A data processing system, server, method, product and storage medium

Abstract

The present invention discloses a data processing system, server, method, product and storage medium, which relates to the field of servers. In order to solve the problem of resource waste and limited adaptation scenarios caused by selecting a network chip with a higher rate, the computing management unit in the system is used to determine the current network transmission state, generate a replacement instruction if the current network transmission state meets the replacement condition, and identify the network transmission device connected to the slot after responding to the replacement instruction through a first data transmission channel, and bind the network transmission device to the data processing unit; including at least one network transmission unit of the network transmission device, the network transmission device is used to transmit the network data between the switching device and the target network through a second data transmission channel after binding with the data processing unit. The present invention can meet the needs of users to select and expand by themselves, is not easy to cause resource waste, and improves the flexibility and adaptability of the data processing system.

Description

Technical Field

[0001] The present invention relates to the field of servers, and in particular to a data processing system, a server, a method, a product and a storage medium. Background Art

[0002] In the digital age, the demand for data processing continues to grow, especially in the fields of artificial intelligence, big data analysis, and network communications. In order to meet these needs, various component manufacturers and users began to develop DPUs (Data Processing units) on their own, encapsulating the data processing unit and the network transmission unit into an integral PCIE (Peripheral Component Interconnect Express, a standard high-speed serial bus for external device interconnection) component and installing it on the server. Since the existing ordinary network cards on the market cannot replace the network transmission unit in the DPU, this leads to increased design and production costs. In addition, once the DPU is designed and produced, its upper limit on network transmission rate is limited by the network chip selected for design. Therefore, when designing, higher-speed network chips are usually selected to support backward compatibility, but this is prone to waste of resources and is limited in adaptation scenarios.

[0003] Therefore, how to provide a solution to the above technical problems is a problem that those skilled in the art need to solve at present. Summary of the invention

[0004] The purpose of the present invention is to provide a data processing system, server, method, product and storage medium, which can realize the needs of users to select and expand configurations by themselves, are not likely to cause waste of resources, and improve the flexibility and adaptability of the data processing system.

[0005] To solve the above technical problems, the present invention provides a data processing system, comprising: a switching device integrating a data processing unit, a computing management unit and at least one slot, a first data transmission channel being provided between the slot and the computing management unit, a second data transmission channel being provided between the slot and the data processing unit, the computing management unit being used to determine a current network transmission status, and generating a replacement instruction if the current network transmission status meets a replacement condition, and identifying a network transmission device connected to the slot after responding to the replacement instruction through the first data transmission channel, and binding the network transmission device to the data processing unit; a network transmission unit comprising at least one of the network transmission devices, the network transmission device being used to transmit network data between the switching device and a target network through the second data transmission channel after being bound to the data processing unit.

[0006] Optionally, the process of determining the current network transmission status includes: obtaining the current network rate of the network transmission device currently bound to the data processing unit; determining whether the current network rate is within the target network rate range corresponding to the current application scenario; if so, determining that the current network transmission status is a normal transmission status; if not, determining that the current network transmission status is an abnormal transmission status.

[0007] Optionally, the replacement condition includes a first replacement condition, and the switching device is further configured to determine that the current network transmission state satisfies the first replacement condition when the current network transmission state is the abnormal transmission state.

[0008] Optionally, the replacement instruction includes a first prompt instruction. If the current network transmission status meets the replacement condition, the replacement instruction is generated. The process of identifying the network transmission device connected to the slot after responding to the replacement instruction through the first data transmission channel includes: if the current network transmission status meets the first replacement condition, the first prompt instruction is generated so that the user can unplug the network transmission device currently bound to the data processing unit from the slot based on the first prompt instruction, and insert a new network transmission device into the slot, so that the computing management unit can identify the new network transmission device through the first data transmission channel, and the network rate of the new network transmission device is within the target network rate range.

[0009] Optionally, the replacement instruction includes a first switching instruction. If the current network transmission status meets the replacement condition, the replacement instruction is generated, and the process of identifying the network transmission device connected to the slot after responding to the replacement instruction through the first data transmission channel includes: if the current network transmission status meets the first replacement condition, and the network transmission unit includes multiple network transmission devices, determine the first target transmission device whose network rate is within the target network rate range among the multiple network transmission devices, and generate the first switching instruction based on the first target transmission device. The first switching instruction is used to disconnect the electrical connection between the network transmission device currently bound to the data processing unit and its corresponding slot, and to conduct the electrical connection between the first target transmission device and its corresponding slot, so that the computing management unit can identify the first target transmission device through the first data transmission channel.

[0010] Optionally, the process of determining the current network transmission status includes: obtaining the usage status of a second target transmission device; the second target transmission device is a network transmission device currently bound to the data processing unit; the usage status includes a normal state or a fault state; if the usage status is the normal state, determining that the current network transmission status is a normal transmission state; if the usage status is the fault state, determining that the current network transmission status is an abnormal transmission state.

[0011] Optionally, the replacement condition includes a second replacement condition, and the switching device is further configured to determine that the current network transmission state satisfies the second replacement condition when the current network transmission state is the abnormal transmission state.

[0012] Optionally, the replacement instruction includes a second prompt instruction. If the current network transmission status meets the replacement condition, a replacement instruction is generated. The process of identifying the network transmission device connected to the slot in response to the replacement instruction through the first data transmission channel includes: if the current network transmission status meets the second replacement condition, the second prompt instruction is generated so that the user can unplug the network transmission device currently bound to the data processing unit from the slot based on the second prompt instruction, and insert a new network transmission device into the slot, so that the computing management unit can identify the new network transmission device through the first data transmission channel, and the usage status of the new network transmission device is normal.

[0013] Optionally, the replacement instruction includes a second switching instruction. If the current network transmission status meets the replacement condition, the replacement instruction is generated, and the process of identifying the network transmission device connected to the slot after responding to the replacement instruction through the first data transmission channel includes: if the current network transmission status meets the second replacement condition, and the network transmission unit includes multiple network transmission devices, determine the third target transmission device whose usage status is normal among the multiple network transmission devices, and generate the second switching instruction based on the third target transmission device. The second switching instruction is used to disconnect the electrical connection between the network transmission device currently bound to the data processing unit and its corresponding slot, and to conduct the electrical connection between the third target transmission device and its corresponding slot, so that the computing management unit can identify the third target transmission device through the first data transmission channel.

[0014] Optionally, the network transmission device is a network transmission device whose firmware is custom compiled based on an application scenario.

[0015] Optionally, the data processing unit is used to receive network data transmitted by the network transmission device bound to itself through the second data transmission channel, and process the network data according to a preset processing logic to obtain result data, and send the result data to the network transmission device bound to itself through the second data transmission channel.

[0016] Optionally, the adapter device further includes a plurality of interfaces, and the plurality of interfaces are used to transmit a power supply signal and at least one communication signal between the adapter device and a server mainboard.

[0017] Optionally, the server mainboard is used to perform a first handshake operation with the adapter device after being powered on, so that a local processor of the server mainboard and a data processing unit in the adapter device establish a communication connection.

[0018] Optionally, the adapter device is used to initiate a second handshake operation between itself and the server mainboard after identifying information of a network transmission device connected to itself, so as to respond to a query instruction sent by the server mainboard and return reply information corresponding to the query instruction; the server mainboard is also used to configure the working parameters of the adapter device based on the received reply information; the working parameters include at least one of the bandwidth of the first data transmission channel, the bandwidth of the second data transmission channel, the enable function parameters of the adapter device, the disable function parameters of the adapter device, and the memory mapping address corresponding to the adapter device.

[0019] Optionally, the data processing system also includes a storage unit, which includes a storage device provided on the adapter device and / or a hard disk connected to the server mainboard; the data processing unit is also used to select the storage device and / or the hard disk to store the result data according to the storage status of the storage unit.

[0020] Optionally, the computing management unit is also used to detect the operating status and operating information of the data processing system, and the operating status includes the current network transmission status.

[0021] Optionally, the operating information includes a current substrate temperature; the adapter device also includes a heat dissipation unit; the calculation management unit is also used to control the operation of the heat dissipation unit according to the current substrate temperature, so that the current substrate temperature of the adapter device is within a target temperature range.

[0022] In order to solve the above technical problems, the present invention further provides a server, comprising a server mainboard and a data processing system as described in any one of the above items connected to the server mainboard.

[0023] In order to solve the above technical problems, the present invention also provides a data processing method, which is applied to a data processing system as described in any one of the above texts, wherein the data processing system includes a network transmission unit, the network transmission unit includes at least one network transmission device, and the data processing system also includes a switching device integrating a data processing unit, a computing management unit and at least one slot, a first data transmission channel is provided between the slot and the computing management unit, and a second data transmission channel is provided between the slot and the data processing unit. The data processing method includes: determining the current network transmission status through the computing management unit, generating a replacement instruction if the current network transmission status meets the replacement condition, and identifying the network transmission device connected to the slot after responding to the replacement instruction through the first data transmission channel, and binding the network transmission device to the data processing unit; after the network transmission device is bound to the data processing unit, transmitting the network data between the switching device and the target network through the second data transmission channel.

[0024] In order to solve the above technical problem, the present invention also provides a computer program product, including a computer program / instruction, which implements the steps of the data processing method described above when executed by a processor.

[0025] In order to solve the above technical problem, the present invention further provides a non-volatile storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the data processing method described above are implemented.

[0026] The present invention provides a data processing system, which separates a data processing unit from a network transmission unit. The network transmission device in the network transmission unit is detachably connected to a slot on a switching device. The network transmission device connected to the slot can be changed according to actual needs, thereby realizing flexible switching of network rates. Different network transmission devices can be bound to the data processing unit to form a new data processing system, realizing the user's self-selection and expansion needs, not easily causing waste of resources, and improving the flexibility and adaptability of the data processing system.

[0027] The present invention also provides a server, a data processing method, a computer program product and a non-volatile storage medium, which have the same beneficial effects as the above data processing system. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. 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.

[0029] Figure 1 This is a structural schematic diagram of a data processing system provided by the present invention.

[0030] Figure 2 This is a schematic diagram of the structure of another data processing system provided by the present invention.

[0031] Figure 3 The present invention is a schematic structural diagram of a switching device provided by the present invention.

[0032] Figure 4 This is a schematic structural diagram of another switching device provided by the present invention.

[0033] Figure 5 The present invention provides a flowchart of the steps of a data processing method.

[0034] Figure 6 A schematic diagram of the structure of a non-volatile storage medium provided by the present invention. DETAILED DESCRIPTION

[0035] The core of the present invention is to provide a data processing system, server, method, product and storage medium, which can realize the needs of users to select and expand configurations by themselves, is not easy to cause waste of resources, and improves the flexibility and adaptability of the data processing system.

[0036] 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 ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0037] First, please refer to Figure 1 The present invention provides a data processing system, including: a switching device 1 integrated with a data processing unit 11, a computing management unit 13 and at least one slot 12, a first data transmission channel between the slot 12 and the computing management unit 13, a second data transmission channel between the slot 12 and the data processing unit 11, the computing management unit 13 is used to determine the current network transmission state, generate a replacement instruction if the current network transmission state meets the replacement condition, and identify the network transmission device 21 connected to the slot 12 after responding to the replacement instruction through the first data transmission channel, and bind the network transmission device 21 to the data processing unit 11; a network transmission unit 2 including at least one network transmission device 21, the network transmission device 21 is used to transmit network data between the switching device 1 and the target network through the second data transmission channel after being bound to the data processing unit 11.

[0038] The data processing system in this embodiment is a separate DPU, which consists of a network transmission unit 2 and a switching device 1. Among them, the switching device 1 is integrated with a data processing unit 11, a computing management unit 13 and at least one slot 12. The data processing unit 11 is the data processing service core of the DPU. On the one hand, as the HOST end (host end), it and the network transmission unit 2 together form the DPU, realizing the unloading engine of the CPU (Central Processing Unit), taking over infrastructure layer services such as network virtualization and hardware resource pooling, releasing the CPU computing power to the upper layer application, and realizing many DPU functions such as flow table service unloading, Virtio (Virtual Input / Output, virtual input / output) hardware acceleration, full unloading of security-related functions, data decompression, etc. based on business needs. The network transmission unit 2 is composed of an ordinary network card or other network transmission device 21, providing network data transmission function. The user can realize flexible adaptation of the DPU by replacing ordinary network cards of different models, solving the problem of repeated development and unchangeability of the network transmission unit 2 in the DPU.

[0039] Specifically, this embodiment decouples the DPU architecture, removes the network transmission related functions that can be implemented by a common network card, and integrates the remaining DPU on a substrate, that is, the switching device 1 in this embodiment. Figure 1 As shown, there is a first data transmission channel between the computing management unit 13 on the switching device 1 and the slot 12, and a second data transmission channel between the data processing unit 11 on the switching device 1 and the slot 12. The computing management unit 13 monitors the current network transmission status of the data transmission system, and generates a replacement instruction if the current network transmission status meets the replacement conditions. The replacement instruction can be used to prompt the user to replace the network transmission device 21, or it can directly realize automatic switching. The network transmission device 21 connected to the slot 12 after responding to the replacement instruction is the network transmission device 21 after the change according to the replacement instruction.

[0040] Correspondingly, at least one slot 12 is also provided on the switching device 1 in this embodiment, and a PCIE slot can be selected specifically. Each slot 12 is used to connect a network transmission device 21, so that the relevant information of the network transmission device 21 connected to the slot 12 is transmitted to the corresponding data processing unit 11 and the computing management unit 13 on the switching device 1 through the management bus in the slot 12 through the data transmission channel, so as to bind the data processing unit 11 and the changed network transmission device 21. When any network transmission device 21 is bound to the data processing unit 11, the network data transmission function can be realized through the network transmission device 21 to form a new DPU device. Among them, the network transmission device 21 can adopt a standard ordinary network card that has been mass-produced at present, and its interface type includes but is not limited to PCIE, OCP (Open Compute Project) (depending on the interface type extended by the switching device 1 integrated with the data processing unit 11). By customizing the firmware of the ordinary network card, it can be coupled with the switching device 1 integrated with the data processing unit 11 to form a complete intelligent network card DPU to work together.

[0041] It can be seen that in this embodiment, the data processing unit 11 and the network transmission unit 2 are separated, and the network transmission device 21 in the network transmission unit 2 is detachably connected to the slot 12 on the switching device 1. The network transmission device 21 connected to the slot 12 can be changed according to actual needs, thereby realizing flexible switching of the network rate, and binding different network transmission devices 21 with the data processing unit 11 can form a new data processing system, realizing the user's own selection and expansion needs, not easily causing waste of resources, and improving the flexibility and adaptability of the data processing system.

[0042] Based on the above embodiments:

[0043] In an exemplary embodiment, the process of determining the current network transmission status includes: obtaining the current network rate of the network transmission device 21 currently bound to the data processing unit 11; determining whether the current network rate is within the target network rate range corresponding to the current application scenario; if so, determining that the current network transmission status is a normal transmission status; if not, determining that the current network transmission status is an abnormal transmission status.

[0044] In this embodiment, if the network interface rate of the DPU can no longer meet the current environmental requirements, the network transmission unit 2 can be replaced, that is, the ordinary network card can be replaced to achieve the expansion and improvement of the DPU network rate, thereby realizing the modularization and standardization of the DPU network transmission unit 2, and improving the flexibility and adaptability of the DPU application. Specifically, first monitor in real time whether the network rate of the network transmission device 21 currently bound to the data processing unit 11 is within the target network rate range corresponding to the current application scenario. If so, determine that the current network transmission state is a normal transmission state, and there is no need to replace the network transmission device 21. If not, determine that the current network transmission state is an abnormal transmission state, which means that the currently used network transmission device 21 cannot meet the requirements, and the switching device 1 determines that the current network transmission state meets the first replacement condition.

[0045] In an exemplary embodiment, the replacement instruction includes a first prompt instruction. If the current network transmission status meets the replacement condition, the replacement instruction is generated, and the process of identifying the network transmission device 21 connected to the slot 12 in response to the replacement instruction through the first data transmission channel includes: if the current network transmission status meets the first replacement condition, a first prompt instruction is generated so that the user can unplug the network transmission device 21 currently bound to the data processing unit 11 from the slot 12 based on the first prompt instruction, and insert the new network transmission device 21 into the slot 12, so that the computing management unit 13 can identify the new network transmission device 21 through the first data transmission channel, and the network rate of the new network transmission device 21 is within the target network rate range.

[0046] In this embodiment, after receiving the first prompt instruction, the user determines that the interface transmission rate of the currently used network transmission device 21 does not meet the requirements of the current application scenario, unplugs the network transmission device 21 from the slot 12, and selects a network transmission device 21 whose network rate meets the requirements of the current application scenario and inserts it into the corresponding slot 12 to achieve capacity expansion and improvement of the network rate.

[0047] In an exemplary embodiment, the replacement instruction includes a first switching instruction. If the current network transmission status meets the replacement condition, the replacement instruction is generated. The process of identifying the network transmission device 21 connected to the slot 12 after responding to the replacement instruction through the first data transmission channel includes: if the current network transmission status meets the first replacement condition, and the network transmission unit 2 includes multiple network transmission devices 21, a first target transmission device whose network rate is within the target network rate range is determined among the multiple network transmission devices 21, and a first switching instruction is generated based on the first target transmission device. The first switching instruction is used to disconnect the electrical connection between the network transmission device 21 currently bound to the data processing unit 11 and the corresponding slot 12, and to connect the electrical connection between the first target transmission device and the corresponding slot 12, so that the computing management unit 13 can identify the first target transmission device through the first data transmission channel.

[0048] In this embodiment, the network transmission unit 2 may include multiple ordinary network cards as hot standby at the same time. When the network rate of one of the ordinary network cards cannot meet the requirements of the current application scenario, the target network card that can meet the requirements of the current application scenario among the remaining ordinary network cards in the network transmission unit 2 is enabled for network transmission to achieve capacity expansion and improvement of the network rate.

[0049] It is understandable that by determining whether to replace the network transmission device 21 based on the network rate, possible bottlenecks of the network can be predicted in advance, so as to take preventive measures to avoid performance degradation. Moreover, replacing the network card before the network rate starts to decline can avoid business interruptions and additional costs caused by performance problems. Replacing the network card in time can immediately increase the network bandwidth and meet the bandwidth requirements of business growth. In addition, replacing the network card in advance can also reduce business interruptions caused by the decline in network rate and ensure the continuity and stability of the network.

[0050] In an exemplary embodiment, the process of determining the current network transmission status includes: obtaining the usage status of the second target transmission device; the second target transmission device is the network transmission device 21 currently bound to the data processing unit 11; the usage status includes a normal state or a fault state; if the usage status is a normal state, determining that the current network transmission status is a normal transmission state; if the usage status is a fault state, determining that the current network transmission status is an abnormal transmission state.

[0051] In this embodiment, the current network transmission status can also be determined based on the usage status of the second target transmission device. If the usage status is normal, the current network transmission status is determined to be a normal transmission status, and there is no need to replace the second target transmission device. If the usage status is a fault status, the current network transmission status is determined to be an abnormal transmission status, and the second target transmission device may not be able to implement the network data transmission function. It is determined that the current network transmission status meets the second replacement condition, and the second target transmission device needs to be replaced.

[0052] In an exemplary embodiment, the switching device 1 is further configured to determine whether the current network transmission state satisfies a second replacement condition when the current network transmission state is an abnormal transmission state.

[0053] In an exemplary embodiment, the replacement instruction includes a second prompt instruction. If the current network transmission status meets the replacement condition, the replacement instruction is generated, and the process of identifying the network transmission device 21 connected to the slot 12 in response to the replacement instruction through the first data transmission channel includes: if the current network transmission status meets the second replacement condition, a second prompt instruction is generated so that the user can unplug the network transmission device 21 currently bound to the data processing unit 11 from the slot 12 based on the second prompt instruction, and insert the new network transmission device 21 into the slot 12, so that the computing management unit 13 can identify the new network transmission device 21 through the first data transmission channel, and the usage status of the new network transmission device 21 is normal.

[0054] If the current network transmission status meets the second replacement condition, a second prompt instruction corresponding to replacing the faulty network card is generated, so that after the user receives the second prompt instruction, the network transmission device 21 in the faulty state can be unplugged from the slot 12, and a network transmission device 21 in a normal state can be selected and inserted into the corresponding slot 12 to realize flexible switching of the network transmission device 21.

[0055] In an exemplary embodiment, the replacement instruction includes a second switching instruction. If the current network transmission status meets the replacement condition, the replacement instruction is generated, and the process of identifying the network transmission device 21 connected to the slot 12 after responding to the replacement instruction through the first data transmission channel includes: if the current network transmission status meets the second replacement condition, and the network transmission unit 2 includes multiple network transmission devices 21, a third target transmission device with a normal usage status is determined among the multiple network transmission devices 21, and a second switching instruction is generated based on the third target transmission device. The second switching instruction is used to disconnect the electrical connection between the network transmission device 21 currently bound to the data processing unit 11 and the corresponding slot 12, and to conduct the electrical connection between the third target transmission device and the corresponding slot 12, so that the computing management unit 13 can identify the third target transmission device through the first data transmission channel.

[0056] The network transmission unit 2 can include multiple ordinary network cards as hot standby. When one of the ordinary network cards fails, the remaining ordinary network cards in the network transmission unit 2 are enabled for network transmission, and the fault information is transmitted to the user through the management unit to replace the faulty network card.

[0057] It can be understood that by monitoring the usage status of the network transmission device 21 to make a decision on replacement, potential hidden dangers of failure can be discovered, so that measures can be taken to prevent the occurrence of failures. Timely replacement of network cards in poor working condition can avoid performance degradation and business interruption caused by network card failures, and can also ensure that all network cards are in the best working condition, optimize the use of network resources, improve the overall performance of the network, improve the reliability of the network, and reduce the possibility of failures.

[0058] In an exemplary embodiment, the data processing unit 11 is used to receive network data transmitted by the network transmission device 21 bound to itself through a second data transmission channel, and process the network data according to a preset processing logic to obtain result data, and send the result data to the network transmission device 21 bound to itself through the second data transmission channel.

[0059] In this embodiment, the data processing unit 11 includes SoC (System on Chip), CPLD (Complex Programmable Logic Device), storage device (memory), BIOS (Basic Input / Output System), etc. This part is used to replace part of the data processing work of the mainboard CPU and realize the core business processing of the DPU.

[0060] When the network transmission device 21 is changed, the new network transmission device 21 is inserted into the PCIE Slot slot on the adapter device 1 integrated in the data processing unit 11. At this time, the newly replaced network transmission device 21 transmits information to the computing management unit 13 on the adapter device 1 integrated in the data processing unit 11 through the management bus in the PCIE slot 12. After the computing management unit 13 recognizes the new network transmission unit 2, it binds the new network transmission unit 2 with the data processing unit 11 according to the pre-written logic to form a new DPU device. The computing management unit 13 includes but is not limited to BMC (Baseboard Management Controller).

[0061] In an exemplary embodiment, the switching device 1 further includes a plurality of interfaces, and the plurality of interfaces are used to transmit a power supply signal and at least one communication signal between the switching device 1 and the server mainboard.

[0062] The adapter device 1 is connected to the server mainboard as an adapter board through a gold finger or a cable to obtain uplink communication signals (such as PCIE signals, I2C (Inter-Integrated Circuit) signals, etc.) and power supply signals. At the same time, one or more PCIE Slot slots are provided on the substrate to expand support for ordinary network cards and other PCIE devices as network transmission units 2.

[0063] In an exemplary embodiment, the server mainboard is used to perform a first handshake operation with the switching device 1 after being powered on, so that a local processor of the server mainboard and a data processing unit 11 in the switching device 1 establish a communication connection.

[0064] It can be understood that after the server is powered off or the data processing system is powered off separately, the network card to be installed will be installed through the PCIE\OCP standard interface and fixed by a locking device. After that, the connected data processing system will be connected to the corresponding interface on the server motherboard and the power supply and data protocol line connections will be completed.

[0065] After the switching device 1 is connected to the network transmission device 21, the server can be powered on directly or remotely, and then the separate DPU can be powered on. At this time, the following process will occur in chronological order: After the server is powered on, the first handshake operation between the server local data processor and the separate DPU will be completed according to the server startup self-test process. The process of the first handshake operation includes the power chip of the server motherboard completing the power distribution, the server firmware and the operating system completing the power-on and clock alignment with the module. After that, the server motherboard will perform link initialization and link training, configure and initialize the physical layer, port and related links of the device, so that normal data packet traffic can be carried out on the link. During the link initialization and training process, the LTSSM (Link Training and Status State Machine) guides the completion of functions such as device identification, rate negotiation, power management, reset, link width channel sequence negotiation, etc. After the server motherboard completes the connection with the separate DPU, it will start the information connection between the DPU and the network card. When the data processing unit 11 is connected to the local processor, the network transmission unit 2 is connected to the data processing unit 11 again as a separate secondary device. The link is initialized and trained again to complete functions such as device identification, rate negotiation, power management, reset, link width channel sequence negotiation, etc. This connection process is implemented based on the standard PCIE or OCP protocol. The component takeover process is the same as the information network card information identification process in general products, so it will not be repeated.

[0066] In an exemplary embodiment, the switching device 1 is used to initiate a second handshake operation between itself and the server mainboard after identifying the information of the network transmission device 21 connected to itself, so as to respond to the query instruction sent by the server mainboard and return the reply information corresponding to the query instruction;

[0067] The server motherboard is also used to configure the working parameters of the adapter device 1 based on the received reply information; the working parameters include the bandwidth of the first data transmission channel, the bandwidth of the second data transmission channel, the enable function parameters of the adapter device 1, the disable function parameters of the adapter device 1, and at least one of the memory mapping addresses corresponding to the adapter device 1.

[0068] In this embodiment, the secondary handshake process between the separate DPU and the server firmware and OS (Operating System) includes that after the separate DPU completes the identification of its own network card information, it will automatically trigger the secondary information identification with the server firmware. The DPU establishes communication with the server firmware through the PCIe interface. The server firmware sends a series of query instructions to the DPU to obtain detailed information about the DPU, such as model, firmware version, supported functions, etc. The DPU will respond to these query instructions and send corresponding reply information to the server firmware. The server firmware will verify whether the information sent by the DPU is consistent with expectations. If the verification is successful, the server firmware will perform corresponding configurations based on the information of the DPU, such as setting the bandwidth of the PCIe channel, enabling or disabling specific DPU functions, etc. At the same time, the server firmware will also update its internal database to record the existence and status of the DPU. In combination with the above, when the specifications of the network transmission unit 2 cannot meet the current needs, a higher-specification network transmission unit 2 can be replaced separately. First, the entire separate DPU is shut down and powered off through the CPLD of the local server. During this process, the network transmission unit 2 is first disconnected from the data processing unit 11, and then the data processing unit 11 is disconnected from the local server. The entire separate DPU can be removed from the local server, and then the network transmission unit 2 is unplugged from the separate DPU and replaced with a new network transmission unit 2. The entire separate DPU is then installed back on the server, and the steps of adding a new network card configuration are repeated. The network transmission unit 2 and the data processing unit 11 are then reinstalled and identified. After link initialization and link training are completed, data transmission and processing can be carried out normally.

[0069] In an exemplary embodiment, please refer to Figure 2 The data processing system also includes a storage unit 14, which includes a storage device arranged on the adapter device 1 and / or a hard disk connected to the server motherboard; the data processing unit 11 is also used to select a storage device and / or a hard disk to store result data according to the storage status of the storage unit 14.

[0070] The storage unit 14 in this embodiment includes storage devices such as a hard disk and an EEPROM (Electrically erasable programmable read only memory), which are used to store data processed by the computing management unit 13 and information such as the firmware files and addresses of the chip. It should be noted that in the present invention, in addition to the traditional DPU installing M.2 on the substrate as a storage device, the hard disk on the external server can also be used as an extended storage device to further expand the storage data space.

[0071] In an exemplary embodiment, the computing management unit 13 is further used to detect the operating status and operating information of the data processing system, where the operating status includes the current network transmission status.

[0072] It can be understood that the computing management unit 13 also includes a sensor, a JTAG (Joint Test Action Group) interface, and a USB (Universal Serial Bus) interface to monitor the operating status and operating information (voltage, current, temperature) of the entire DPU, thereby realizing out-of-band management, FW upgrade, debug and other operation and maintenance management of the separated DPU. Among them, the sensor is used to monitor and transmit key information such as temperature and voltage of the DPU; in addition, there may be BMC Switch (switching component), level conversion, MOS tube (Metal-Oxide-Semiconductor Field-Effect Transistor), VR (Voltage Regulator) and other electronic components on the substrate to ensure that the circuit design of the substrate meets the requirements.

[0073] On the basis of the above embodiment, the data received by the ordinary network card is transmitted to the Soc and CPLD on the substrate through the PCIE Slot. The data processing unit 11 composed of the Soc and CPLD processes the received data, and the processed data is stored in the M.2 hard disk controlled by the SoC. At the same time, the BMC on the substrate manages the working status and operation information of the adapter device 1 and the ordinary network card through the I2C bus. When it is necessary to refresh the FW of the BMC on the substrate, it can be refreshed through the UART (Universal Asynchronous Receiver / Transmitter) interface, and the new FW is stored in the EEPROM under the BMC. The server motherboard realizes out-of-band management and information identification of the DPU and other external PCIE devices through the SMBUS (System Management Bus) management bus on the substrate gold finger or the cable.

[0074] Furthermore, other PCIE devices transferred by the transfer device 1 do not interact with the baseboard for data, but are still processed by the mainboard CPU through the link from the PCIE Slot to the gold finger and then to the mainboard.

[0075] In an exemplary embodiment, the operation information includes the current substrate temperature; the adapter 1 also includes a heat dissipation unit; the calculation management unit 13 is also used to control the heat dissipation unit to operate according to the current substrate temperature so that the current substrate temperature of the adapter 1 is within the target temperature range.

[0076] In this embodiment, the heat dissipation unit includes but is not limited to an active or passive heat sink / heat sink, which is used to meet the heat dissipation requirements of the substrate chip.

[0077] Based on the above embodiment, the structural diagram of the switching device 1 is shown in FIG. Figure 3 and Figure 4 As shown, it includes but is not limited to a gold finger 16 , a memory slot 141 , a BMC 151 , a BIOS 131 , a SoC 132 , a CPLD 133 , an EEPROM chip 142 , a first external interface 17 , a second external interface 18 , and a third external interface 19 .

[0078] In summary, the separate DPU adapter device 1 provided by the present invention first decouples the current DPU architecture, decouples the DPU into an adapter device 1 integrating a network transmission unit 2 and a data processing unit 11, adopts an ordinary network card as the network transmission unit 2, integrates the data processing unit 11 of the DPU on the adapter card, forms an adapter device 1 of the data processing unit 11, and constructs a DPU by connecting the ordinary network card and the adapter device 1 of the data processing unit 11. The DPU can also be used as an adapter card to expand other PCIE devices, thereby realizing flexible changes and adaptations of the network transmission unit 2 of the DPU without affecting the PCIE scalability of the entire server.

[0079] In a second aspect, the present invention further provides a server, comprising a server mainboard and a data processing system connected to the server mainboard and as described in any one of the above embodiments.

[0080] The data processing system includes: a switching device integrated with a data processing unit, a computing management unit and at least one slot, a first data transmission channel is provided between the slot and the computing management unit, and a second data transmission channel is provided between the slot and the data processing unit. The computing management unit is used to determine the current network transmission status, generate a replacement instruction if the current network transmission status meets the replacement condition, and identify the network transmission device connected to the slot after responding to the replacement instruction through the first data transmission channel, and bind the network transmission device to the data processing unit; a network transmission unit including at least one network transmission device, the network transmission device is used to transmit network data between the switching device and the target network through the second data transmission channel after being bound to the data processing unit.

[0081] In an exemplary embodiment, the process of determining the current network transmission status includes: obtaining the current network rate of the network transmission device currently bound to the data processing unit; determining whether the current network rate is within the target network rate range corresponding to the current application scenario; if so, determining that the current network transmission status is a normal transmission status; if not, determining that the current network transmission status is an abnormal transmission status.

[0082] In an exemplary embodiment, the replacement condition includes a first replacement condition, and the switching device is further configured to determine that the current network transmission state satisfies the first replacement condition when the current network transmission state is an abnormal transmission state.

[0083] In an exemplary embodiment, the replacement instruction includes a first prompt instruction, and the replacement instruction is generated if the current network transmission status meets the replacement condition. The process of identifying the network transmission device connected to the slot in response to the replacement instruction through the first data transmission channel includes generating a first prompt instruction if the current network transmission status meets the first replacement condition, so that the user can unplug the network transmission device currently bound to the data processing unit from the slot based on the first prompt instruction, and insert a new network transmission device into the slot, so that the computing management unit can identify the new network transmission device through the first data transmission channel, and the network rate of the new network transmission device is within the target network rate range.

[0084] In an exemplary embodiment, the replacement instruction includes a first switching instruction. If the current network transmission status meets the replacement condition, the replacement instruction is generated, and the process of identifying the network transmission device connected to the slot after responding to the replacement instruction through the first data transmission channel includes: if the current network transmission status meets the first replacement condition, and the network transmission unit includes multiple network transmission devices, determine the first target transmission device whose network rate is within the target network rate range among the multiple network transmission devices, and generate a first switching instruction based on the first target transmission device. The first switching instruction is used to disconnect the electrical connection between the network transmission device currently bound to the data processing unit and its corresponding slot, and to connect the electrical connection between the first target transmission device and its corresponding slot, so that the computing management unit can identify the first target transmission device through the first data transmission channel.

[0085] In an exemplary embodiment, the process of determining the current network transmission status includes: obtaining the usage status of the second target transmission device; the second target transmission device is a network transmission device currently bound to the data processing unit; the usage status includes a normal state or a fault state; if the usage status is a normal state, determining that the current network transmission status is a normal transmission state; if the usage status is a fault state, determining that the current network transmission status is an abnormal transmission state.

[0086] In an exemplary embodiment, the replacement condition includes a second replacement condition, and the switching device is further configured to determine that the current network transmission state satisfies the second replacement condition when the current network transmission state is an abnormal transmission state.

[0087] In an exemplary embodiment, the replacement instruction includes a second prompt instruction. If the current network transmission status meets the replacement condition, the replacement instruction is generated, and the process of identifying the network transmission device connected to the slot in response to the replacement instruction through the first data transmission channel includes: if the current network transmission status meets the second replacement condition, a second prompt instruction is generated so that the user can unplug the network transmission device currently bound to the data processing unit from the slot based on the second prompt instruction, and insert a new network transmission device into the slot, so that the computing management unit can identify the new network transmission device through the first data transmission channel, and the usage status of the new network transmission device is normal.

[0088] In an exemplary embodiment, the replacement instruction includes a second switching instruction. If the current network transmission status meets the replacement condition, the replacement instruction is generated, and the process of identifying the network transmission device connected to the slot after responding to the replacement instruction through the first data transmission channel includes: if the current network transmission status meets the second replacement condition, and the network transmission unit includes multiple network transmission devices, a third target transmission device with a normal usage status is determined among the multiple network transmission devices, and a second switching instruction is generated based on the third target transmission device. The second switching instruction is used to disconnect the electrical connection between the network transmission device currently bound to the data processing unit and its corresponding slot, and to conduct the electrical connection between the third target transmission device and its corresponding slot, so that the computing management unit can identify the third target transmission device through the first data transmission channel.

[0089] In an exemplary embodiment, the network transmission device is a network transmission device whose firmware is customized and compiled based on an application scenario.

[0090] In an exemplary embodiment, the data processing unit is used to receive network data transmitted by the network transmission device bound to itself through a second data transmission channel, and process the network data according to a preset processing logic to obtain result data, and send the result data to the network transmission device bound to itself through the second data transmission channel.

[0091] In an exemplary embodiment, the switching device further includes a plurality of interfaces, and the plurality of interfaces are used to transmit a power supply signal and at least one communication signal between the switching device and the server mainboard.

[0092] In an exemplary embodiment, the server mainboard is used to perform a first handshake operation with the switching device after the server mainboard is powered on, so that a local processor of the server mainboard and a data processing unit in the switching device establish a communication connection.

[0093] In an exemplary embodiment, the adapter device is used to initiate a second handshake operation between itself and the server mainboard after identifying information of a network transmission device connected to itself, so as to respond to a query instruction sent by the server mainboard and return reply information corresponding to the query instruction; the server mainboard is also used to configure the working parameters of the adapter device based on the received reply information; the working parameters include at least one of the bandwidth of the first data transmission channel, the bandwidth of the second data transmission channel, the enable function parameters of the adapter device, the disable function parameters of the adapter device, and the memory mapping address corresponding to the adapter device.

[0094] In an exemplary embodiment, the data processing system also includes a storage unit, which includes a storage device disposed on the adapter device and / or a hard disk connected to the server motherboard; the data processing unit is also used to select the storage device and / or hard disk to store the result data according to the storage status of the storage unit.

[0095] In an exemplary embodiment, the computing management unit is further used to detect the operating status and operating information of the data processing system, where the operating status includes the current network transmission status.

[0096] In an exemplary embodiment, the operation information includes a current substrate temperature; the transfer device further includes a heat dissipation unit; the calculation management unit is further used to control the heat dissipation unit to operate according to the current substrate temperature so that the current substrate temperature of the transfer device is within a target temperature range.

[0097] Third, please refer to Figure 5 The present invention also provides a data processing method, which is applied to a data processing system as described in any one of the embodiments above, wherein the data processing system includes a network transmission unit, the network transmission unit includes at least one network transmission device, and the data processing system also includes a switching device integrated with a data processing unit, a computing management unit and at least one slot, a first data transmission channel is provided between the slot and the computing management unit, and a second data transmission channel is provided between the slot and the data processing unit. The data processing method includes: S101: determining the current network transmission status through the computing management unit, generating a replacement instruction if the current network transmission status meets the replacement condition, and identifying the network transmission device connected to the slot after responding to the replacement instruction through the first data transmission channel, and binding the network transmission device to the data processing unit; S102: transmitting the network data between the switching device and the target network through the second data transmission channel after the network transmission device is bound to the data processing unit.

[0098] For an introduction to a data processing method provided by the present invention, please refer to the above-mentioned embodiment, and the present invention will not be described in detail here.

[0099] A data processing method provided by the present invention has the same beneficial effects as the above data processing method.

[0100] In a fourth aspect, the present invention further provides a computer program product, comprising a computer program / instruction, which implements the steps of the above data processing method when executed by a processor.

[0101] For an introduction to a computer program product provided by the present invention, please refer to the above embodiments, and the present invention will not be described in detail here.

[0102] A computer program product provided by the present invention has the same beneficial effects as the above data processing method.

[0103] Fifth, please refer to Figure 6 The present invention further provides a non-volatile storage medium 30, on which a computer program 31 is stored. When the computer program 31 is executed by a processor, the steps of the above data processing method are implemented.

[0104] The non-volatile storage medium 30 may include: ROM (Read-only memory), PROM (Programmable read-only memory), EAROM (Electrically alterable read only memory), EPROM (Erasable programmable read only memory), EEPROM, Flash memory, and other media that can store program codes.

[0105] For an introduction to a non-volatile storage medium 30 provided by the present invention, please refer to the above embodiment, and the present invention will not be described in detail here.

[0106] The non-volatile storage medium 30 provided by the present invention has the same beneficial effects as the above data processing method.

[0107] It should also be noted that, in this specification, 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 "comprises", "comprising" 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, an element defined by the statement "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or device including the element.

[0108] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A data processing system, characterized in that: The data processing system is a separate DPU, wherein the separate DPU separates a data processing unit and a network transmission unit, and the data processing system comprises: A switching device integrated with the data processing unit, a computing management unit and at least one slot, wherein a first data transmission channel is provided between the slot and the computing management unit, and a second data transmission channel is provided between the slot and the data processing unit, wherein the computing management unit is used to determine a current network transmission state, generate a replacement instruction if the current network transmission state meets a replacement condition, and identify a network transmission device connected to the slot after the change in response to the replacement instruction through the first data transmission channel, and bind the network transmission device to the data processing unit; The network transmission unit includes at least one network transmission device, and the network transmission device is used to transmit network data between the switching device and the target network through the second data transmission channel after being bound to the data processing unit; the network transmission device and the slot are detachably connected.

2. The data processing system according to claim 1, characterized in that The process of determining the current network transmission status includes: Acquire the current network rate of the network transmission device currently bound to the data processing unit; Determine whether the current network rate is within the target network rate range corresponding to the current application scenario; If so, determining that the current network transmission state is a normal transmission state; If not, it is determined that the current network transmission state is an abnormal transmission state.

3. The data processing system according to claim 2, characterized in that: The replacement condition includes a first replacement condition, and the switching device is further configured to determine, when the current network transmission state is the abnormal transmission state, that the current network transmission state satisfies the first replacement condition.

4. The data processing system according to claim 3, characterized in that: The replacement instruction includes a first prompt instruction, and if the current network transmission state meets the replacement condition, a replacement instruction is generated, and a process of identifying, through the first data transmission channel, a network transmission device connected to the slot that has been changed in response to the replacement instruction includes: If the current network transmission status meets the first replacement condition, the first prompt instruction is generated so that the user can unplug the network transmission device currently bound to the data processing unit from the slot based on the first prompt instruction and insert a new network transmission device into the slot, so that the computing management unit can identify the new network transmission device through the first data transmission channel, and the network rate of the new network transmission device is within the target network rate range.

5. The data processing system according to claim 3, characterized in that: The replacement instruction includes a first switching instruction, and if the current network transmission state meets the replacement condition, the replacement instruction is generated, and the process of identifying the network transmission device connected to the slot after the change in response to the replacement instruction through the first data transmission channel includes: If the current network transmission status satisfies the first replacement condition and the network transmission unit includes multiple network transmission devices, a first target transmission device whose network rate is within the target network rate range is determined among the multiple network transmission devices, and the first switching instruction is generated based on the first target transmission device. The first switching instruction is used to disconnect the electrical connection between the network transmission device currently bound to the data processing unit and its corresponding slot, and to connect the electrical connection between the first target transmission device and its corresponding slot, so that the computing management unit can identify the first target transmission device through the first data transmission channel.

6. The data processing system according to claim 1, characterized in that: The process of determining the current network transmission status includes: Acquire the usage status of a second target transmission device; the second target transmission device is a network transmission device currently bound to the data processing unit; the usage status includes a normal state or a fault state; If the usage status is the normal status, determining that the current network transmission status is the normal transmission status; If the usage status is the fault status, it is determined that the current network transmission status is an abnormal transmission status.

7. The data processing system according to claim 6, characterized in that: The replacement condition includes a second replacement condition, and the switching device is further configured to determine, when the current network transmission state is the abnormal transmission state, that the current network transmission state satisfies the second replacement condition.

8. The data processing system according to claim 7, characterized in that: The replacement instruction includes a second prompt instruction, and if the current network transmission state meets the replacement condition, the replacement instruction is generated, and the process of identifying the network transmission device connected to the slot after the change in response to the replacement instruction through the first data transmission channel includes: If the current network transmission status meets the second replacement condition, the second prompt instruction is generated so that the user can unplug the network transmission device currently bound to the data processing unit from the slot based on the second prompt instruction and insert a new network transmission device into the slot so that the computing management unit can identify the new network transmission device through the first data transmission channel, and the usage status of the new network transmission device is normal.

9. The data processing system according to claim 7, characterized in that: The replacement instruction includes a second switching instruction, and if the current network transmission state satisfies the replacement condition, a replacement instruction is generated, and a process of identifying, through the first data transmission channel, a network transmission device connected to the slot that has been changed in response to the replacement instruction includes: If the current network transmission status satisfies the second replacement condition, and the network transmission unit includes multiple network transmission devices, determine the third target transmission device whose usage status is normal among the multiple network transmission devices, and generate the second switching instruction based on the third target transmission device. The second switching instruction is used to disconnect the electrical connection between the network transmission device currently bound to the data processing unit and its corresponding slot, and to conduct the electrical connection between the third target transmission device and its corresponding slot, so that the computing management unit can identify the third target transmission device through the first data transmission channel.

10. The data processing system according to claim 1, characterized in that: The network transmission device is a network transmission device whose firmware is customized and compiled based on the application scenario.

11. The data processing system according to any one of claims 1 to 10, characterized in that: The data processing unit is used to receive network data transmitted by the network transmission device bound to itself through the second data transmission channel, process the network data according to the preset processing logic, obtain result data, and send the result data to the network transmission device bound to itself through the second data transmission channel.

12. The data processing system according to claim 11, characterized in that: The adapter device further comprises a plurality of interfaces, and the plurality of interfaces are used to transmit a power supply signal and at least one communication signal between the adapter device and a server mainboard.

13. The data processing system according to claim 12, characterized in that: The server mainboard is used to perform a first handshake operation with the switching device after being powered on, so that a local processor of the server mainboard and a data processing unit in the switching device establish a communication connection.

14. The data processing system according to claim 13, characterized in that: The switching device is used to initiate a second handshake operation between itself and the server mainboard after identifying the information of the network transmission device connected to itself, so as to respond to the query instruction sent by the server mainboard and return reply information corresponding to the query instruction; The server mainboard is also used to configure the working parameters of the adapter device based on the received reply information; the working parameters include the bandwidth of the first data transmission channel, the bandwidth of the second data transmission channel, the enable function parameters of the adapter device, the disable function parameters of the adapter device, and at least one of the memory mapping address corresponding to the adapter device.

15. The data processing system according to claim 12, characterized in that: The data processing system further comprises a storage unit, wherein the storage unit comprises a storage device disposed on the adapter device and / or a hard disk connected to the server mainboard; The data processing unit is further configured to select the storage device and / or the hard disk to store the result data according to the storage status of the storage unit.

16. The data processing system according to claim 11, characterized in that: The computing management unit is also used to detect the operating status and operating information of the data processing system, and the operating status includes the current network transmission status.

17. The data processing system according to claim 16, characterized in that: The operation information includes the current substrate temperature; The switching device also includes a heat dissipation unit; The calculation management unit is further used to control the heat dissipation unit to operate according to the current substrate temperature, so that the current substrate temperature of the transfer device is within a target temperature range.

18. A server, characterized in that: A data processing system as claimed in any one of claims 1 to 17, comprising a server mainboard and connected to the server mainboard.

19. A data processing method, characterized in that: Applied to the data processing system according to any one of claims 1 to 17, the data processing system is a separate DPU, the separate DPU separates a data processing unit and a network transmission unit, the data processing system includes the network transmission unit, the network transmission unit includes at least one network transmission device, the data processing system also includes a switching device integrated with the data processing unit, a computing management unit and at least one slot, a first data transmission channel is provided between the slot and the computing management unit, a second data transmission channel is provided between the slot and the data processing unit, and the data processing method includes: Determine the current network transmission state through the computing management unit, generate a replacement instruction if the current network transmission state meets the replacement condition, and identify the network transmission device connected to the slot after the change in response to the replacement instruction through the first data transmission channel, and bind the network transmission device to the data processing unit; After the network transmission device is bound to the data processing unit, the network data between the switching device and the target network is transmitted through the second data transmission channel; the network transmission device and the slot are detachably connected.

20. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the data processing method according to claim 19 are implemented.

21. A non-volatile storage medium, characterized in that: The non-volatile storage medium stores a computer program, and when the computer program is executed by the processor, the steps of the data processing method according to claim 19 are implemented.

Images