8 bandwidth is also used to receive the bus signal of the second computing node, and the network card is extended through the first expansion interface.

2. The server according to claim 1, characterized in that: A hard disk drawer is provided in the chassis, and the hard disk drawer is used to be drawn out from the front window of the chassis; The fan board and the hard disk backplane are connected to the management board through lines, the mechanical hard disk is placed in the hard disk drawer, and the mechanical hard disk is connected to the hard disk backplane through lines; When the bus expansion card is pushed into the chassis from the front window of the chassis, it is plugged into the management board through the second high-density connector, and the power board is plugged into the management board through the third high-density connector.

3. The server according to claim 1, characterized in that: The hard disk backplane includes a plurality of hard disk connectors, and the hard disk connectors are used to connect a plurality of extended solid state hard disks. The hard disk backplane is provided with a first ventilation hole that penetrates the hard disk backplane.

4. The server according to claim 2, characterized in that: The first computing node and the second computing node are stacked up and down, and the hard disk drawer is placed at the lower layer of the computing nodes.

5. The server according to claim 2, characterized in that: The computing node includes a processor and multiple memories. The first high-density connector provided on the computing node includes two bus interfaces, a power interface and a low-speed control signal interface.

6. The server according to claim 2, characterized in that: The side of the management board facing the first side is the back side, and the side facing the second side is the front side. The first high-density connector, the third high-density connector, the fan board interface and the hard disk backplane interface are arranged on the front side of the management board; the second high-density connector, the baseboard management controller and the power-on management module are arranged on the back side of the management board.

7. The server according to claim 6, characterized in that: The cooling component is a fan component. The fan board includes a fourth high-density connector, the fan board interface and a fan management module. The fan component is plugged into the fan board through the fourth high-density connector. The fan management module is used to control the fan component.

8. The server according to claim 2, characterized in that: The third high-density connector is arranged on the back side of the power board, the power connector of the power module is arranged on the front side of the power board, and a second ventilation hole penetrating the power board is arranged on the power board; wherein the back side of the power board faces the first side, and the front side of the power board faces the second side.

9. The server according to claim 4, characterized in that: The bus expansion card includes a second expansion interface, a third expansion interface and the second high-density connector, the second expansion interface is used to receive the bus signal of the first computing node, and the third expansion interface is used to receive the bus signal of the second computing node.

10. The server according to claim 9, characterized in that The second extension interface and the third extension interface have an X8 bandwidth specification.

11. The server according to claim 4, characterized in that: The bus expansion card includes a first expansion interface, a second expansion interface and a third expansion interface, and the first expansion interface, the second expansion interface and the third expansion interface are used to select to receive the bus signal of the first computing node, or to receive the bus signal of the second computing node according to the control signal of the management board.

12. The server according to claim 4, characterized in that: The bus expansion card includes a first expansion interface, a second expansion interface and a third expansion interface, and the first expansion interface, the second expansion interface and the third expansion interface are all used to receive the bus signal of the first computing node, or are all used to receive the bus signal of the second computing node.

13. The server according to claim 4, characterized in that: The hard disk drawer and the node drawer are separated by a tray, the slide of the first computing node drawer is set on the tray, the slide of the second computing node drawer is set on the first computing node drawer, the first computing node is placed in the first computing node drawer, and the second computing node is placed in the second computing node drawer.

14. The server according to claim 13, characterized in that: The computing node is fixed in the node drawer by screws.

15. The server according to claim 13, characterized in that: The portion of the node drawer exposed outside the chassis is provided with a power-assisting handle, and the power-assisting handle is embedded in the side wall of the chassis and is flush with the side wall of the chassis.

Images