Network card fault detection method and system, electronic equipment and storage medium

Abstract

The invention provides a network card fault detection method which can be applied to the field of financial science and technology. The network card fault detection method comprises the following steps: receiving a detection request sent by a computing power cluster; the computing power cluster is obtained by configuring a plurality of network cards; acquiring configuration information of the computing power cluster based on the detection request; generating a detection instruction based on the configuration information of the computing power cluster; and sending the detection instruction to the computing power cluster, so that the computing power cluster performs fault detection on each network card according to a detection task indicated by the detection instruction. The invention further provides a network card fault detection device, electronic equipment and a storage medium.

Description

Technical Field

[0001] This application relates to the field of financial technology, specifically to a method, apparatus, electronic device, and storage medium for detecting network interface card (NIC) faults. Background Technology

[0002] With the rapid growth in computing power demand from high-frequency trading, risk calculation, and robo-advisory services, the number of network interface cards (NICs) in large-scale computing clusters has increased dramatically. Traditional NIC fault detection methods mainly rely on the device's own monitoring mechanisms for fault reporting. This passive fault perception method suffers from delayed fault detection and low fault location efficiency, and cannot meet the requirements of financial services for high system availability and rapid fault recovery. Summary of the Invention

[0003] In view of the above problems, this application provides a network interface card (NIC) fault detection method, apparatus, device, medium, and program product.

[0004] According to the first aspect of this application, a network interface card (NIC) fault detection method is provided, comprising:

[0005] Receives probe requests sent by a computing power cluster; the computing power cluster is configured from multiple network interface cards (NICs).

[0006] Based on the probe request, obtain the configuration information of the computing power cluster;

[0007] Detection commands are generated based on the configuration information of the computing cluster;

[0008] The detection command is sent to the computing power cluster so that the computing power cluster can perform fault detection on each of the network cards according to the detection task indicated by the detection command.

[0009] According to an embodiment of this application, generating a probe command based on the configuration information of the computing power cluster includes:

[0010] The hierarchical structure of the computing power cluster is determined based on the computing power cluster configuration information;

[0011] According to the hierarchical structure, a detection task is assigned to at least one target network interface card (NIC) in the computing power cluster; the detection task is used to instruct the target NIC to perform fault detection on other NICs in the cluster.

[0012] Detection commands are generated based on the detection tasks of each target network interface card.

[0013] According to an embodiment of this application, the hierarchical structure is composed of a first preset number of computing power units; each computing power unit includes a second preset number of computing power nodes, and each computing power node includes a third preset number of network interface cards; the first preset number, the second preset number, and the third preset number may be the same or different.

[0014] According to an embodiment of this application, the step of assigning a probe task to at least one target network interface card in the computing power cluster according to the hierarchical structure includes:

[0015] When the first preset quantity, the second preset quantity, and the third preset quantity are the same, for each computing node in the same computing power unit, the i-th network card in the i-th computing power node is determined as the target network card, and a probe task within the unit is assigned to the target network card. The probe task within the unit is used to instruct the target network card to probe the i-th network card of all computing power nodes in the computing power unit, where i is a positive integer.

[0016] For each computing unit within the same computing power cluster, the network card of the j-th computing node in the j-th computing unit is identified as the target network card, and an intra-cluster probing task is assigned to the target network card. The intra-cluster probing task is used to instruct the target network card to probe the network card of the j-th computing node in all computing units within the computing power cluster, where j is a positive integer.

[0017] According to an embodiment of this application, the step of assigning a probe task to at least one target network interface card in the computing power cluster according to the hierarchical structure includes:

[0018] When the first preset quantity, the second preset quantity, and the third preset quantity are different, for the same computing power unit, when the number of computing power nodes in the computing power unit is less than the second preset quantity, the detection tasks corresponding to the portion exceeding the number of computing power nodes are allocated to other computing power nodes in the computing power unit in a preset order.

[0019] For each computing unit within the same computing power cluster, when the number of computing power units in the computing power cluster is less than the first preset number, the detection tasks corresponding to the portion exceeding the number of computing power units are allocated to other computing power units in the computing power cluster in a preset order.

[0020] According to an embodiment of this application, obtaining the configuration information of the computing power cluster based on the probe request includes:

[0021] The configuration information of the computing cluster indicated by the probe request is obtained from the first database.

[0022] According to embodiments of this application, the method further includes:

[0023] The system receives the detection results fed back by the computing power cluster and saves the detection results to the second database.

[0024] According to embodiments of this application, the method further includes:

[0025] The detection results are sent to a message queue cluster, so that the message queue cluster sends the detection results to a display terminal for display.

[0026] A second aspect of this application provides a network interface card (NIC) fault detection system, comprising:

[0027] A computing power cluster, which is configured from multiple network interface cards;

[0028] A control device is configured to receive a probe request sent by the computing power cluster; obtain configuration information of the computing power cluster based on the probe request; generate a probe command based on the configuration information of the computing power cluster; and send the probe command to the computing power cluster.

[0029] The computing power cluster is also used to perform fault detection on each of the network cards according to the detection task indicated by the detection command to obtain the detection results.

[0030] According to an embodiment of this application, the system further includes:

[0031] The first database is used to store the configuration information of the computing power cluster;

[0032] A second database is used to store the detection results.

[0033] According to an embodiment of this application, the system further includes:

[0034] Display terminal;

[0035] A message queue cluster is used to receive the detection results and send them to a display terminal for display.

[0036] A third aspect of this application provides an electronic device comprising: one or more processors; and a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the method described above.

[0037] A fourth aspect of this application also provides a computer-readable storage medium having a computer program or instructions stored thereon, which, when executed by a processor, implement the steps of the above-described method. Attached Figure Description

[0038] The above-mentioned contents, other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0039] Figure 1 This illustration schematically depicts an application scenario of a network interface card (NIC) fault detection method, system, device, and medium according to embodiments of this application.

[0040] Figure 2 The flowchart illustrates a network interface card (NIC) fault detection method according to an embodiment of this application.

[0041] Figure 3 This illustration schematically shows a computing power cluster hierarchy structure according to an embodiment of this application;

[0042] Figures 4A-4B This illustration schematically shows a probe task allocation diagram under a normalized architecture according to an embodiment of this application;

[0043] Figure 5 This illustration schematically shows a reassignment diagram of a probe task under a non-normalized architecture according to an embodiment of this application;

[0044] Figure 6 This schematically illustrates a structural block diagram of a network interface card (NIC) fault detection system according to an embodiment of this application;

[0045] Figure 7 A block diagram schematically illustrates an electronic device suitable for implementing a network interface card (NIC) fault detection method according to an embodiment of this application. Detailed Implementation

[0046] The embodiments of this application will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this application. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this application for ease of explanation. However, it will be apparent that one or more embodiments may be implemented without these specific details. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0047] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0048] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0049] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0050] Embodiments of this application provide a network interface card (NIC) fault detection method, including:

[0051] Receive probe requests sent by the computing power cluster; the computing power cluster is configured from multiple network interface cards;

[0052] Obtain the configuration information of the computing cluster based on the probe request;

[0053] Generate probe commands based on the configuration information of the computing cluster;

[0054] The detection command is sent to the computing power cluster so that the computing power cluster can perform fault detection on each network card according to the detection task indicated by the detection command.

[0055] By adopting the embodiments of this application, an active communication is established between the computing power cluster and the cluster by receiving a probe request, thus changing the traditional passive approach of relying on the device's own monitoring mechanism for fault reporting. By obtaining the configuration information of the computing power cluster based on the probe request, the complete topology and distribution of network interface cards (NICs) within the cluster can be grasped in real time, providing an accurate target location basis for subsequent fault detection. By generating probe commands based on the configuration information of the computing power cluster, the problem of incomplete detection coverage caused by the lack of a unified coordination mechanism in related technologies is avoided. By sending the probe commands to the computing power cluster, enabling it to perform fault detection on each NIC according to the probe tasks indicated by the commands, a unified and coordinated distributed detection mechanism is established, which can quickly locate the specific location and fault type of the faulty NIC when a fault occurs. Compared with the passive fault perception methods in related technologies, the embodiments of this application shorten the fault discovery time, improve the accuracy and efficiency of fault location, effectively solve the technical problems of delayed fault discovery and low location efficiency, and meet the requirements of financial services for high availability and rapid fault recovery of computing power clusters.

[0056] Figure 1 The illustration shows an application scenario of the network card fault detection method, system, device and medium according to embodiments of this application.

[0057] like Figure 1As shown, application scenario 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 serves as a medium for providing communication links between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired or wireless communication links, or fiber optic cables, etc.

[0058] Users can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 via the network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social media platform software, etc. (for example only).

[0059] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with displays and support web browsing, including but not limited to smartphones, tablets, laptops, and desktop computers.

[0060] Server 105 can be a server that provides various services, such as a backend management server that supports websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103 (this is just an example). The backend management server can analyze and process data such as received user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal devices.

[0061] It should be noted that the network card fault detection method provided in this application embodiment can generally be executed by server 105. Correspondingly, the network card fault detection system provided in this application embodiment can generally be set up in server 105. The network card fault detection method provided in this application embodiment can also be executed by a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105. Correspondingly, the network card fault detection system provided in this application embodiment can also be set up in a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105.

[0062] It should be understood that Figure 1The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0063] The following will be based on Figure 1 The following describes the network card fault detection method according to the embodiments of this application, based on the described scenario.

[0064] Figure 2 The flowchart illustrates a network interface card (NIC) fault detection method according to an embodiment of this application.

[0065] like Figure 2 As shown, the network card fault detection method of this embodiment includes operations S210 to S240, and the network card fault detection method can be executed by a control device.

[0066] Operate S210 to receive probe requests sent by the computing power cluster; the computing power cluster is configured from multiple network interface cards.

[0067] Operate S220 to obtain the configuration information of the computing cluster based on the probe request;

[0068] Operate S230 to generate probe commands based on the configuration information of the computing cluster;

[0069] Operate S240 to send probe commands to the computing cluster, so that the computing cluster can perform fault detection on each network card according to the probe tasks indicated by the probe commands.

[0070] In operation S210, a computing power cluster refers to a distributed computing system composed of multiple hardware devices with computing capabilities connected via a network. In this embodiment, it can be understood as a cluster system composed of multiple computing power servers equipped with network interface cards (NICs), used to execute large-scale parallel computing tasks and data processing tasks. Each computing power server in the computing power cluster is equipped with one or more NICs. These NICs are responsible for enabling data communication and information exchange between the computing power servers within the cluster. The normal operation of the NICs directly affects the communication stability and computing efficiency of the entire computing power cluster.

[0071] The detection request refers to a message sent by the computing server in the computing power cluster to the control terminal to request a fault detection instruction. In this embodiment, it can be understood as a data packet containing the computing server identification information and detection requirement information, which is used to trigger the control terminal to allocate a corresponding network card fault detection task to the computing server.

[0072] Specifically, the control device can receive probe requests sent by the computing power cluster through a preset API interface. Each computing power server in the computing power cluster actively sends probe requests to the API interface of the control terminal at preset time intervals (such as every minute). After receiving the probe request, the control terminal parses the computing power server identification information contained in the request to confirm the location and role of the computing power server in the computing power cluster.

[0073] In operation S220, configuration information refers to data information describing the internal structure of the computing power cluster and the distribution status of network interface card devices. In this embodiment, it can be understood as a data set containing key parameters such as the identification information of each computing power server in the computing power cluster, the number and distribution of network interface cards, and the network address of each network interface card device, which is used to generate targeted detection commands for the control device.

[0074] Specifically, after receiving a probe request from the computing power cluster, the control device queries and retrieves the configuration information of the corresponding computing power cluster from a pre-established configuration database based on the computing power server identification information contained in the probe request. The configuration database stores the complete topology of the computing power cluster, including the total number of computing power servers in the cluster, the network interface card (NIC) information configured on each computing power server, and the IP address allocation of the NICs. The control terminal parses the cluster identifier or server identifier in the probe request to locate the corresponding configuration record and extract the current configuration status of the computing power cluster, including which computing power servers are active, whether the NICs on each server are properly registered, and the connection relationships of the NICs, among other key configuration information.

[0075] In operation S230, the probe instruction refers to the set of instructions generated by the control device based on the computing power cluster configuration information to guide the network card devices in the cluster to perform fault detection. In this embodiment, it can be understood as a structured data packet containing information such as probe task allocation, probe target network card address, and probe execution rules, which is used to enable each network card device in the computing power cluster to perform fault detection operations according to a unified strategy and rules.

[0076] Specifically, after acquiring the configuration information of the computing cluster, the control device first parses the configuration information to identify all active computing servers and their corresponding network interface cards (NICs) within the cluster. Based on the current scale of the computing cluster and the distribution of NICs, the control device determines the range of NICs requiring fault detection. Based on a preset detection strategy, the control device assigns corresponding detection tasks to the computing servers sending detection requests, including parameters such as the list of target NICs to be detected, the detection method (e.g., connectivity test), and the detection frequency. The control device encapsulates this detection task information into a standardized detection command format. The detection command contains key parameters such as task identifier, target NIC network address, detection type, and execution time, ensuring that the computing servers can accurately understand and execute the corresponding fault detection operations.

[0077] Operate S240 to send probe commands to the computing cluster, so that the computing cluster can perform fault detection on each network card according to the probe tasks indicated by the probe commands.

[0078] Specifically, after generating a probe command, the control device sends the command to the corresponding computing server through a pre-established communication channel. Based on the source information of the probe request, the control device identifies the network address of the computing server that sent the request and returns the probe command to that server via an API response. Upon receiving the probe command, the computing server parses the probe task information contained within it, including the list of target network cards to be probed, the probe method, and execution parameters.

[0079] Following the probe instructions, the computing server initiates its local fault detection program, performing fault detection operations on each network interface card (NIC) within the computing cluster according to the specified probe tasks. The detection operations include sending connectivity test packets to the target NIC and detecting the NIC's response status and communication quality. During the execution of the probe tasks, the computing server collects real-time probe results data for each NIC, including key indicators such as connectivity status, response time, and packet loss rate. Through this distributed probe execution mechanism, the computing cluster can achieve comprehensive fault detection coverage of its internal NIC devices, enabling timely detection and location of faulty NIC devices.

[0080] By adopting the embodiments of this application, an active communication is established between the computing power cluster and the cluster by receiving a probe request, thus changing the traditional passive approach of relying on the device's own monitoring mechanism for fault reporting. By obtaining the configuration information of the computing power cluster based on the probe request, the complete topology and distribution of network interface cards (NICs) within the cluster can be grasped in real time, providing an accurate target location basis for subsequent fault detection. By generating probe commands based on the configuration information of the computing power cluster, the problem of incomplete detection coverage caused by the lack of a unified coordination mechanism in related technologies is avoided. By sending the probe commands to the computing power cluster, enabling it to perform fault detection on each NIC according to the probe tasks indicated by the commands, a unified and coordinated distributed detection mechanism is established, which can quickly locate the specific location and fault type of the faulty NIC when a fault occurs. Compared with the passive fault perception methods in related technologies, the embodiments of this application shorten the fault discovery time, improve the accuracy and efficiency of fault location, effectively solve the technical problems of delayed fault discovery and low location efficiency, and meet the requirements of financial services for high availability and rapid fault recovery of computing power clusters.

[0081] In actual computing cluster deployments, the number of network interface cards (NICs) is large and their distribution is complex. If a full network scan is used for fault detection, it will result in excessive detection traffic and low detection efficiency. At the same time, the lack of an effective task allocation mechanism can easily lead to detection blind spots or repeated detection.

[0082] To address the aforementioned problems, based on the above embodiments, as an optional embodiment, operation S230 may further include the following operations:

[0083] Operate S310 to determine the hierarchical structure of the computing power cluster based on the computing power cluster configuration information;

[0084] Operate S320 to assign a probe task to at least one target network card in the computing power cluster according to the hierarchical structure; the probe task is used to instruct the target network card to perform fault detection on other network cards in the cluster.

[0085] Operate the S330 to generate detection commands based on the detection tasks of each target network card.

[0086] In operation S310, the hierarchical structure refers to the hierarchical management architecture established according to the physical distribution and logical relationship of network interface card devices in the computing power cluster. In this embodiment, it can be understood as dividing the computing power cluster into multiple levels according to preset organizational rules. Each level contains a different number of management units to realize the structured management and orderly detection of large-scale network interface card devices.

[0087] Specifically, after obtaining the configuration information of the computing cluster, the control device first parses the configuration information, including the number of computing servers, the network interface card (NIC) configuration of each server, and the distribution status of the NICs. Based on preset hierarchical division rules, the control device groups and categorizes the NICs in the computing cluster according to their respective computing servers and physical locations. Based on the scale of the computing cluster and the total number of NICs, the control device determines a suitable hierarchical structure and establishes a tree-like structure with multiple management levels, where each level is responsible for managing NICs within a specific range.

[0088] In operation S320, the target network interface card (NIC) refers to the NIC device selected in the hierarchical structure to undertake the fault detection task. In this embodiment, it can be understood as a NIC device with detection execution capability and located in a specific position in the hierarchical structure, used to perform connectivity detection and status monitoring on other NIC devices within a specified range according to the assigned detection task. The detection task refers to the specific detection work content formulated for the target NIC. In this embodiment, it can be understood as a task instruction containing parameters such as the detection target range, detection execution method, and detection frequency, used to guide the target NIC to perform fault detection operations on other NICs in the cluster in an orderly manner.

[0089] Specifically, based on the established hierarchical structure, the control device analyzes the distribution and interrelationships of network interface cards (NICs) within each level, and selects suitable NICs as target NICs in each level according to a preset task allocation strategy. The control device assigns a corresponding detection responsibility range to each target NIC, ensuring that each NIC is included in the detection range of at least one target NIC, while avoiding duplicate allocation of detection tasks.

[0090] In operating the S330, after assigning detection tasks, the control device organizes and encapsulates the detection task information for each target network interface card (NIC). Based on the detection responsibility scope and task type of different target NICs, it generates corresponding detection instructions. The control device creates a standardized instruction format for each target NIC's detection task, containing key information such as the target NIC's identification information, a list of NICs to be detected, detection method parameters, and execution time requirements. According to the coordination relationship between the target NICs in the hierarchical structure, the control device adds necessary synchronization and coordination information to the detection instructions to ensure that different target NICs avoid conflicts and interference when executing detection tasks. The control device then groups and organizes the generated detection instructions according to the target NIC's affiliation, forming a complete set of detection instructions.

[0091] By employing the embodiments of this application, a hierarchical structure is determined based on the computing cluster configuration information, organizing the originally complex and disordered large-scale network interface card (NIC) devices into a layered architecture with clear management boundaries, providing a structured management foundation for subsequent probing task allocation. Probing tasks are allocated to target NICs according to the hierarchical structure, achieving a reasonable division of labor in the probing work and avoiding the problems of incomplete probing coverage or duplicate probing in related technologies. Standardized probing instructions are generated based on the probing tasks of each target NIC, ensuring unified coordination and orderly execution of the probing work, and improving the efficiency and accuracy of NIC fault detection throughout the computing cluster.

[0092] Based on the above embodiments, as an optional embodiment, the hierarchical structure is composed of a first preset number of computing power units; each computing power unit includes a second preset number of computing power nodes, and each computing power node includes a third preset number of network cards; the first preset number, the second preset number, and the third preset number may be the same or different.

[0093] In this context, a computing power unit refers to a basic management unit that undertakes specific computing tasks in the computing power cluster hierarchy. In the embodiments of this application, it can be understood as a logical group composed of multiple computing power nodes, used to realize regional computing resource management and fault isolation in the computing power cluster.

[0094] A computing node refers to a hardware device with independent computing capabilities within a computing unit. In this embodiment, it can be understood as a computing server configured with multiple network interface cards (NICs) to perform specific computing tasks and network communication functions. The first preset quantity refers to the total number of computing units in the computing cluster, the second preset quantity refers to the number of computing nodes contained in each computing unit, and the third preset quantity refers to the number of NICs configured on each computing node. These three preset quantity parameters can be configured independently according to actual deployment needs, providing flexible architecture adaptation capabilities for computing clusters of different sizes and types.

[0095] Figure 3 The illustration shows a schematic diagram of a computing power cluster hierarchy provided according to an embodiment of this application.

[0096] like Figure 3 As shown, the computing power cluster adopts a three-layer architecture design. The top layer is the computing power cluster layer, which contains multiple computing power units, Group1, Group2 to GroupN. Each computing power unit contains multiple computing power nodes. Taking Group1 as an example, it contains computing power nodes such as Unit1, Unit2 to Unit8. Each computing power node is configured with multiple network interface cards (NICs). Taking the Npu1 node as an example, it contains 8 NICs.

[0097] Specifically, after acquiring the computing cluster configuration information, the control device first identifies the overall device scale and distribution contained in the configuration information. Based on the actual deployment architecture of the computing cluster, it determines the specific values ​​of the first, second, and third preset quantities. Based on these three preset quantity parameters, the control device constructs a complete hierarchical structure model of the computing cluster, hierarchically classifying and arranging all network interface cards (NICs) according to their respective computing nodes and computing units. According to the inclusion relationships between levels in the hierarchical structure, the control device assigns a unique hierarchical identifier to each NIC. This identifier includes the computing unit number, computing node number, and NIC number within the node, forming a complete device location system.

[0098] By adopting the embodiments of this application, a hierarchical structure built based on a first preset quantity, a second preset quantity, and a third preset quantity provides a unified organizational architecture model for computing power clusters of different sizes and configurations, solving the problem of poor adaptability of fixed architecture models. By allowing configurations where the three preset quantity parameters are the same or different, the personalized deployment needs of computing power clusters in different application scenarios are met, improving the versatility of the hierarchical structure design.

[0099] Based on the above embodiments, as an optional embodiment, when the first preset quantity, the second preset quantity, and the third preset quantity are the same, operation S320 may further include the following operations:

[0100] Operation S410: For each computing node in the same computing power unit, the i-th network card in the i-th computing power node is identified as the target network card, and a probe task within the unit is assigned to the target network card. The probe task within the unit is used to instruct the target network card to probe the i-th network card of all computing power nodes in the computing power unit, where i is a positive integer.

[0101] Operation S420 identifies the network card of the j-th computing node in the j-th computing unit as the target network card for each computing unit within the same computing power cluster, and assigns an intra-cluster probing task to the target network card. The intra-cluster probing task is used to instruct the target network card to probe the network card of the j-th computing node in all computing units within the computing power cluster, where j is a positive integer.

[0102] In operation S410, the intra-unit detection task refers to the network card fault detection work performed within the computing power unit. In this embodiment, it can be understood as a detection task in which a target network card at a specific location within the computing power unit is responsible for performing connectivity detection and status monitoring on network card devices with the same location number on all computing power nodes within the same computing power unit, in order to achieve full coverage fault detection of network card devices within the computing power unit.

[0103] Specifically, after determining that the computing power cluster adopts a regularized architecture with the same first, second, and third preset quantities, the control device utilizes this regularity to design a symmetrical detection task allocation strategy. For each computing unit in the computing power cluster, the control device selects the target network interface card (NIC) according to the correspondence between NIC location numbers. Specifically, within each computing unit, the i-th NIC in the i-th computing node is selected as the target NIC for the intra-unit detection task, where i ranges from 1 to the second preset quantity. The control device assigns a corresponding intra-unit detection task to each selected target NIC. The task includes detecting the connectivity of the i-th NIC on all computing nodes within the computing unit, ensuring that all NICs with the same location number within the same computing unit are included in the detection range. Based on the distribution of the target NICs and the execution requirements of the detection task, the control device sets appropriate detection frequency and detection method parameters in the intra-unit detection task, enabling the target NICs to perform periodic fault detection operations on NICs within a specified range according to a unified standard.

[0104] In operation S420, the cluster-wide probing task refers to the network card fault detection work performed across computing power units. In this embodiment, it can be understood as the target network card at a specific location within the computing power cluster being responsible for the connectivity detection and status monitoring of network card devices on computing power nodes at the same location in all computing power units within the entire computing power cluster. This is used to realize cross-unit fault detection of network card devices within the global scope of the computing power cluster.

[0105] Specifically, based on the allocation of detection tasks within a unit, the control device further designs a cluster-wide detection task allocation strategy, fully utilizing the symmetry of the regularized architecture to achieve cross-computing unit detection coverage. For each computing unit in the computing power cluster, the control device selects the target network interface card (NIC) according to the correspondence between the computing unit and the computing node location number. Specifically, it selects the NIC on the j-th computing node in the j-th computing unit as the target NIC for performing the cluster-wide detection task, where j ranges from 1 to a first preset number. The control device assigns a corresponding cluster-wide detection task to each selected target NIC. The task includes connectivity detection of the NIC on the j-th computing node in all computing units within the computing power cluster, achieving cross-computing unit NIC fault detection coverage.

[0106] Figures 4A-4B The illustration shows a schematic diagram of the probe task allocation under a normalized architecture according to an embodiment of this application.

[0107] like Figure 4AAs shown, in the Unit level, each computing unit contains 8 computing nodes (Npu1 to Npu8), and each computing node is configured with 8 network interface cards (NICs). According to the allocation rules of the probing tasks within the unit, the first NIC in the first computing node Npu1 is selected as the target NIC, responsible for probing the first NICs of all computing nodes within the computing unit, forming a probing relationship distributed in a diagonal pattern.

[0108] Similarly, the second network interface card (NIC) in the second computing node Npu2 is responsible for detecting the second NIC of all computing nodes, and so on until the eighth NIC in the eighth computing node Npu8 is responsible for detecting the eighth NIC of all computing nodes.

[0109] The aforementioned symmetrical allocation method ensures that each network interface card (NIC) device within the computing unit is included in the detection range of the corresponding target NIC, while achieving a balanced distribution of detection tasks among different computing nodes.

[0110] like Figure 4B As shown, the computing power cluster Group level contains multiple computing power units (Unit1 to Unit8). According to the task allocation rules within the cluster, the network interface card (NIC) of the first computing power node in the first computing power unit (Unit1) is selected as the target NIC, responsible for probing the NIC devices of the first computing power node in all computing power units within the cluster. The NIC of the second computing power node in the second computing power unit (Unit2) is responsible for probing the NIC devices of the second computing power node in all computing power units, and so on, forming a cross-computing power unit probing coverage network.

[0111] By adopting the embodiments of this application, and utilizing the regularized architecture feature where the first, second, and third preset quantities are all the same, a symmetrical detection task allocation strategy is designed, achieving a balanced distribution of detection load among various network interface card (NIC) devices. Through the layered design of intra-unit and intra-cluster detection tasks, a dual detection mechanism covering both within and across computing units is constructed, ensuring comprehensive monitoring coverage of all NIC devices and eliminating the existence of detection blind spots.

[0112] Based on the above embodiments, as an optional embodiment, when the first preset quantity, the second preset quantity, and the third preset quantity are different, operation S320 may further include the following operations:

[0113] Operation S510: For the same computing power unit, when the number of computing power nodes in the computing power unit is less than the second preset number, the detection tasks corresponding to the portion exceeding the number of computing power nodes are assigned to other computing power nodes in the computing power unit in a preset order.

[0114] Operation S520: For each computing unit within the same computing power cluster, when the number of computing power units in the computing power cluster is less than a first preset number, the detection tasks corresponding to the portion exceeding the number of computing power units are allocated to other computing power units in the computing power cluster in a preset order.

[0115] In operation S510, the portion exceeding the number of computing power nodes refers to the portion of probing tasks that should have been undertaken by the missing computing power nodes but now need to be reallocated when the actual number of computing power nodes within the computing power unit is less than the second preset number. In this embodiment, this can be understood as a gap in probing tasks caused by insufficient computing power node configuration, which needs to be addressed through a reallocation mechanism to ensure the integrity of the probing coverage. The preset order refers to the rule sequence used to guide the reallocation of probing tasks. In this embodiment, this can be understood as the allocation order determined according to the computing power node number or other predefined priority rules, used to ensure the orderliness and predictability of probing task allocation.

[0116] Specifically, when the control device detects that the actual number of computing nodes within a computing unit is less than a second preset number, it first identifies the location of the missing computing nodes and the corresponding probing tasks. Based on a preset probing task allocation strategy, the control device determines the probing tasks that should have been undertaken by these missing computing nodes, including intra-unit probing tasks and potentially cross-unit probing tasks. Following a preset order rule, the control device reassigns these probing tasks, which cannot be executed due to missing computing nodes, to other available computing nodes within the computing unit, ensuring that each probing task has a corresponding execution entity.

[0117] When redistributing detection tasks, the control device considers the current detection load of each computing node and adopts a load balancing principle to allocate tasks, avoiding any computing node from bearing an excessive number of detection tasks and affecting its normal operating performance. The control device updates the detection task allocation record and saves the redistributed detection task information to the configuration database to ensure that subsequent detection command generation can reflect the latest task allocation status.

[0118] When operating the S520, the control device fully considers the impact of network topology and inter-unit communication delay when redistributing detection tasks at the computing unit level. It prioritizes assigning detection tasks to computing units that are closer to the network or have better communication quality, so as to improve the efficiency and accuracy of detection task execution.

[0119] Figure 5 This illustration schematically shows a reassignment diagram of a probe task under a non-normalized architecture according to an embodiment of this application.

[0120] like Figure 5As shown, in the actual deployment of the computing power cluster, an irregular configuration occurred. A certain computing unit was configured with only 6 computing nodes instead of the standard 8, resulting in the absence of the 7th and 8th computing nodes. According to the redistribution strategy of operation S510, the intra-unit probing tasks that should have been handled by the missing 7th computing node were redistributed to the 1st computing node in a preset order, while the probing tasks that should have been handled by the missing 8th computing node were assigned to the 2nd computing node. Similarly, at the computing power cluster level, if some computing unit positions are missing, the corresponding intra-cluster probing tasks will be redistributed to other available computing units according to a similar redistribution strategy, ensuring the integrity and continuity of the entire probing coverage network.

[0121] By adopting the embodiments of this application, a flexible detection task reallocation mechanism is established when facing irregular computing power cluster configurations. By reallocating detection tasks according to a preset order, the orderliness and predictability of task allocation are ensured, avoiding the load imbalance problem that may result from arbitrary allocation. By establishing reallocation strategies both within the computing power unit and at the computing power cluster level, adaptive handling of configuration changes at different levels is achieved, improving the fault detection system's adaptability to changes in the actual deployment environment.

[0122] Based on the above embodiments, as an optional embodiment, operation S220 may further include the following operations:

[0123] Retrieve the configuration information of the computing cluster indicated by the probe request from the first database.

[0124] The first database refers to a data storage system specifically used to store and manage the configuration information of the computing power cluster. In this embodiment, it can be understood as a structured database system with high availability, high concurrency access capability and data consistency guarantee, used to centrally manage the configuration parameters, topology relationships and status information of all computing power units, computing power nodes and network card devices in the computing power cluster.

[0125] Specifically, after receiving a probe request from the computing power cluster, the control device first parses the computing power cluster identification information and computing power server identification information contained in the probe request. This identification information is used to locate the specific source of the probe request and the range of the computing power cluster to which it belongs. Based on the parsed identification information, the control device constructs a corresponding database query statement and initiates a configuration information query request to the first database through a pre-established database connection interface. After receiving the query request, the first database retrieves the complete configuration information of the corresponding computing power cluster from the stored configuration data table according to the provided computing power cluster identification information. This includes key configuration data such as the hierarchical structure parameters of the computing power cluster, the composition of each computing power unit, the device configuration of each computing power node, and the network address allocation of the network interface card.

[0126] In one feasible implementation, the first database establishes a comprehensive data update mechanism. When the configuration of the computing power cluster changes, such as adding or removing computing power nodes or replacing network card devices due to failure, the relevant configuration change information will be updated to the first database in a timely manner to ensure that the control device can always obtain the latest configuration status.

[0127] By adopting the embodiments of this application, a centralized configuration information management mechanism based on a first database is established. By uniformly storing configuration information in a dedicated database system, standardized management and efficient access to configuration data are achieved, improving the response speed and accuracy of control devices in obtaining configuration information.

[0128] Based on the above embodiments, as an optional embodiment, the above method may further include the following operations:

[0129] Receive the detection results from the computing cluster and save the detection results to the second database.

[0130] The detection results refer to the detection data and status information generated by each target network interface card (NIC) in the computing power cluster after performing fault detection tasks. In this embodiment, it can be understood as a structured data set containing key information such as NIC connectivity status, response time, packet loss rate, detection timestamp, and fault type, used to reflect the real-time operating status and fault conditions of the NIC devices within the computing power cluster. The second database refers to a data storage system specifically used to store and manage the detection results data. In this embodiment, it can be understood as a database system with large-capacity storage capacity, high write performance, and time-series data processing characteristics, used to store the detection results data for a long time and support efficient data query and analysis operations.

[0131] Specifically, the control device continuously monitors the probe result feedback from the computing power cluster through a data receiving interface. After completing their assigned probe tasks, each computing server in the cluster sends the probe result data to the control device according to a predetermined data format and transmission protocol. Upon receiving the probe result data, the control device preprocesses it, including data format standardization, outlier identification and handling, and data completion, ensuring that the data stored in the second database has good quality and consistency. Based on the time-series characteristics and query requirements of the probe result data, the control device designs a suitable data storage structure and indexing strategy, and writes the preprocessed probe result data into the second database in batches, while assigning unique identifiers and timestamp information to each data record.

[0132] The second database adopts a time-series database architecture, specifically optimized for the time-series characteristics of the probe results data. It supports high-frequency data write operations and fast query functions based on time ranges. The second database implements data partitioning and compression mechanisms, automatically partitioning and storing data according to the time distribution of the probe results data, and compressing historical data to save storage space while maintaining efficient data access. The second database establishes a comprehensive data backup and recovery mechanism, regularly backing up the probe results data to ensure rapid recovery of probe data in the event of system failure or data corruption, guaranteeing data security and reliability.

[0133] By adopting the embodiments of this application, the detection results are uniformly stored in a dedicated second database, achieving persistent storage and standardized management of the detection data, providing a reliable data foundation for fault analysis, performance monitoring, and system optimization. The time-series data processing capabilities and high-performance storage architecture of the second database ensure efficient storage and rapid retrieval of large-scale detection result data, meeting the dual needs of real-time monitoring and historical analysis of the computing cluster.

[0134] Based on the above embodiments, as an optional embodiment, the above method may further include the following operations:

[0135] The detection results are sent to a message queue cluster, which then sends the results to a display terminal for presentation.

[0136] A message queue cluster refers to a distributed message passing system composed of multiple message queue nodes. In this embodiment, it can be understood as a message middleware cluster with high throughput, low latency, and high availability, used to realize asynchronous transmission and multi-point distribution of probe result data. A display terminal refers to a user interface device used to display the network card fault detection results of the computing power cluster. In this embodiment, it can be understood as including various types of display devices such as monitoring screens, operation and maintenance management platforms, and mobile terminals, used to provide operation and maintenance personnel with intuitive network card status monitoring and fault information display.

[0137] Specifically, while saving the detection results to the second database, the control device simultaneously sends the detection result data to the message queue cluster, implementing a dual-path processing mechanism for the detection results. Based on the content characteristics and urgency of the detection result data, the control device sends the detection results to the corresponding queues in the message queue cluster according to different topics or categories, such as normal state detection results, fault alarm information, and performance anomaly data, facilitating subsequent classification and targeted distribution. When sending detection results to the message queue cluster, the control device adopts an asynchronous sending mode to avoid blocking the main detection process, and sets an appropriate message persistence strategy to ensure that important fault information is not lost due to system anomalies. The control device adds necessary metadata information to the detection result data sent to the message queue cluster, including data priority, expiration time, and target display terminal type.

[0138] After receiving the probe result data from the control device, the message queue cluster pushes the data to the corresponding display terminals according to pre-configured distribution rules and the subscription relationships of the display terminals. The message queue cluster implements multiple message passing modes, including point-to-point transmission, publish-subscribe, and broadcast modes, to meet the differentiated needs of various display terminals for probe result data. The message queue cluster has load balancing and failover capabilities; when a queue node fails, it can automatically transfer message passing tasks to other healthy nodes, ensuring the continuity and reliability of probe result data transmission.

[0139] The display terminal receives the latest detection results data in real time by subscribing to the corresponding topics in the message queue cluster, and then visualizes this data in various forms such as charts, lists, and alarms.

[0140] By adopting the embodiments of this application, a real-time data distribution mechanism based on a message queue cluster is established. Through the asynchronous transmission and multi-point distribution capabilities of the message queue cluster, the real-time push and concurrent display of detection result data are realized, improving the timeliness of fault information transmission and the efficiency of operation and maintenance response.

[0141] Based on the above-mentioned network interface card (NIC) fault detection method, this application also provides a NIC fault detection system. The following will combine... Figure 6 The device is described in detail.

[0142] Figure 6 A schematic block diagram of a network interface card (NIC) fault detection system according to an embodiment of this application is shown.

[0143] In one feasible implementation, the network card fault detection system 600 of this embodiment includes a computing cluster 610 and a control device 620.

[0144] The computing power cluster 610 is configured from multiple network interface cards;

[0145] Control device 620 is used to receive a probe request sent by computing power cluster 610; obtain configuration information of computing power cluster 610 based on the probe request; generate a probe command based on the configuration information of computing power cluster 610; and send the probe command to computing power cluster 610.

[0146] The computing cluster 610 is also used to perform fault detection on each network card according to the detection task indicated by the detection command and obtain the detection results.

[0147] Based on the above embodiments, as a feasible implementation method, the network card fault detection system 600 further includes:

[0148] The first database 630 is used to store the configuration information of the computing power cluster 610;

[0149] The second database 640 is used to store the detection results.

[0150] Based on the above embodiments, as a feasible implementation method, the network card fault detection system 600 further includes:

[0151] Display terminal 660;

[0152] The message queue cluster 650 is used to receive the detection results and send the detection results to the display terminal 660 for display.

[0153] Based on the above embodiments, as a feasible implementation method, the computing power cluster 610 consists of multiple computing power servers configured with network interface cards (NICs) for performing large-scale parallel computing tasks. The computing power servers in the computing power cluster 610 communicate with each other through NICs. Each computing power server is configured with a management IP and a NIC IP to enable data transmission and fault detection within the cluster.

[0154] Based on the above embodiments, as a feasible implementation method, the first database 630 is used to store the configuration information of the computing power cluster 610, including the hierarchical structure parameters of the computing power cluster, the composition of each computing power unit, the device configuration of each computing power node, the network address allocation of the network card device, and other key configuration data. The first database 630 adopts a distributed database architecture, which has high availability and strong consistency characteristics, ensuring the accuracy and real-time performance of the configuration information.

[0155] Based on the above embodiments, as a feasible implementation method, the control device 620, as the core control node of the entire fault detection system, is used to receive detection requests sent by the computing power cluster 610; obtain the configuration information of the computing power cluster 610 from the first database 630 based on the detection requests; generate detection instructions based on the configuration information of the computing power cluster 610; and send the detection instructions to the computing power cluster 610 so that the computing power cluster 610 performs fault detection on each network card according to the detection tasks indicated by the detection instructions. The control device 620 is also responsible for providing API interface services, responding to the detection requests of the computing power servers, and allocating corresponding detection tasks to each computing power server according to the ping_mesh design rules.

[0156] Based on the above embodiments, as a feasible implementation method, the second database 640 is used to store the detection results generated after the computing cluster 610 performs fault detection. The second database 640 adopts a time-series database architecture, which is specifically optimized for the time-series characteristics of the detection result data, supporting high-frequency data write operations and fast query functions based on time ranges. The detection results stored in the second database 640 include key information such as network card connectivity status, response time, packet loss rate, detection timestamp, and fault type.

[0157] Based on the above embodiments, as a feasible implementation method, the message queue cluster 650 is used to receive the detection results sent by the control device 620 and distribute the detection results to the display terminal 660 in real time. The message queue cluster 650 adopts a distributed message passing architecture, which has the characteristics of high throughput, low latency, and high availability, and supports multiple message passing modes, including point-to-point transmission, publish-subscribe mode, etc., to ensure the continuity and reliability of detection result data transmission.

[0158] Based on the above embodiments, as a feasible implementation method, the display terminal 660 is used to receive probe result data from the message queue cluster 650 and visualize this data in various forms such as charts, lists, and alarms. The display terminal 660 provides maintenance personnel with an intuitive interface for monitoring network card status and displaying fault information, supporting functions such as real-time monitoring, historical data query, and fault alarms.

[0159] Based on the above embodiments, as a feasible implementation method, the data flow process of the network card fault detection system 600 is as follows:

[0160] Configuration information management process: Each computing server in the computing power cluster 610 writes its own server IP information to the second database 640 every 5 minutes. When the number of computing server nodes changes, the control device 620 executes the corresponding script, names the newly generated IP information according to the rules, and synchronously sends it to the first database 630. The display terminal 660 pulls data from the first database 630 every 5 minutes, and dynamically updates the front-end page after detecting a change in the number of computing power nodes.

[0161] Network interface card (NIC) parameter transmission process: Every minute, each computing server in the computing power cluster 610 writes its own RoCE NIC parameter information to the second database 640. Every minute, the control device 620 reads the RoCE NIC parameter information from the second database 640 and sends it to the message queue cluster 650 according to the agreed-upon format. Every minute, the display terminal 660 reads the corresponding RoCE NIC parameter indicators from the message queue cluster 650.

[0162] Fault Detection Process: Every minute, each computing server in the computing power cluster 610 sends a probe request to the API interface of the control device 620. The control device 620 returns whether the computing server status is normal and the IP address of the RoCE network card of the computing server it needs to probe. The computing server performs a ping_mesh operation and writes the computing server status information and ping_mesh results to the second database 640 every minute, while also sending the results to the message queue cluster 650. The display terminal 660 reads the corresponding computing server status indicators and RoCE network card ping_mesh indicators from the message queue cluster 650 every minute.

[0163] Figure 7 A block diagram schematically illustrates an electronic device suitable for implementing a network interface card (NIC) fault detection method according to an embodiment of this application.

[0164] like Figure 7 As shown, an electronic device 700 according to an embodiment of this application includes a processor 701, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 702 or a program loaded from a storage portion 708 into a random access memory (RAM) 703. The processor 701 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 701 may also include onboard memory for caching purposes. The processor 701 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of this application.

[0165] RAM 703 stores various programs and data required for the operation of electronic device 700. Processor 701, ROM 702, and RAM 703 are interconnected via bus 704. Processor 701 executes various operations of the method flow according to embodiments of this application by executing programs in ROM 702 and / or RAM 703. It should be noted that the programs may also be stored in one or more memories other than ROM 702 and RAM 703. Processor 701 may also execute various operations of the method flow according to embodiments of this application by executing programs stored in said one or more memories.

[0166] According to embodiments of this application, the electronic device 700 may further include an input / output (I / O) interface 705, which is also connected to a bus 704. The electronic device 700 may also include one or more of the following components connected to the input / output (I / O) interface 705: an input section 706 including a keyboard, mouse, etc.; an output section 707 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 708 including a hard disk, etc.; and a communication section 709 including a network interface card such as a LAN card, modem, etc. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the input / output (I / O) interface 705 as needed. A removable medium 711, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 710 as needed so that computer programs read from it can be installed into the storage section 708 as needed.

[0167] This application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.

[0168] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium, such as including but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to embodiments of this application, the computer-readable storage medium may include ROM 702 and / or RAM 703 and / or one or more memories other than ROM 702 and RAM 703 described above.

[0169] Embodiments of this application also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowchart. When the computer program product is run on a computer system, the program code is used to enable the computer system to implement the network card fault detection method provided in the embodiments of this application.

[0170] When the computer program is executed by the processor 701, it performs the functions defined in the system / apparatus of this application embodiment. According to the embodiments of this application, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0171] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and may be downloaded and installed via the communication section 709, and / or installed from a removable medium 711. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0172] In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 709, and / or installed from the removable medium 711. When the computer program is executed by the processor 701, it performs the functions defined in the system of this application embodiment. According to the embodiments of this application, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0173] According to embodiments of this application, program code for executing the computer programs provided in the embodiments of this application can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0174] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0175] Those skilled in the art will understand that the features described in the various embodiments of this application can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in this application. In particular, the features described in the various embodiments of this application can be combined and / or combined in various ways without departing from the spirit and teachings of this application. All such combinations and / or combinations fall within the scope of this application.

Claims

1. A method for detecting network interface card (NIC) faults, characterized in that, include: Receive probe requests sent by the computing power cluster; The computing power cluster is configured with multiple network interface cards; Based on the probe request, obtain the configuration information of the computing power cluster; Detection commands are generated based on the configuration information of the computing cluster; The detection command is sent to the computing power cluster so that the computing power cluster can perform fault detection on each of the network cards according to the detection task indicated by the detection command.

2. The method according to claim 1, characterized in that, The generation of detection commands based on the configuration information of the computing power cluster includes: The hierarchical structure of the computing power cluster is determined based on the computing power cluster configuration information; According to the hierarchical structure, a detection task is assigned to at least one target network interface card (NIC) in the computing power cluster; the detection task is used to instruct the target NIC to perform fault detection on other NICs in the cluster. Detection commands are generated based on the detection tasks of each target network interface card.

3. The method according to claim 2, characterized in that, The hierarchical structure is composed of a first preset number of computing power units; each computing power unit includes a second preset number of computing power nodes, and each computing power node includes a third preset number of network cards; the first preset number, the second preset number, and the third preset number may be the same or different.

4. The method according to claim 3, characterized in that, Assigning probing tasks to at least one target network interface card in the computing power cluster according to the hierarchical structure includes: When the first preset quantity, the second preset quantity, and the third preset quantity are the same, for each computing node in the same computing power unit, the i-th network card in the i-th computing power node is determined as the target network card, and a probe task within the unit is assigned to the target network card. The probe task within the unit is used to instruct the target network card to probe the i-th network card of all computing power nodes in the computing power unit, where i is a positive integer. For each computing unit within the same computing power cluster, the network card of the j-th computing node in the j-th computing unit is identified as the target network card, and an intra-cluster probing task is assigned to the target network card. The intra-cluster probing task is used to instruct the target network card to probe the network card of the j-th computing node in all computing units within the computing power cluster, where j is a positive integer.

5. The method according to claim 3, characterized in that, Assigning probing tasks to at least one target network interface card in the computing power cluster according to the hierarchical structure includes: When the first preset quantity, the second preset quantity, and the third preset quantity are different, for the same computing power unit, when the number of computing power nodes in the computing power unit is less than the second preset quantity, the detection tasks corresponding to the portion exceeding the number of computing power nodes are allocated to other computing power nodes in the computing power unit in a preset order. For each computing unit within the same computing power cluster, when the number of computing power units in the computing power cluster is less than the first preset number, the detection tasks corresponding to the portion exceeding the number of computing power units are allocated to other computing power units in the computing power cluster in a preset order.

6. The method according to claim 1, characterized in that, The step of obtaining the configuration information of the computing cluster based on the probe request includes: The configuration information of the computing cluster indicated by the probe request is obtained from the first database.

7. The method according to claim 1, characterized in that, The method further includes: The system receives the detection results fed back by the computing power cluster and saves the detection results to the second database.

8. The method according to claim 7, characterized in that, The method further includes: The detection results are sent to a message queue cluster, so that the message queue cluster sends the detection results to a display terminal for display.

9. A network card fault detection system, characterized in that, include: A computing power cluster, which is configured from multiple network interface cards; Control device, used to receive probe requests sent by the computing power cluster; Based on the probe request, obtain the configuration information of the computing power cluster; Detection commands are generated based on the configuration information of the computing cluster; The detection command is sent to the computing power cluster; The computing power cluster is also used to perform fault detection on each of the network cards according to the detection task indicated by the detection command to obtain the detection results.

10. The system according to claim 9, characterized in that, The system also includes: The first database is used to store the configuration information of the computing power cluster; A second database is used to store the detection results.

11. The system according to claim 9, characterized in that, The system also includes: Display terminal; A message queue cluster is used to receive the detection results and send them to a display terminal for display.

12. An electronic device, characterized in that, include: One or more processors; Memory, used to store one or more computer programs. The characteristic feature is that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1 to 8.

13. A computer-readable storage medium, characterized in that, It stores a computer program or instructions thereon, characterized in that, when the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 8.

Images