A server and multi-motherboard device thereof

Abstract

The present application discloses a server and a multi-motherboard device thereof, which are applied to the field of networking technology, and include: a first CPU, a first and a second connector arranged in a first motherboard; a second CPU, a third and a fourth connector arranged in a second motherboard; in a first mode, the first CPU is connected to a shared network card through a first connector, a transit connector and a fourth connector, and the second CPU is connected to the shared network card, so that the first and second CPUs use the shared network card; in a second mode, the first interface of the first CPU is connected to the first network card through the first and second connectors in sequence, and the second interface of the first CPU is connected to the first network card, so that the first CPU uses the first network card; in a second mode, the second CPU uses the second network card. Applying the solution of the present application can support CPUs in different motherboards to share the same network card, and also support CPUs in different motherboards to use their own network cards, and reduce costs.

Description

Technical Field

[0001] The present invention relates to the technical field of networking, and in particular to a server and multi-motherboard equipment thereof. Background Art

[0002] At present, the cost of network cards and optical modules is increasing with the increase of network speed, resulting in high networking costs. The shared network card (Multi-Host) solution supports sharing the same network card between multiple CPUs (Central Processing Units), which can reduce the number of network cards required for computing nodes in the entire server cluster, thereby reducing networking costs. With the same number of computing nodes, reducing networking costs can make the computing cluster solution more competitive.

[0003] See also Figure 1 , which is a diagram of a traditional commonly used networking method. Usually in the same server, two CPUs share the same shared network card. Figure 1 The CPU 0, CPU 1 and shared network card are set in the same motherboard, and both CPU 0 and CPU 1 can use the shared network card to realize external data transmission. Figure 1 In this design, the management interface and other I / O (Input / Output) ports are set on the same side of the server as the shared network card to facilitate centralized layout of the interfaces and reduce space occupancy.

[0004] but, Figure 1 Such a solution only supports CPUs on the same motherboard to share the same network card, and cannot allow CPUs in different motherboards to share the same network card. Although there are currently some solutions that can enable CPUs in different motherboards to share the same network card, they are all relatively complicated.

[0005] In addition, some users have the need to switch the way the network card is used. For example, the user needs to let the CPUs in different motherboards share the same network card at the current time, and then wants the CPUs in different motherboards to use their own network cards at a later time. For users with such needs, the current solution usually provides two sets of equipment for them, and the user chooses the corresponding equipment to work according to their needs, but this will be very costly.

[0006] In summary, how to effectively implement networking, both to support CPUs in different motherboards to share the same network card and to support CPUs in different motherboards to use their own network cards, and to reduce costs, is a technical problem that technical personnel in this field urgently need to solve. Summary of the invention

[0007] The purpose of the present invention is to provide a server and a multi-motherboard device thereof to effectively realize networking, which can support CPUs in different motherboards to share the same network card, and can support CPUs in different motherboards to use their own network cards, and reduce costs.

[0008] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0009] In a first aspect, the present invention provides a multi-motherboard device, comprising: a first central processing unit, a first connector, and a second connector arranged in a first motherboard; a second central processing unit, a third connector, and a fourth connector arranged in a second motherboard;

[0010] In the first mode, the first interface of the first central processor is connected to the first interface of the shared network card through the first connector, the transit connector and the fourth connector in sequence, and the second interface of the second central processor is connected to the second interface of the shared network card, so that in the first mode, the shared network card is shared by the first central processor and the second central processor;

[0011] In the second mode, the first interface of the first central processor is connected to the first interface of the first network card through the first connector and the second connector in sequence, and the second interface of the first central processor is connected to the second interface of the first network card, so that in the second mode, the first central processor uses the first network card;

[0012] In the second mode, the first interface of the second central processor is connected to the first interface of the second network card through the third connector and the fourth connector in sequence, and the second interface of the second central processor is connected to the second interface of the second network card, so that in the second mode, the second central processor uses the second network card.

[0013] On the other hand, the first interface and the second interface of the first central processor, the first interface and the second interface of the second central processor, the first interface and the second interface of the shared network card, the first interface and the second interface of the first network card, and the first interface and the second interface of the second network card are all 8-channel peripheral component rapid interconnection interfaces;

[0014] The transit connector, the first connector, the second connector, the third connector and the fourth connector are all connectors using an 8-channel peripheral component rapid interconnect interface.

[0015] On the other hand, the first main board is arranged above the second main board, and in the first mode, the fourth connector is connected to the female connector of the transit connector via a detachable male connector.

[0016] On the other hand, the connection line between the first interface of the first central processor and the first connector is a connection line fixedly set in the first mainboard; the connection line between the second interface of the first central processor and the external interface on the first mainboard is a connection line fixedly set in the first mainboard; the connection line between the second connector and the external interface on the first mainboard is a connection line fixedly set in the first mainboard;

[0017] The connection line between the first interface of the second central processing unit and the third connector is a connection line fixedly set in the second mainboard; the connection line between the second interface of the second central processing unit and the external interface on the second mainboard is a connection line fixedly set in the second mainboard; the connection line between the fourth connector and the external interface on the second mainboard is a connection line fixedly set in the second mainboard.

[0018] On the other hand, in the first mode, the connection line between the first connector and the relay connector, and the connection line between the relay connector and the fourth connector are both detachable connection lines.

[0019] On the other hand, in the second mode, the connection line between the first connector and the second connector, and the connection line between the third connector and the fourth connector are both detachable connection lines.

[0020] On the other hand, the transfer connector, the first connector, the second connector, the third connector and the fourth connector all have the same structure and are all multi-channel input and output connectors.

[0021] On the other hand, the first central processor is further configured to: in the first mode, when a connection interruption or bandwidth degradation with the shared network card is detected, execute a preset fault handling operation; in the second mode, when a connection interruption or bandwidth degradation with the first network card is detected, execute the fault handling operation;

[0022] The second central processor is also used to: in the first mode, when it detects that the connection with the shared network card is interrupted or the bandwidth is degraded, execute the fault handling operation; in the second mode, when it detects that the connection with the second network card is interrupted or the bandwidth is degraded, execute the fault handling operation.

[0023] On the other hand, the fault handling operation performed includes:

[0024] The fault time is recorded, and when the restart conditions are met, the mainboard is automatically restarted.

[0025] On the other hand, the transit connector is a right-angle connector, and the first connector, the second connector, the third connector and the fourth connector are all vertical connectors.

[0026] On the other hand, in the first mode, the transit connector is fixed to the outside of the first mainboard by a preset fixing component.

[0027] On the other hand, each external interface in the first mainboard is arranged in the direction of the cabinet rear window, and each interface in the second mainboard is arranged in the direction of the cabinet rear window;

[0028] In the first mode, the shared network card and the transit connector are both arranged in the direction of the rear window of the cabinet; in the second mode, the first network card and the second network card are both arranged in the direction of the rear window of the cabinet.

[0029] On the other hand, each external button and indicator light in the first main board are arranged in the direction of the rear window of the cabinet, and each external button and indicator light in the second main board are arranged in the direction of the rear window of the cabinet.

[0030] In a second aspect, the present invention provides a server, comprising the multi-motherboard device as described above.

[0031] On the other hand, the server includes N of the multi-motherboard devices, where N is a positive integer not less than 2, and the structures of the multi-motherboard devices are consistent.

[0032] By applying the technical solution provided by the embodiment of the present invention, the multi-motherboard device can support both the first mode and the second mode through the design of the corresponding connector. Specifically, the first motherboard is provided with a first central processor, a first connector and a second connector, and the second motherboard is provided with a second central processor, a third connector and a fourth connector. In the first mode, the first interface of the first central processor needs to be connected to the first interface of the shared network card through the first connector, the transit connector and the fourth connector in sequence, and the second interface of the second central processor is connected to the second interface of the shared network card. It can be seen that in the first mode, the shared network card is shared by the first central processor and the second central processor, that is, the need for CPUs in different motherboards to share the same network card is met. In the second mode, the first interface of the first central processor can be connected to the first interface of the first network card through the first connector and the second connector in sequence, and the second interface of the first central processor is connected to the second interface of the first network card. It can be seen that in the second mode, the first network card is used by the first central processor and is not shared with the second central processor. Correspondingly, in the second mode, the first interface of the second central processor is connected to the first interface of the second network card through the third connector and the fourth connector in sequence, and the second interface of the second central processor is connected to the second interface of the second network card. It can be seen that in the second mode, the second network card is used by the second central processor and is not shared with the first central processor. Therefore, in the second mode, the purpose of CPUs in different motherboards using their own network cards is met. It can also be seen that the present application does not need to provide two different servers for users, but can achieve the purpose through mode switching. When switching modes, it is not necessary to adjust the number of components in the first motherboard and the second motherboard, but only to change the connection relationship of the relevant connectors, making the present application low-cost.

[0033] In summary, the present application scheme can realize networking, which can support CPUs in different motherboards to share the same network card, and can also support CPUs in different motherboards to use their own network cards, and reduces costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0035] Figure 1 This is a schematic diagram of a traditional commonly used networking method;

[0036] Figure 2A schematic diagram of the structure of a multi-motherboard device provided by a specific implementation mode of the present invention, and the connection mode is the first mode;

[0037] Figure 3 A schematic diagram of the structure of a multi-motherboard device provided by a specific implementation mode of the present invention, and the connection mode is the second mode;

[0038] Figure 4 A schematic diagram of the structure of a first mainboard in a multi-mainboard device provided in a specific embodiment of the present invention;

[0039] Figure 5 This is a schematic diagram of another traditional networking method;

[0040] Figure 6 This is a front view of a multi-mainboard device viewed from the rear window of a cabinet in a specific implementation manner of the present invention. DETAILED DESCRIPTION

[0041] The core of the present invention is to provide a server and a multi-motherboard device thereof, which can realize networking, can support CPUs in different motherboards to share the same network card, can support CPUs in different motherboards to use their own network cards, and reduce costs.

[0042] In order to enable those skilled in the art to better understand the scheme of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0043] Please refer to Figure 2 and Figure 3 , Figure 2 A schematic diagram of a multi-motherboard device provided by a specific embodiment of the present invention, and Figure 2 The connection method is the connection method in the first mode. Figure 3 This is the connection method of the multi-motherboard device in the second mode provided by this specific implementation method.

[0044] The multi-motherboard device includes: a first central processor 21, a first connector 11 and a second connector 12 arranged in a first motherboard; and a second central processor 22, a third connector 13 and a fourth connector 14 arranged in a second motherboard.

[0045] In the first mode, the first interface of the first central processor 21 is connected to the first interface of the shared network card through the first connector 11, the transit connector and the fourth connector 14 in sequence, and the second interface of the second central processor 22 is connected to the second interface of the shared network card, so that in the first mode, the shared network card is shared by the first central processor 21 and the second central processor 22.

[0046] In the second mode, the first interface of the first central processor 21 is connected to the first interface of the first network card through the first connector 11 and the second connector 12 in sequence, and the second interface of the first central processor 21 is connected to the second interface of the first network card, so that in the second mode, the first central processor 21 uses the first network card.

[0047] In the second mode, the first interface of the second central processor 22 is connected to the first interface of the second network card through the third connector 13 and the fourth connector 14 in sequence, and the second interface of the second central processor 22 is connected to the second interface of the second network card, so that in the second mode, the second central processor 22 uses the second network card.

[0048] Specifically, the first mode is described first. In the first mode, the first central processing unit 21 and the second central processing unit 22 share the same shared network card to meet the requirement that CPUs in different motherboards share the same network card.

[0049] In this regard, in the first mode, the first interface of the first CPU 21 needs to be connected to the first connector 11. Figure 2 and Figure 3 In the example, the second interface is located on the left side of the first CPU 21, and the first interface is located on the upper left side of the first CPU 21. Figure 2 and Figure 3 In the example, the second CPU 22 has its second interface on the left side and the first interface on the upper left side.

[0050] The first interface of the first central processor 21 is connected to the first connector 11, the first connector 11 is connected to the transit connector, the transit connector is connected to the fourth connector 14 in the second mainboard, and the fourth connector 14 is connected to the first interface of the shared network card. Therefore, it can be seen that in the first mode, the first interface of the first central processor 21 is connected to the first interface of the shared network card in sequence through the first connector 11, the transit connector and the fourth connector 14. In addition, in the first mode, the second connector 12 in the first mainboard and the second interface of the first central processor 21 are not used, that is, they are in an idle state.

[0051] In the first mode, the second interface of the second CPU 22 is connected to the second interface of the shared network card, so that in the first mode, the first CPU 21 and the second CPU 22 share the same shared network card.

[0052] The second mode is described below. In the second mode, the central processing units in different motherboards can use their own network cards.

[0053] Specifically, in the second mode, the first interface of the first central processor 21 needs to be connected to the first interface of the first network card through the first connector 11 and the second connector 12 in sequence, and the second interface of the first central processor 21 can be directly connected to the second interface of the first network card. Therefore, it can be seen that in the second mode, the first network card is used by the first central processor 21 and is not shared with the second central processor 22.

[0054] Accordingly, in the second mode, the first interface of the second central processor 22 needs to be connected to the first interface of the second network card through the third connector 13 and the fourth connector 14 in sequence, and the second interface of the second central processor 22 can be directly connected to the second interface of the second network card. Therefore, it can be seen that in the second mode, the second network card is used by the second central processor 22 and is not shared with the first central processor 21.

[0055] It can be seen that in the second mode, there is no need to set up a transfer connector, and there is no need to connect any connector in the first mainboard with any connector in the second mainboard.

[0056] In addition, from the above description, it can be seen that when the user switches between the first mode and the second mode, it is not necessary to adjust the number of components in the first main board and the second main board, but only needs to change the connection relationship of the relevant connectors, which makes the present application solution low-cost and easy to implement.

[0057] In a specific embodiment of the present invention, the first interface and the second interface of the first central processor 21, the first interface and the second interface of the second central processor 22, the first interface and the second interface of the shared network card, the first interface and the second interface of the first network card, and the first interface and the second interface of the second network card are all 8-channel peripheral component rapid interconnection interfaces;

[0058] The transit connector, the first connector 11, the second connector 12, the third connector 13 and the fourth connector 14 are all connectors using an 8-channel peripheral component rapid interconnect interface.

[0059] It is understandable that in the solution of the present application, when making connections between connectors, or when making connections between any connector and the corresponding central processor, or when making connections between any connector and the corresponding network card, the same channel type should be used to ensure data transmission.

[0060] Furthermore, this implementation method takes into account that the current network card usually uses a PCIe (Peripheral Component Interconnect express) x16 structure, that is, a 16-channel peripheral component interconnect express interface is used, and in the present application scheme, it is required that the shared network card can be used by the first central processor 21 and the second central processor 22 at the same time. Therefore, the PCIex16 of the first central processor 21 in the first motherboard can be divided into two PCIex8s, one of which is the first interface of the first central processor 21, and the other PCIex8 is the second interface of the first central processor 21, that is, the first interface and the second interface of the first central processor 21 both use an 8-channel peripheral component interconnect express interface. In the first mode, the first interface of the first central processor 21 is connected to the first interface of the shared network card through the first connector 11, the transit connector and the fourth connector 14 in sequence. Therefore, the first connector 11, the transit connector, the fourth connector 14 and the first interface of the shared network card are all 8-channel peripheral component interconnect express interfaces. Figure 2 The x8 shown in the figure represents PCIex8.

[0061] Correspondingly, the PCIex16 of the second central processor 22 in the second motherboard also needs to be divided into two PCIex8s, one of which is the first interface of the second central processor 22, and the other PCIex8 is the second interface of the second central processor 22. In the first mode, the second interface of the second central processor 22 is connected to the second interface of the shared network card. Therefore, it can be seen that in the first mode, the PCIex16 of the shared network card is divided into two PCIex8s, one as the first interface of the shared network card, which is used by the first central processor 21, and the other as the second interface of the shared network card, which is used by the second central processor 22.

[0062] In the second mode, the first interface of the first central processor 21 is connected to the first interface of the first network card through the first connector 11 and the second connector 12 in sequence, and the first connector 11 and the second connector 12 are both in the form of PCIex8. At the same time, the second interface of the first central processor 21 is connected to the second interface of the first network card. Therefore, the PCIex16 of the first network card is divided into two PCIex8s as its first interface and second interface. Similarly, in the second mode, the PCIex16 of the second network card is divided into two PCIex8s as its first interface and second interface. The first interface of the second network card is connected to the first interface of the second central processor 22 through the third connector 13 and the fourth connector 14, and the second interface of the second network card can be directly connected to the second interface of the second central processor 22.

[0063] In a specific embodiment of the present invention, the first main board is arranged above the second main board, and in the first mode, the fourth connector 14 is connected to the female connector of the transit connector via a detachable male connector.

[0064] The multi-motherboard device in the present application scheme can usually be used in a multi-motherboard server, that is, in a multi-node server. For example, on a 2U4-node multi-node server, it can be divided into 2 groups, each with upper and lower nodes. The first motherboard and the second motherboard in the multi-motherboard device can constitute one group of upper and lower nodes. And this embodiment takes into account that in the first mode of the present application scheme, the fourth connector 14 in the second motherboard needs to be connected to the transit connector, while in the second mode, there is no need to set up a transit connector, and the fourth connector 14 does not need to be connected to the transit connector. Therefore, the fourth connector 14 can be connected to the female head of the transit connector through a detachable male head, so that when the transit connector does not need to be connected, the detachable male head can be removed.

[0065] In addition, this embodiment takes into account that in the first mode, the transit connector needs to be connected to the first motherboard, and the shared network card needs to be connected to the second motherboard. Since the slot position of the network card connected to the motherboard is usually set at the bottom in the height direction of the motherboard, the first motherboard can usually be set above the second motherboard, so that in the first mode, the fourth connector 14 can be connected to the female head of the transit connector through a detachable male head, and is not easily affected by the position of the shared network card.

[0066] In a specific embodiment of the present invention, the connection line between the first interface of the first central processor 21 and the first connector 11 is a connection line fixedly set in the first mainboard; the connection line between the second interface of the first central processor 21 and the external interface on the first mainboard is a connection line fixedly set in the first mainboard; the connection line between the second connector 12 and the external interface on the first mainboard is a connection line fixedly set in the first mainboard;

[0067] The connection line between the first interface of the second central processor 22 and the third connector 13 is a connection line fixedly set in the second mainboard; the connection line between the second interface of the second central processor 22 and the external interface on the second mainboard is a connection line fixedly set in the second mainboard; the connection line between the fourth connector 14 and the external interface on the second mainboard is a connection line fixedly set in the second mainboard.

[0068] In the present application, ideally, a certain motherboard should be able to be used both as the first motherboard and as the second motherboard. Therefore, in some implementations, the structures of the first motherboard and the second motherboard of the present application can be exactly the same, which is beneficial to reducing production and maintenance costs.

[0069] And this implementation method takes into account that although there are certain differences in the connection methods in the first mode and the second mode, some connection lines in the mainboard can be fixed, so that when switching between the first mode and the second mode, only the connection of a few lines needs to be adjusted.

[0070] Specifically, it can be seen from the above description of the first mode and the second mode that, regardless of the first mode or the second mode, the first interface of the first central processor 21 needs to be connected to the first connector 11, so the connection line between the two is a connection line fixedly set in the first mainboard. Since the second mainboard also needs to adopt the same design, the connection line between the first interface of the second central processor 22 and the third connector 13 will also be fixedly set in the second mainboard. In addition, it should be noted that, although in the first mode, the connection line between the second central processor 22 and the third connector 13 exists, since the third connector 13 is idle at this time, it will not cause any impact on the second central processor 22.

[0071] Correspondingly, the connection line between the second interface of the first central processor 21 and the external interface on the first mainboard, and the connection line between the second connector 12 and the external interface on the first mainboard, are both fixedly arranged in the first mainboard. Although in the first mode, these two connection lines are idle, it will not cause any impact. Since the second mainboard will also adopt the same design, the connection line between the second interface of the second central processor 22 and the external interface on the second mainboard, and the connection line between the fourth connector 14 and the external interface on the second mainboard, are both fixedly arranged in the second mainboard.

[0072] It can be seen that in this implementation, the first main board and the second main board are supported to adopt exactly the same structure, which facilitates the production and operation and maintenance of the main board, and some connection lines in the main board are fixed, thereby improving the convenience of switching between the first mode and the second mode.

[0073] In addition, for ease of viewing, Figure 2 and Figure 3 In FIG. 1 , the fixed lines in this implementation mode are bolded.

[0074] Furthermore, in a specific embodiment of the present invention, in the first mode, the connection line between the first connector 11 and the transit connector, and the connection line between the transit connector and the fourth connector 14 are both detachable connection lines.

[0075] In the above implementation, some of the connection lines in the first board and the second board are fixedly set on the corresponding mainboard, which will not affect the switching between the first mode and the second mode, and achieves the purpose of keeping the structure of the first mainboard and the second mainboard consistent. In this implementation, it is considered that in the first mode, in addition to the connection lines fixedly set on the corresponding mainboard, the connection lines between the first connector 11 and the transit connector, and the connection lines between the transit connector and the fourth connector 14 are also needed. Both of these connection lines can be set as detachable connection lines to ensure the convenience of disassembly. For example, through the design of male and female heads, the detachable connection lines can be easily realized.

[0076] In a specific embodiment of the present invention, in the second mode, the connection line between the first connector 11 and the second connector 12, and the connection line between the third connector 13 and the fourth connector 14 are both detachable connection lines.

[0077] Similar to the above embodiment, in the second mode, in addition to the above-mentioned connection lines fixed on the corresponding mainboard, it is also necessary to use the connection line between the first connector 11 and the second connector 12, and the connection line between the third connector 13 and the fourth connector 14. Both of these connection lines can be set as detachable connection lines to ensure the convenience of disassembly. For example, through the design of male and female heads, the detachable connection line can be easily realized.

[0078] The specific structural form of each connector of the present application solution can be set and adjusted according to actual needs. In a specific implementation of the present invention, considering the convenience of production, the structures of the transit connector, the first connector 11, the second connector 12, the third connector 13 and the fourth connector 14 can be the same, and considering that the multi-channel input and output connector (MCIO connector) can provide high-speed and stable data transmission to meet the use requirements in actual applications, therefore, each connector can be a multi-channel input and output connector.

[0079] In a specific embodiment of the present invention, the transit connector is a right-angle connector, and the first connector 11, the second connector 12, the third connector 13 and the fourth connector 14 are all vertical connectors.

[0080] This embodiment takes into account that the first connector 11, the second connector 12, the third connector 13 and the fourth connector 14 are all arranged in the mainboard, and in order to facilitate connection to the relevant external interfaces of the mainboard, the first connector 11, the second connector 12, the third connector 13 and the fourth connector 14 can usually be set as vertical connectors. The transit connector is usually arranged on the side of the first mainboard and needs to be connected to the fourth connector 14 of the second mainboard, so in order to facilitate wiring, it can usually be set as a right-angle connector.

[0081] In a specific embodiment of the present invention, in the first mode, the transit connector is fixed to the outside of the first mainboard by a preset fixing component.

[0082] This embodiment takes into account that, in the first mode, the transit connector needs to connect the first connector 11 in the first main board and the fourth connector 14 in the second main board. To facilitate the placement of the transit connector, in the first mode, the transit connector can be fixed to the outside of the first main board by a preset fixing component, such as being fixed to the outside of the first main board by a snap-on fixing component, which can achieve both fixation and easy disassembly.

[0083] In a specific embodiment of the present invention, the first central processor 21 may also be used to: in the first mode, when a connection interruption or bandwidth degradation with the shared network card is detected, execute a preset fault handling operation; in the second mode, when a connection interruption or bandwidth degradation with the first network card is detected, execute a fault handling operation;

[0084] The second central processor 22 is further used to: in the first mode, when a connection interruption or bandwidth degradation with the shared network card is detected, perform a fault handling operation; in the second mode, when a connection interruption or bandwidth degradation with the second network card is detected, perform a fault handling operation.

[0085] This implementation takes into account that in both the first mode and the second mode, the first central processor 21 can detect whether the network card connected to itself has a connection interruption or bandwidth degradation. If so, it can execute preset fault handling operations to try to resolve the fault and improve system reliability.

[0086] Similarly, in the first mode and in the second mode, the second central processor 22 can also detect whether the network card connected to itself has a connection interruption or bandwidth degradation. If so, it can execute preset fault handling operations to try to resolve the fault and improve system reliability.

[0087] The specific content of the fault handling operation performed can be set and adjusted according to actual needs. For example, in a specific implementation of the present invention, it can include: recording the fault time, and automatically restarting the mainboard when the restart conditions are met.

[0088] This implementation is designed to take into account that, in some cases, the aforementioned connection interruption or bandwidth degradation can be resolved by automatically restarting the mainboard, and the fault handling operation is simple and convenient to implement. Of course, the mainboard will be automatically restarted only when the restart conditions are met, for example, the mainboard will be automatically restarted only when the current business is completed to avoid affecting the business. Other restart conditions can be set in other occasions.

[0089] In a specific embodiment of the present invention, please refer to Figure 4 The first main board may also include a switching switch 30 and a third central processor 40, wherein the first end of the switching switch 30 is connected to the external interface on the first main board, the second end of the switching switch 30 is connected to the second interface of the first central processor 21, and the third end of the switching switch 30 is connected to the third central processor 40.

[0090] In the first mode and the second mode, the first end of the switch 30 is connected to the second end thereof. In the third mode, the first end of the switch 30 is connected to the third end thereof, so that in the third mode, the first central processor 21 and the third central processor 40 share the first network card. It can be understood that in the third mode, the first interface of the first central processor 21 needs to be connected to the first interface of the first network card through the first connector 11 and the second connector 12 in sequence.

[0091] In this implementation, two different CPUs in the same motherboard are supported to share the same network card. In the first mode and the second mode, the first end of the switch 30 is connected to the second end thereof, which can be set as the default state of the switch 30. Moreover, the switch can be switched to a non-default state under the direct or indirect control of the first CPU 21 or the third CPU 40, that is, the first end of the switch 30 is connected to the third end thereof.

[0092] In this implementation mode, it is understandable that in the third mode, the first network card needs to be configured to support sharing. In addition, it is understandable that the first motherboard is taken as an example here, and the second motherboard also supports such a design.

[0093] In a specific embodiment of the present invention, each external interface in the first mainboard is arranged in the direction of the cabinet rear window, and each interface in the second mainboard is arranged in the direction of the cabinet rear window;

[0094] In the first mode, the shared network card and the transit connector are both arranged in the direction of the rear window of the cabinet; in the second mode, the first network card and the second network card are both arranged in the direction of the rear window of the cabinet.

[0095] In some traditional solutions, it is possible to support CPUs in different motherboards to share the same network card. Figure 5 , CPU 0 and CPU 1 are on different motherboards and can be shared Figure 5 Similarly, CPU 2 and CPU 3 are on different motherboards and can be shared. Figure 5 Shared network card 2 in. Figure 5 In such a structure, in order to ensure that plugging in or out of any motherboard does not affect other motherboards, the shared network card is placed outside the motherboard. Figure 5It can be seen that the CPU 0 and CPU 1 from different motherboards need to be connected to the shared network card 1 through the mid-backplane, so the management interface and other I / O ports of the CPU 0 and CPU 1 are set on a different side of the server from the shared network card 1. For example, the shared network card 1 is usually set on the front window of the server, and the management interface and other I / O ports of the CPU 0 and CPU 1 are set on the back window of the server.

[0096] for Figure 5 This solution is very inconvenient for operators, as it requires one person to operate the front and rear windows or one person to operate back and forth. There are cables on both sides, and when the cabinet is mounted, if the front and rear doors are not deep enough, it is easy to cause the cabinet door to be unable to close.

[0097] In the solution of the present application, it can be seen from the above analysis and description that the solution of the present application realizes flexible switching between the first mode and the second mode through the design of relevant connectors and line structures, without any other restrictions, and can also support the interfaces on each network card and the relevant central processor to be set on the same side of the server, thereby shortening the required cabinet depth.

[0098] Specifically, in this embodiment, each external interface in the first mainboard is arranged in the direction of the cabinet rear window, and each interface in the second mainboard is arranged in the direction of the cabinet rear window;

[0099] In the first mode, the shared network card and the transit connector are both arranged in the direction of the rear window of the cabinet; in the second mode, the first network card and the second network card are both arranged in the direction of the rear window of the cabinet.

[0100] In this implementation, each external interface in the first motherboard and each interface in the second motherboard are arranged in the direction of the rear window of the server cabinet. Similarly, the shared network card and the transfer connector in the first mode and the first network card and the second network card in the second mode are arranged in the direction of the rear window of the server cabinet.

[0101] The specific number and type of each external interface in the first mainboard can be arranged according to actual needs, for example Figure 6 In a specific implementation manner, a front view of a multi-motherboard device viewed from the rear window of a cabinet is shown. Figure 6 The figure shows the USB (Universal Serial Bus) interface, RJ45 interface, and a slot for connecting other PCIe plug-in cards. Figure 6 This is just an example. In actual applications, there may be many external interfaces on the motherboard, such as debug interface, etc.

[0102] In a specific embodiment of the present invention, each external button and indicator light in the first mainboard are arranged in the direction of the cabinet rear window, and each external button and indicator light in the second mainboard are arranged in the direction of the cabinet rear window.

[0103] This implementation takes into account that each external button and indicator light in the first mainboard can also be set in the direction of the cabinet rear window, so that the staff can operate uniformly at the rear window. The external button can include, for example Figure 6 The reset button, etc., as exemplified in the above, the indicator light may include a positioning indicator light, etc. Similarly, each external button and indicator light in the second mainboard may also be arranged in the direction of the rear window of the cabinet.

[0104] By applying the technical solution provided by the embodiment of the present invention, the multi-motherboard device can support both the first mode and the second mode through the design of the corresponding connector. Specifically, the first motherboard is provided with a first central processor 21, a first connector 11 and a second connector 12, and the second motherboard is provided with a second central processor 22, a third connector 13 and a fourth connector 14. In the first mode, the first interface of the first central processor 21 needs to be connected to the first interface of the shared network card through the first connector 11, the transit connector and the fourth connector 14 in sequence, and the second interface of the second central processor 22 is connected to the second interface of the shared network card. It can be seen that in the first mode, the first central processor 21 and the second central processor 22 share the shared network card, which satisfies the need for CPUs in different motherboards to share the same network card. In the second mode, the first interface of the first central processor 21 can be connected to the first interface of the first network card through the first connector 11 and the second connector 12 in sequence, and the second interface of the first central processor 21 is connected to the second interface of the first network card. It can be seen that in the second mode, the first network card is used by the first central processor 21 and is not shared with the second central processor 22. Correspondingly, in the second mode, the first interface of the second central processor 22 is connected to the first interface of the second network card through the third connector 13 and the fourth connector 14 in sequence, and the second interface of the second central processor 22 is connected to the second interface of the second network card. It can be seen that in the second mode, the second network card is used by the second central processor 22 and is not shared with the first central processor 21. Therefore, in the second mode, the purpose of CPUs in different motherboards using their own network cards is met. And it can be seen that the present application scheme does not need to provide users with 2 sets of different servers, but can achieve the purpose through mode switching. When switching modes, it is not necessary to adjust the number of components in the first motherboard and the second motherboard, but only to change the connection relationship of the relevant connectors, so that the present application scheme has a low cost.

[0105] In summary, the present application scheme can realize networking, which can support CPUs in different motherboards to share the same network card, and can also support CPUs in different motherboards to use their own network cards, and reduces costs.

[0106] Corresponding to the above embodiment of the multi-motherboard device, the embodiment of the present invention further provides a multi-node server, which can be referred to in correspondence with the above. The multi-node server may include a multi-motherboard device as in any of the above embodiments. The multi-node server can realize networking, which can support CPUs in different motherboards to share the same network card, and can support CPUs in different motherboards to use their own network cards, and effectively reduce costs.

[0107] Further, in a specific embodiment of the present invention, the multi-node server may include N multi-motherboard devices, where N is a positive integer not less than 2, and the structures of the multi-motherboard devices are consistent, so as to facilitate implementation. For example, for the 2U4-node multi-node server described above, N is equal to 2, and the multi-node server includes 2 multi-motherboard devices, each of which includes two upper and lower motherboards, and the structures of the two multi-motherboard devices are consistent.

[0108] It should also be noted that, in this application, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0109] Those skilled in the art may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in this application can be implemented by electronic hardware, computer software or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to the function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention. Specific examples are used in this application to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the technical solution and its core idea of ​​the present invention. It should be pointed out that for those of ordinary skill in the art, without departing from the principles of the present invention, the present invention can also be improved and modified in a number of ways, and these improvements and modifications also fall within the scope of protection of the present invention.

Claims

1. A multi-motherboard device, characterized in that: include: A first central processing unit, a first connector and a second connector are arranged in a first mainboard; A second central processing unit, a third connector and a fourth connector are arranged in the second mainboard; In the first mode, the first interface of the first central processor is connected to the first interface of the shared network card through the first connector, the transit connector and the fourth connector in sequence, and the second interface of the second central processor is connected to the second interface of the shared network card, so that in the first mode, the shared network card is shared by the first central processor and the second central processor; In the second mode, the first interface of the first central processor is connected to the first interface of the first network card through the first connector and the second connector in sequence, and the second interface of the first central processor is connected to the second interface of the first network card, so that in the second mode, the first central processor uses the first network card; In the second mode, the first interface of the second central processor is connected to the first interface of the second network card through the third connector and the fourth connector in sequence, and the second interface of the second central processor is connected to the second interface of the second network card, so that in the second mode, the second network card is used by the second central processor.

2. The multi-motherboard device according to claim 1, characterized in that: The first interface and the second interface of the first central processor, the first interface and the second interface of the second central processor, the first interface and the second interface of the shared network card, the first interface and the second interface of the first network card, and the first interface and the second interface of the second network card are all 8-channel peripheral component rapid interconnection interfaces; The transit connector, the first connector, the second connector, the third connector and the fourth connector are all connectors using an 8-channel peripheral component rapid interconnect interface.

3. The multi-motherboard device according to claim 1, characterized in that: The first main board is arranged above the second main board. In the first mode, the fourth connector is connected to the female connector of the transit connector via a detachable male connector.

4. The multi-motherboard device according to claim 1, characterized in that: The connection line between the first interface of the first central processor and the first connector is a connection line fixedly set in the first mainboard; the connection line between the second interface of the first central processor and the external interface on the first mainboard is a connection line fixedly set in the first mainboard; The connection line between the second connector and the external interface on the first mainboard is a connection line fixedly arranged in the first mainboard; The connection line between the first interface of the second central processing unit and the third connector is a connection line fixedly arranged in the second mainboard; The connection line between the second interface of the second central processor and the external interface on the second mainboard is a connection line fixedly arranged in the second mainboard; The connection line between the fourth connector and the external interface on the second mainboard is a connection line fixedly arranged in the second mainboard.

5. The multi-motherboard device according to claim 4, characterized in that: In the first mode, the connection line between the first connector and the relay connector, and the connection line between the relay connector and the fourth connector are both detachable connection lines.

6. The multi-motherboard device according to claim 4, characterized in that: In the second mode, the connection line between the first connector and the second connector, and the connection line between the third connector and the fourth connector are both detachable connection lines.

7. The multi-motherboard device according to claim 1, characterized in that: The transfer connector, the first connector, the second connector, the third connector and the fourth connector all have the same structure and are all multi-channel input and output connectors.

8. The multi-motherboard device according to claim 1, characterized in that: The first central processor is further configured to: in the first mode, when a connection interruption or bandwidth degradation with the shared network card is detected, execute a preset fault handling operation; in the second mode, when a connection interruption or bandwidth degradation with the first network card is detected, execute the fault handling operation; The second central processor is also used to: in the first mode, when it detects that the connection with the shared network card is interrupted or the bandwidth is degraded, execute the fault handling operation; in the second mode, when it detects that the connection with the second network card is interrupted or the bandwidth is degraded, execute the fault handling operation.

9. The multi-motherboard device according to claim 8, characterized in that: The fault handling operations performed include: The fault time is recorded, and when the restart conditions are met, the mainboard is automatically restarted.

10. The multi-motherboard device according to claim 1, characterized in that: The transit connector is a right-angle connector, and the first connector, the second connector, the third connector and the fourth connector are all vertical connectors.

11. The multi-motherboard device according to claim 1, characterized in that: In the first mode, the transit connector is fixed to the outside of the first mainboard by a preset fixing component.

12. The multi-motherboard device according to any one of claims 1 to 11, characterized in that: Each external interface in the first mainboard is arranged in the direction of the cabinet rear window, and each interface in the second mainboard is arranged in the direction of the cabinet rear window; In the first mode, the shared network card and the transit connector are both arranged in the direction of the rear window of the cabinet; in the second mode, the first network card and the second network card are both arranged in the direction of the rear window of the cabinet.

13. The multi-motherboard device according to claim 12, characterized in that: Each external button and indicator light in the first main board are arranged in the direction of the rear window of the cabinet, and each external button and indicator light in the second main board are arranged in the direction of the rear window of the cabinet.

14. A server, characterized in that: Comprising a multi-motherboard device as described in any one of claims 1 to 13.

15. The server according to claim 14, characterized in that: The server includes N multi-motherboard devices, where N is a positive integer not less than 2, and each of the multi-motherboard devices has the same structure.

Images