Network management method and device, computer device, storage medium and program product

Abstract

The application relates to a network management method and device, computer equipment, a storage medium and a computer program product. The method comprises the following steps: detecting whether the initial connection state of a to-be-probed connection link in an infinite bandwidth network is in a normal state; the to-be-probed connection link comprises a first connection end and a second connection end; if the initial connection state of the to-be-probed connection link is in the normal state, a fault detection instruction is sent from the first connection end of the to-be-probed connection link to the second connection end of the to-be-probed connection link; the fault detection instruction comprises management data reports corresponding to a subnet management interface in the infinite bandwidth network; the to-be-probed connection link is subjected to fault detection according to the fault detection instruction, a fault detection result is generated, and the infinite bandwidth network is subjected to network management based on the fault detection result. The method can improve the security of the IB network.

Description

Technical Field

[0001] This application relates to the field of network communication technology, and in particular to a network management method, apparatus, computer equipment, storage medium, and computer program product. Background Technology

[0002] Infiniband (IB) is a high-speed network technology and standard. High-performance clusters typically use networks corresponding to Infiniband (i.e., Infiniband networks, or IB networks) as high-speed computing or storage networks. To improve the availability and reliability of IB networks, network management is required.

[0003] The traditional method is to manage the IB network through a subnet manager.

[0004] However, in traditional IB network management, when the subnet manager receives a large number of processing requests simultaneously, it can overload the subnet manager, leading to IB network anomalies or even network paralysis. Therefore, the traditional method of managing IB networks through subnet managers suffers from low security. Summary of the Invention

[0005] Therefore, it is necessary to provide a network management method, apparatus, computer equipment, computer-readable storage medium, and computer program product that can improve the security of IB networks in response to the above-mentioned technical problems.

[0006] Firstly, this application provides a network management method. The method includes:

[0007] The initial connection status of the link to be probed in the unlimited bandwidth network is detected to determine whether it is in a normal state; the link to be probed includes a first connection end and a second connection end;

[0008] If the initial connection state of the link to be probed is normal, then the first connection end of the link to be probed is controlled to send a fault detection command to the second connection end of the link to be probed; the fault detection command includes a management data packet corresponding to the subnet management interface in the unlimited bandwidth network;

[0009] The fault detection command is used to detect faults in the connection link to be detected, generate fault detection results, and perform network management on the unlimited bandwidth network based on the fault detection results.

[0010] Because the fault detection command in this embodiment includes management data packets corresponding to the subnet management interface, it is possible to directly control the first connection end of the link to be detected to send the fault detection command to the second connection end of the link to be detected. This avoids sending ICMP packets or path query requests to the subnet manager, thus preventing the excessive load on the subnet manager, which can lead to IB network anomalies or even IB network paralysis, as is common in traditional methods where link fault detection is performed through the subnet manager. Subsequently, fault detection results are generated directly on the link to be detected based on the fault detection command, and network management can be performed on the unlimited bandwidth network based on these results, improving the security of the IB network.

[0011] In one embodiment, the step of controlling the first connection end of the connection to be detected to send a fault detection command to the second connection end of the connection to be detected includes:

[0012] Obtain the management data packet corresponding to the subnet management interface, and generate the fault detection command based on the management data packet corresponding to the subnet management interface;

[0013] The fault detection command is sent from the first connection end to the second connection end by controlling the local identifier.

[0014] In this embodiment, a fault detection command is generated based on the management data packet corresponding to the subnet management interface. Since the local identifier of the first connection end includes the path information of the first connection end, and the local identifier of the second connection end includes the path information of the second connection end, the first connection end can be directly controlled to send a fault detection command containing the management data packet corresponding to the subnet management interface to the second connection end through the local identifier. Therefore, link fault detection can be performed directly on the connection link to be detected through the management data packet corresponding to the subnet management interface, without sending ICMP packets or path query requests to the subnet manager. This avoids the problem of excessive load on the subnet manager, leading to IB network anomalies or even IB network paralysis, which is common in traditional methods when link fault detection is performed through the subnet manager. Therefore, the security of the IB network can be improved.

[0015] In one embodiment, the step of performing fault detection on the connection link to be detected according to the fault detection command and generating a fault detection result includes:

[0016] The system detects whether the first connection end receives a fault response instruction sent by the second connection end within a first preset time period after sending the fault detection instruction; the fault response instruction is generated in response to the fault detection instruction, and the fault response instruction includes a management data packet corresponding to the subnet management interface in the unlimited bandwidth network.

[0017] If the first connection end receives a fault response command sent by the second connection end within a first preset time period, then the fault detection result is determined to be that the current connection status of the connection link to be detected is in a normal state.

[0018] Since the fault response command is generated in response to the fault detection command, and the fault response command includes the management datagram corresponding to the subnet management interface in the unlimited bandwidth network, this embodiment can directly perform link fault detection on the probed connection link through the fault response command containing the management datagram corresponding to the subnet management interface. This avoids sending ICMP packets or path query requests to the subnet manager, thus preventing the excessive load on the subnet manager caused by traditional methods that rely on the subnet manager for link fault detection, which could lead to IB network anomalies or even IB network paralysis. Therefore, it improves the security of the IB network.

[0019] In one embodiment, the method further includes:

[0020] If the first connection end does not receive a fault response instruction sent by the second connection end within a first preset time period, then the fault detection result is determined to be that the current connection status of the connection link to be detected is not in a normal state.

[0021] The process involves repeatedly sending a fault detection command from the first connection end of the link to be detected to the second connection end of the link to be detected; detecting whether the first connection end receives a fault response command from the second connection end within a first preset time period after sending the fault detection command, until it is determined that the fault detection result indicates that the current connection status of the link to be detected is normal.

[0022] In this embodiment, if the first connection end does not receive a fault response instruction from the second connection end within a first preset time period, the link fault detection steps described in the above embodiment are executed repeatedly. This allows for periodic link fault detection of the link to be probed based on the fault response instruction containing the management datagram corresponding to the subnet management interface, until the fault detection result indicates that the current connection status of the link to be probed is normal. This embodiment avoids sending ICMP packets or path query requests to the subnet manager, thus avoiding the excessive load on the subnet manager that can lead to IB network anomalies or even IB network paralysis, as is common in traditional methods. Therefore, it improves the security of the IB network.

[0023] In one embodiment, the method further includes:

[0024] If the initial connection state of the link to be probed is not in a normal state, then it is detected whether the current connection state of the link to be probed is in a probe state; the state not in a normal state includes abnormal state and state of not receiving results;

[0025] If the current connection state of the connection to be detected is in the detection state, then it is detected whether a fault response command sent by the second connection end is received within the second preset time period.

[0026] If the current connection state of the link to be probed is not in the probe state, then return to the execution of sending a fault detection command from the first connection end of the link to be probed to the second connection end of the link to be probed; perform fault detection on the link to be probed according to the fault detection command, generate fault detection results, and perform network management on the unlimited bandwidth network based on the fault detection results.

[0027] In this embodiment, when the initial connection state of the link to be probed is not in a normal state, the current connection state of the link to be probed is checked to see if it is in a probe state. Link fault detection is then performed based on whether the current connection state of the link to be probed is in a probe state and the fault response command. This allows for periodic link fault detection of the link to be probed, based on the fault response command containing the management datagram corresponding to the subnet management interface, until the fault detection result indicates that the current connection state of the link to be probed is in a normal state. Network management can then be performed on the unlimited bandwidth network based on the fault detection result. This embodiment does not require sending ICMP packets or path query requests to the subnet manager, thus avoiding the problem of excessive load on the subnet manager caused by link fault detection through the subnet manager in traditional methods, which could lead to IB network anomalies or even IB network paralysis. Therefore, it improves the security of the IB network.

[0028] In one embodiment, the method further includes:

[0029] If the fault detection result indicates that the current connection status of the connection link to be detected is normal, then the first connection end is controlled to send a reservation connection instruction to the second connection end; the reservation connection instruction carries the reservation connection information corresponding to the connection link to be detected, and the reservation connection instruction includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network;

[0030] The second connection terminal is controlled to send a connection response instruction to the first connection terminal according to the reservation connection instruction; the connection response instruction carries the invitation connection information corresponding to the connection link to be probed, and the connection response instruction includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network;

[0031] Based on the invitation connection information, control the first connection end to establish a connection with the second connection end.

[0032] In this embodiment, for a probed connection link whose connection status remains normal, the reservation connection instruction can carry the reservation connection information corresponding to the probed connection link, and the connection response instruction can carry the invitation connection information corresponding to the probed connection link. This establishes a connection between the first and second connection ends of the probed connection link through a "reservation-registration-invitation-connection" process. Furthermore, this embodiment can determine a preset rate based on the subnet manager's processing efficiency and the number of links to be connected, thus controlling the frequency of connection establishment. This avoids the problem of excessive load on the subnet manager caused by receiving a large number of path query requests in a short period during traditional connection establishment processes, which could lead to IB network anomalies or even IB network paralysis. Therefore, establishing a connection between the first and second connection ends of the probed connection link through the "reservation-registration-invitation-connection" process improves the security of the IB network.

[0033] In one embodiment, controlling the second connection end to send a connection response instruction to the first connection end according to the reserved connection instruction includes:

[0034] The second connection terminal is controlled to register the reservation connection information on the second connection terminal according to the reservation connection instruction;

[0035] Update the priority of the link to be probed in the list of links to be connected based on the reservation connection information, and obtain the target priority of the link to be probed in the list of links to be connected;

[0036] Based on the target priority, the second connection end is controlled to send a connection response command to the first connection end at a preset rate; the preset rate is determined based on the processing efficiency of the subnet manager and the number of links to be connected.

[0037] This embodiment can determine the preset rate based on the subnet manager's processing efficiency and the number of links to be connected. This preset rate controls the frequency of connection establishment, thus avoiding the problem of excessive load on the subnet manager caused by receiving too many path query requests in a short period during traditional connection establishment processes, which could lead to IB network anomalies or even IB network paralysis. Therefore, establishing a connection between the first and second connection ends of the link to be probed through a "reservation-registration-invitation-connection" method improves the security of the IB network.

[0038] In one embodiment, the method further includes:

[0039] If the fault detection result indicates that the current connection status of the link to be detected has switched from an abnormal state to a normal state, then the second connection end is controlled to send a connection notification instruction to the first connection end; the connection notification instruction includes a management datagram corresponding to the subnet management interface in the unlimited bandwidth network;

[0040] According to the connection notification instruction, control the first connection end to establish a connection with the second connection end.

[0041] In this embodiment, for a probed connection link whose current connection state changes from an abnormal state to a normal state, the second connection end in the probed connection link can actively notify the first connection end in the probed connection link to establish a connection. This not only ensures that the connection is established promptly after the fault in the probed connection link is resolved, but also avoids the problem of the first connection end in the probed connection link frequently attempting to establish a connection while the fault in the probed connection link has not yet been resolved. This would prevent the subnet manager from receiving too many path query requests in a short period, leading to excessive load on the subnet manager, which could cause IB network anomalies or even paralysis. Therefore, having the second connection end actively notify the first connection end to establish a connection improves the security of the IB network.

[0042] Secondly, this application also provides a network management method. The method includes:

[0043] If the fault detection result indicates that the current connection status of the link to be detected is normal, then the first connection end is controlled to send a reservation connection instruction to the second connection end; the reservation connection instruction carries the reservation connection information corresponding to the link to be detected, and the reservation connection instruction includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network; the fault detection result is the result of fault detection on the link to be detected in the unlimited bandwidth network.

[0044] The second connection terminal is controlled to send a connection response instruction to the first connection terminal according to the reservation connection instruction; the connection response instruction carries the invitation connection information corresponding to the connection link to be probed, and the connection response instruction includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network;

[0045] Based on the invitation connection information, control the first connection end to establish a connection with the second connection end.

[0046] In this embodiment, for a probed connection link whose connection status remains normal, the reservation connection instruction can carry the reservation connection information corresponding to the probed connection link, and the connection response instruction can carry the invitation connection information corresponding to the probed connection link. This establishes a connection between the first and second connection ends of the probed connection link through a "reservation-registration-invitation-connection" process. Furthermore, this embodiment can determine a preset rate based on the subnet manager's processing efficiency and the number of links to be connected, thus controlling the frequency of connection establishment. This avoids the problem of excessive load on the subnet manager caused by receiving a large number of path query requests in a short period during traditional connection establishment processes, which could lead to IB network anomalies or even IB network paralysis. Therefore, establishing a connection between the first and second connection ends of the probed connection link through the "reservation-registration-invitation-connection" process improves the security of the IB network.

[0047] Thirdly, this application also provides a network management device. The device includes:

[0048] The detection module is used to detect whether the initial connection status of the link to be detected in the unlimited bandwidth network is in a normal state; the link to be detected includes a first connection end and a second connection end;

[0049] The sending module is configured to, if the initial connection state of the connection to be probed is in a normal state, control the first connection end of the connection to be probed to send a fault detection command to the second connection end of the connection to be probed; the fault detection command includes a management data packet corresponding to the subnet management interface in the unlimited bandwidth network;

[0050] The network management module is used to perform fault detection on the connection link to be detected according to the fault detection instruction, generate fault detection results, and perform network management on the unlimited bandwidth network based on the fault detection results.

[0051] Fourthly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the methods in any of the above embodiments.

[0052] Fifthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements the steps of the methods in any of the above embodiments.

[0053] Sixthly, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, implements the steps of the methods in any of the above embodiments.

[0054] The aforementioned network management method, apparatus, computer equipment, storage medium, and computer program product detect whether the initial connection state of a link to be probed in an unlimited bandwidth network is normal. The link to be probed includes a first connection end and a second connection end. If the initial connection state of the link to be probed is normal, the first connection end of the link to be probed sends a fault detection command to the second connection end of the link to be probed. The fault detection command includes a management data packet corresponding to the subnet management interface in the unlimited bandwidth network. Fault detection is performed on the link to be probed according to the fault detection command, a fault detection result is generated, and network management is performed on the unlimited bandwidth network based on the fault detection result. Since the fault detection command of this application includes a management data packet corresponding to the subnet management interface, it is possible to directly control the first connection end of the link to be probed to send a fault detection command to the second connection end of the link to be probed. This avoids the problem of excessive load on the subnet manager, leading to IB network anomalies or even IB network paralysis, which is common in traditional methods when link fault detection is performed through the subnet manager. Subsequently, based on the fault detection command, fault detection is directly performed on the connection link to be detected to generate fault detection results, and network management can be performed on the unlimited bandwidth network based on the fault detection results, which can improve the security of the IB network. Attached Figure Description

[0055] Figure 1 This is a diagram illustrating the application environment of a network management method in one embodiment;

[0056] Figure 2 This is a flowchart illustrating a network management method in one embodiment;

[0057] Figure 3 This is a flowchart illustrating the fault detection command sending steps in one embodiment;

[0058] Figure 4 This is a flowchart illustrating the steps for generating fault detection results in one embodiment;

[0059] Figure 5 This is a flowchart illustrating the first loop step in one embodiment;

[0060] Figure 6 This is a flowchart illustrating the second loop step in one embodiment;

[0061] Figure 7This is a flowchart illustrating the first connection establishment step in one embodiment;

[0062] Figure 8 This is a schematic diagram of the process of the first connection end actively establishing a connection in one embodiment;

[0063] Figure 9 This is a flowchart illustrating the second connection establishment step in one embodiment;

[0064] Figure 10 This is a schematic diagram of the process of the first connection end passively establishing a connection in another embodiment;

[0065] Figure 11 This is a flowchart illustrating the network management method in one optional embodiment;

[0066] Figure 12 This is a structural block diagram of a network management device in one embodiment;

[0067] Figure 13 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0068] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0069] Infiniband (IB) is a high-speed network technology and standard. High-performance clusters typically use networks corresponding to Infiniband (i.e., Infiniband networks, or IB networks) as high-speed computing or storage networks. To improve the availability and reliability of IB networks, network management is required.

[0070] Traditionally, IB networks are managed through a Subnet Manager (SM). The process of managing an IB network includes link failure detection and connectivity management. Link failure detection encompasses network failure detection and node failure detection.

[0071] The traditional method for link failure detection in an IB network involves the following steps: First, each node's System Management Agent (SMA) sends Internet Control Message Protocol (ICMP) packets to the Subnet Manager. Second, the Subnet Manager receives the ICMP packets and processes the data within them. Third, the Subnet Manager sends response packets to each SMA. Finally, each SMA determines whether a link failure has occurred based on the response packets. Furthermore, during ICMP packet-based link failure detection, each node's SMA sends a path query request to the Subnet Manager to retrieve path information.

[0072] The traditional steps for managing IB network connectivity are as follows: If a normal link is detected, connectivity management is performed through the Communication Manager Application Program Interface (CM API) corresponding to Remote Direct Memory Access (RDMA). During this process, the subnet management agent for each node in the cluster needs to send a path query request to the subnet manager to retrieve path information.

[0073] However, when the cluster contains a large number of nodes, the subnet management agent corresponding to each node sends ICMP packets or path lookup requests to the subnet manager, causing the subnet manager to receive a large number of ICMP packets or path lookup requests. Furthermore, the subnet manager's processing capacity is limited, so it cannot process the data in multiple ICMP packets or path lookup requests in a timely manner, leading to network outages for some nodes in the cluster, or even paralyzing the entire cluster network. Therefore, the traditional method of network management of the IB network through ICMP packets and the corresponding communication management application interface of RDMA can overload the subnet manager when it receives a large number of processing requests simultaneously, causing IB network anomalies or even paralysis. Thus, the traditional method of network management of the IB network through the subnet manager has a low security issue.

[0074] The network management method provided in this application embodiment can be applied to, for example, Figure 1The application environment is shown. In this environment, the subnet manager 120 communicates with the cluster 140 via a network. The cluster 140 includes multiple computer devices 142, which can be servers acting as nodes within the cluster 140. Each computer device 142 is equipped with a subnet management agent. The data storage system can store data that the subnet manager 120 and the cluster 140 need to process. The data storage system can be integrated into the subnet manager 120, the cluster 140, or it can be hosted on a cloud or other network server. Computer device 142 detects whether the initial connection status of a link to be probed in the unlimited bandwidth network is normal. The link to be probed includes a first connection end and a second connection end. If the initial connection status of the link to be probed is normal, the computer device 142 controls the first connection end of the link to send a fault detection command to the second connection end of the link to be probed. The fault detection command includes a management data packet corresponding to the subnet management interface in the unlimited bandwidth network. The computer device 142 performs fault detection on the link to be probed according to the fault detection command, generates a fault detection result, and performs network management on the unlimited bandwidth network based on the fault detection result. The computer device 142 can be implemented using a standalone server or a cluster of multiple servers capable of primary / backup failover.

[0075] In one embodiment, such as Figure 2 As shown, a network management method is provided, which is applied to... Figure 1 Taking computer device 142 as an example, the following steps are included:

[0076] S220, Detect whether the initial connection status of the link to be probed in the unlimited bandwidth network is in a normal state; the link to be probed includes a first connection end and a second connection end.

[0077] Infiniband (IB) is a high-speed network technology and standard. High-performance clusters typically use the network corresponding to Infiniband (i.e., the Infiniband network, or IB network) as a high-speed computing or storage network to enable inter-process communication. Since network management of an IB network includes link fault detection and connection management, and link fault detection includes network fault detection and node fault detection, the link to be probed refers to the link in the Infiniband network that needs fault detection. The link to be probed includes a first connection end and a second connection end. The initial connection state of the link to be probed represents the connection state of the link at the initial moment, which includes, but is not limited to, the normal state.

[0078] Optionally, the computer device 142 can monitor the initial connection status of the link to be probed in the unlimited bandwidth network in real time at the initial moment, or the computer device 142 can obtain the initial connection status of the link to be probed in the unlimited bandwidth network from the historical monitoring results of the link to be probed. Afterwards, the computer device 142 can detect whether the initial connection status of the link to be probed in the unlimited bandwidth network is normal.

[0079] S240, if the initial connection state of the link to be probed is in a normal state, then control the first connection end of the link to be probed to send a fault detection command to the second connection end of the link to be probed; the fault detection command includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network.

[0080] Optionally, since the connection to be probed includes a first connection end and a second connection end, when the initial connection state of the connection to be probed is detected to be normal, the computer device 142 can control the first connection end of the connection to be probed to send a fault detection command to the second connection end of the connection to be probed to perform fault detection on the connection to be probed. The fault detection command is used to detect faults in the connection state of the connection to be probed, and includes a management datagram corresponding to the subnet management interface in the unlimited bandwidth network. The management datagram corresponding to the subnet management interface (SMI MAD) is a common management datagram in unlimited bandwidth networks.

[0081] S260 performs fault detection on the connection link to be detected according to the fault detection command, generates fault detection results, and performs network management on the unlimited bandwidth network based on the fault detection results.

[0082] Optionally, the computer device 142 can perform fault detection on the connection link to be detected based on whether the second connection end of the connection link to be detected receives a fault detection command, and generate a fault detection result; or, the computer device 142 can also perform fault detection on the connection link to be detected based on whether the first connection end of the connection link to be detected receives information returned by the second connection end after the second connection end of the connection link to be detected receives a fault detection command, and generate a fault detection result. The fault detection result is used to characterize whether the current connection state of the connection link to be detected is normal, and the fault detection result can include whether the current connection state of the connection link to be detected is normal or not.

[0083] Subsequently, computer device 142 can perform network management on the unlimited bandwidth network based on the fault detection results. For example, if the fault detection result indicates that the current connection status of the link to be detected is normal, computer device 142 can control the first connection end of the link to be detected to establish a connection with the second connection end of the link to be detected. If the fault detection result indicates that the current connection status of the link to be detected is not normal, computer device 142 can clear invalid connections and information in the link to be detected to prevent network congestion.

[0084] In the aforementioned network management method, the initial connection state of the link to be probed in the unlimited bandwidth network is detected to ensure it is in a normal state. The link to be probed includes a first connection end and a second connection end. If the initial connection state of the link to be probed is normal, the first connection end of the link to be probed is controlled to send a fault detection command to the second connection end of the link to be probed. The fault detection command includes a management data packet corresponding to the subnet management interface in the unlimited bandwidth network. Fault detection is performed on the link to be probed according to the fault detection command, a fault detection result is generated, and network management is performed on the unlimited bandwidth network based on the fault detection result. Since the fault detection command of this application includes a management data packet corresponding to the subnet management interface, the first connection end of the link to be probed can be directly controlled to send a fault detection command to the second connection end of the link to be probed. This avoids the problem of excessive load on the subnet manager, leading to IB network anomalies or even IB network paralysis, which is common in traditional methods when link fault detection is performed through the subnet manager. Subsequently, based on the fault detection command, fault detection is directly performed on the connection link to be detected to generate fault detection results, and network management can be performed on the unlimited bandwidth network based on the fault detection results, which can improve the security of the IB network.

[0085] The above embodiments describe controlling the first connection end of the link to be probed to send a fault detection command to the second connection end of the link to be probed. The implementation steps are described below. In one embodiment, such as... Figure 3 As shown, controlling the first connection end of the link to be detected to send a fault detection command to the second connection end of the link to be detected includes:

[0086] S320: Obtain the management data packet corresponding to the subnet management interface, and generate a fault detection command based on the management data packet corresponding to the subnet management interface.

[0087] Optionally, the computer device 142 can obtain the corresponding subnet management interface of the link to be probed based on the first and second connection ends of the link. Then, it can obtain the management datagram (SMI MAD) corresponding to the subnet management interface from the subnet management interface of the link to be probed. Afterwards, the computer device 142 can generate a fault detection command based on the management datagram (SMI MAD) corresponding to the subnet management interface. Here, a datagram refers to the basic unit of data transmitted over a network, and a datagram includes a header and the data itself.

[0088] S340 controls the first connection end to send fault detection commands to the second connection end via a local identifier.

[0089] Optionally, based on the first and second connection ends of the connection link to be probed, the computer device 142 can determine the local identifier of the first connection end and the local identifier of the second connection end of the connection link to be probed. The local identifier (LID) is a symbol used to identify the location information of the connection end. Then, the computer device 142 can control the first connection end to send a fault detection command to the second connection end based on the local identifiers of the first and second connection ends.

[0090] In this embodiment, a fault detection command is generated based on the management data packet corresponding to the subnet management interface. Since the local identifier of the first connection end includes the path information of the first connection end, and the local identifier of the second connection end includes the path information of the second connection end, the first connection end can be directly controlled to send a fault detection command containing the management data packet corresponding to the subnet management interface to the second connection end through the local identifier. Therefore, link fault detection can be performed directly on the connection link to be detected through the management data packet corresponding to the subnet management interface, without sending ICMP packets or path query requests to the subnet manager. This avoids the problem of excessive load on the subnet manager, leading to IB network anomalies or even IB network paralysis, which is common in traditional methods when link fault detection is performed through the subnet manager. Therefore, the security of the IB network can be improved.

[0091] The above embodiments describe the process of detecting faults in the connection link to be detected according to the fault detection command and generating fault detection results. The implementation steps are described below. In one embodiment, such as... Figure 4 As shown, the fault detection is performed on the connection link to be detected according to the fault detection command, and the fault detection results are generated, including:

[0092] S420, detect whether the first connection end receives a fault response command sent by the second connection end within a first preset time period after sending the fault detection command; the fault response command is generated in response to the fault detection command, and the fault response command includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network.

[0093] It should be noted that the principle of link fault detection in this embodiment is as follows: First, link information of the link to be detected is created, and the initial connection state of the link to be detected is initialized. Second, the first connection end of the link to be detected is controlled to directly send a fault detection command to the second connection end of the link to be detected. Third, the second connection end of the link to be detected is controlled to receive the fault detection command and add the link information of the link to be detected to the queue of links to be detected. Then, the second connection end of the link to be detected is controlled to directly send a fault response command to the first connection end of the link to be detected. Finally, whether the connection state of the link to be detected has failed is determined based on whether the first connection end receives the fault response command. Both the fault detection command and the fault response command include the management datagram (SMI MAD) corresponding to the subnet management interface in the unlimited bandwidth network.

[0094] Optionally, the computer device 142 may periodically detect whether the first connection end has received a fault response command from the second connection end within a first preset time period after the first connection end sends the fault detection command. Alternatively, the computer device 142 may directly detect whether the first connection end has received a fault response command from the second connection end at the last moment of the first preset time period after the first connection end sends the fault detection command. The fault response command is generated in response to the fault detection command and includes a management datagram (SMI MAD) corresponding to the subnet management interface in the unlimited bandwidth network. Only when the first connection end receives a fault response command from the second connection end can it be considered that the connection status of the probed link is normal.

[0095] S440, if the first connection end receives a fault response command sent by the second connection end within the first preset time period, then the fault detection result is determined to be that the current connection status of the connection link to be detected is in a normal state.

[0096] Optionally, if the first connection end receives a fault response command sent by the second connection end within a first preset time period, the computer device 142 can update the current connection status of the link to be detected to a normal status in real time, thereby determining that the fault detection result of this link fault detection is that the current connection status of the link to be detected is in a normal state. Here, the current connection status of the link to be detected being in a normal state indicates that the current connection status of the link to be detected is normal, meaning that a connection can be established between the first connection end and the second connection end of the link to be detected. The first preset time period can be set according to actual conditions, and this embodiment does not limit it.

[0097] In this embodiment, it is detected whether the first connection end receives a fault response command from the second connection end within a first preset time period after sending the fault detection command. If the first connection end receives the fault response command from the second connection end within the first preset time period, it is determined that the fault detection result is that the current connection status of the link to be detected is normal. Since the fault response command is generated in response to the fault detection command, and the fault response command includes the management data packet corresponding to the subnet management interface in the infinite bandwidth network, this embodiment can directly perform link fault detection on the link to be detected through the fault response command containing the management data packet corresponding to the subnet management interface. This avoids sending ICMP packets or path query requests to the subnet manager, thus avoiding the problem of excessive load on the subnet manager caused by link fault detection through the subnet manager in traditional methods, which could lead to IB network anomalies or even IB network paralysis. Therefore, it can improve the security of the IB network.

[0098] In one embodiment, such as Figure 5 As shown, a network management method is provided, which also includes:

[0099] S520, if the first connection end does not receive a fault response instruction sent by the second connection end within the first preset time period, the fault detection result is determined to be that the current connection status of the connection link to be detected is not in a normal state.

[0100] Optionally, if the first connection end does not receive a fault response command from the second connection end within a first preset time period, the computer device 142 can determine that the fault detection result is that the current connection state of the connection link to be detected is not in a normal state. Here, "the current connection state of the connection link to be detected is not in a normal state" indicates a fault in the current connection state of the connection link to be detected, meaning that a connection cannot be established between the first connection end and the second connection end of the connection link to be detected. Furthermore, the computer device 142 needs to clear invalid connections and information in the connection link to be detected to prevent network congestion.

[0101] S540, cyclically executes the control to send a fault detection command from the first connection end of the connection to be detected to the second connection end of the connection to be detected; detects whether the first connection end receives a fault response command sent by the second connection end within a first preset time period after sending the fault detection command, until it is determined that the fault detection result is that the current connection status of the connection to be detected is in a normal state.

[0102] Optionally, if the fault detection result indicates that the current connection state of the link to be detected is not in a normal state, return to S240 to re-initiate the link fault detection. This can be understood as follows: First, set the current connection state of the link to be detected to a detection state; then, repeatedly execute the command to send a fault detection instruction from the first connection end of the link to be detected to the second connection end of the link to be detected; and check whether the first connection end receives a fault response instruction from the second connection end within a first preset time period after sending the fault detection instruction. If the first connection end receives a fault response instruction from the second connection end within the first preset time period, it is determined that the fault detection result indicates that the current connection state of the link to be detected is in a normal state. At this point, the loop stops, and a connection can be established between the first connection end and the second connection end of the link to be detected.

[0103] In this embodiment, if the first connection end does not receive a fault response instruction from the second connection end within a first preset time period, the fault detection result is determined to be that the current connection state of the link to be detected is not in a normal state. At this time, by repeatedly executing the link fault detection steps described in the above embodiment, the link to be detected can be periodically checked for faults based on the fault response instruction containing the management datagram corresponding to the subnet management interface, until it is determined that the current connection state of the link to be detected is in a normal state. This embodiment does not require sending ICMP packets or path query requests to the subnet manager, thus avoiding the problem of excessive load on the subnet manager caused by link fault detection through the subnet manager in traditional methods, which could lead to IB network anomalies or even IB network paralysis. Therefore, it can improve the security of the IB network.

[0104] In one embodiment, such as Figure 6 As shown, a network management method is provided, which also includes:

[0105] S620, if the initial connection state of the link to be probed is not in a normal state, then check whether the current connection state of the link to be probed is in a probe state; not in a normal state includes abnormal state and state where no result has been received.

[0106] Optionally, if the initial connection state of the link to be probed is not detected to be in a normal state, and since the initial connection state not being in a normal state includes both an abnormal state and a state where no result has been received, it indicates that the initial connection state of the link to be probed can be in an abnormal state, or the initial connection state of the link to be probed can be in a state where no result has been received (i.e., the link to be probed is initially in a fault detection state, referred to as the probe state). At this time, the computer device 142 can then detect whether the current connection state of the link to be probed is in the probe state.

[0107] S640, if the current connection status of the link to be probed is in the probe state, then detect whether a fault response command sent by the second connection end is received within the second preset time period.

[0108] Optionally, if the current connection state of the link to be detected is in a detection state, it indicates that the link to be detected is still in fault detection at the current moment. In this case, optionally, the computer device 142 can periodically detect whether the first connection end has received a fault response command sent by the second connection end within a second preset time period after the first connection end sends the fault detection command. Alternatively, the computer device 142 can also directly detect whether the first connection end has received a fault response command sent by the second connection end at the last moment of the second preset time period after the first connection end sends the fault detection command. The second preset time period can be set according to actual conditions, and this embodiment does not limit it.

[0109] S660, if the current connection state of the link to be detected is not in the detection state, then return to the execution control to send a fault detection command from the first connection end of the link to be detected to the second connection end of the link to be detected; perform fault detection on the link to be detected according to the fault detection command, generate fault detection results, and perform network management on the unlimited bandwidth network based on the fault detection results.

[0110] Optionally, if the current connection state of the link to be detected is not in the detection state, it indicates that the link to be detected is in an abnormal state at the current moment. In this case, return to S240 to re-initiate link fault detection. This can be understood as follows: First, set the current connection state of the link to be detected to the detection state; then, repeatedly execute the command to send a fault detection instruction from the first connection end of the link to be detected to the second connection end of the link to be detected; and check whether the first connection end receives a fault response instruction from the second connection end within a first preset time period after sending the fault detection instruction. If the first connection end receives a fault response instruction from the second connection end within the first preset time period, it is determined that the current connection state of the link to be detected is in a normal state. At this point, the loop stops, and network management is performed on the unlimited bandwidth network based on the fault detection result.

[0111] In this embodiment, if the initial connection state of the link to be probed is not in a normal state, the current connection state of the link to be probed is checked to see if it is in a probe state. If the current connection state of the link to be probed is in a probe state, it is checked whether a fault response command sent by the second connection end is received within a second preset time period; if the current connection state of the link to be probed is not in a probe state, the link fault detection steps in the above embodiment are executed repeatedly. This allows for periodic link fault detection of the link to be probed based on the fault response command containing the management datagram corresponding to the subnet management interface, until the fault detection result indicates that the current connection state of the link to be probed is in a normal state. Based on the fault detection result, network management of the unlimited bandwidth network can then be performed. This embodiment does not require sending ICMP packets or path query requests to the subnet manager, thus avoiding the problem of excessive load on the subnet manager caused by link fault detection through the subnet manager in traditional methods, which can lead to IB network anomalies or even IB network paralysis. Therefore, it can improve the security of the IB network.

[0112] The above embodiments describe the process of detecting faults in the connection link to be detected according to the fault detection command, and generating fault detection results. This application embodiment can also perform connection management on an unlimited bandwidth network based on the fault detection results; the implementation steps are described below. In one embodiment, such as... Figure 7 As shown, a network management method is provided, which also includes:

[0113] S720, if the fault detection result indicates that the current connection status of the link to be detected is normal, then the first connection end is controlled to send a reservation connection instruction to the second connection end; the reservation connection instruction carries the reservation connection information corresponding to the link to be detected, and the reservation connection instruction includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network.

[0114] The current connection status of the link to be probed being in a normal state includes both a continuously normal connection status and a transition from an abnormal state to a normal state. The connection reservation instruction is used to reserve a connection between the first and second connection ends. The reservation instruction carries the reservation connection information corresponding to the link to be probed and includes the management datagram (SMI MAD) corresponding to the subnet management interface in the unlimited bandwidth network. During the connection establishment process, the first connection end can be referred to as the connection initiator or client, and the second connection end can be referred to as the connection passive end or server.

[0115] Optionally, such as Figure 8 As shown, Figure 8 This is a schematic diagram illustrating the process of the first connection end actively establishing a connection in one embodiment. If the fault detection result indicates that the connection status of the link to be detected remains normal, in step S801, the computer device 142 can control the first connection end to send a reservation connection command carrying the reservation connection information corresponding to the link to be detected to the second connection end at a preset frequency. The preset frequency is specifically set based on the processing efficiency of the subnet manager, and this embodiment does not limit its specific frequency.

[0116] Furthermore, the computer device 142 can also detect whether the first connection end has successfully sent a reservation connection command to the second connection end. If it is detected that the first connection end has failed to send a reservation connection command to the second connection end, or if it does not receive a connection response command from the second connection end within a third preset time period, it waits for a fourth preset time period before resending the reservation connection command. If it is detected that the first connection end has successfully sent a reservation connection command to the second connection end, it adds the link information of the link to be probed to the list of links to be connected, and waits for a connection response command from the second connection end within the third preset time period. The third and fourth preset time periods can be set according to actual conditions, and this embodiment does not limit them.

[0117] S740, the second connection end is controlled to send a connection response instruction to the first connection end according to the reservation connection instruction; the connection response instruction carries the invitation connection information corresponding to the connection link to be probed, and the connection response instruction includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network.

[0118] Optionally, after the second connection end receives the reservation connection instruction sent by the first connection end, the computer device 142 can control the second connection end to send a connection response instruction carrying invitation connection information corresponding to the connection link to be probed to the first connection end according to the reservation connection instruction. The connection response instruction is generated in response to the reservation connection instruction, and carries the invitation connection information corresponding to the connection link to be probed. The connection response instruction also includes a management datagram (SMI MAD) corresponding to the subnet management interface in the unlimited bandwidth network.

[0119] In one embodiment, S740 includes:

[0120] S742 controls the second connection end to register the reservation connection information on the second connection end according to the reservation connection instruction.

[0121] Optionally, combined Figure 8 As shown in step S802, based on the reserved connection information corresponding to the link to be probed carried in the reserved connection instruction, the computer device 142 can control the second connection end to register the reserved connection information on the second connection end. The process of registering the reserved connection information can be understood as follows: after receiving the reserved connection instruction sent by the first connection end, the second connection end checks whether the list of links to be connected includes the link information of the link to be probed. If it is detected that the list of links to be connected does not include the link information of the link to be probed, then the link information of the link to be probed is created, and the link information of the link to be probed is added to the list of links to be connected.

[0122] S744, update the priority of the link to be probed in the list of links to be connected according to the reservation connection information, and obtain the target priority of the link to be probed in the list of links to be connected.

[0123] Optionally, combined Figure 8 As shown, when the computer device 142 detects that the list of links to be connected includes the link to be probed, it can update the priority of the link to be probed in the list of links to be connected based on the reserved connection information, that is, increase the priority of the link to be probed in the list of links to be connected, thereby obtaining the target priority of the link to be probed in the list of links to be connected. The target priority of the link to be probed in the list of links to be connected is higher than the initial priority of the link to be probed in the list of links to be connected.

[0124] Furthermore, the computer device 142 can also detect whether the second connection end receives a connection establishment command corresponding to the connection link to be probed sent by the first connection end within a fifth preset time period. If it is detected that the second connection end does not receive a connection establishment command corresponding to the connection link to be probed sent by the first connection end within the fifth preset time period, then the connection link to be probed and the invalid link information in the connection link to be probed are cleared.

[0125] S746, according to the target priority, controls the second connection end to send a connection response command to the first connection end at a preset rate; the preset rate is determined based on the processing efficiency of the subnet manager and the number of links to be connected.

[0126] Optionally, combined Figure 8 As shown in step S803, based on the target priority of the link to be probed in the list of links to be connected, the computer device 142 can control the second connection end to send a connection response command carrying the invitation connection information corresponding to the link to be probed to the first connection end at a preset rate. The preset rate is the maximum rate that the second connection end can send, determined based on the processing efficiency of the subnet manager and the number of links to be connected. For example, assuming the subnet manager's processing efficiency is 100 links / second and the number of links to be connected at the current moment is 10, then the preset rate (i.e., the maximum rate that the second connection end can send) is equal to the subnet manager's processing efficiency divided by the number of links to be connected at the current moment, i.e., 10 links / second.

[0127] S760 controls the first connection end to establish a connection with the second connection end based on the invitation connection information.

[0128] Optionally, combined Figure 8 As shown in S804, after the first connection end receives the connection response instruction sent by the second connection end carrying the invitation connection information corresponding to the connection link to be detected, the computer device 142 can remove the connection link to be detected from the list of connection links to be connected, and according to the invitation connection information, control the first connection end to send the connection establishment instruction corresponding to the connection link to be detected to the second connection end at a preset rate, thereby controlling the first connection end to establish a connection with the second connection end.

[0129] In this embodiment, for a probed connection link whose connection status remains normal, the reservation connection instruction can carry the reservation connection information corresponding to the probed connection link, and the connection response instruction can carry the invitation connection information corresponding to the probed connection link. This establishes a connection between the first and second connection ends of the probed connection link through a "reservation-registration-invitation-connection" process. Furthermore, this embodiment can determine a preset rate based on the subnet manager's processing efficiency and the number of links to be connected, thus controlling the frequency of connection establishment. This avoids the problem of excessive load on the subnet manager caused by receiving a large number of path query requests in a short period during traditional connection establishment processes, which could lead to IB network anomalies or even IB network paralysis. Therefore, establishing a connection between the first and second connection ends of the probed connection link through the "reservation-registration-invitation-connection" process improves the security of the IB network.

[0130] The above embodiments describe connection management for unlimited bandwidth networks based on fault detection results. The following describes another implementation step for connection management. In one embodiment, such as... Figure 9 As shown, a network management method is provided, which also includes:

[0131] S920, if the fault detection result is that the current connection status of the link to be detected has switched from an abnormal state to a normal state, then control the second connection end to send a connection notification instruction to the first connection end; the connection notification instruction includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network.

[0132] Optionally, such as Figure 10 As shown, Figure 10 This is a schematic diagram of the passive connection establishment process of the first connection end in another embodiment. S1001, the second connection end can pre-save the link information of the link to be detected before a fault occurs. If the fault detection result indicates that the current connection state of the link to be detected has switched from an abnormal state to a normal state, the computer device 142 can control the second connection end to load the link information of the link to be detected. S1002, the computer device 142 can control the second connection end to send a connection notification instruction to the first connection end at a preset rate. The connection notification instruction is sent by the second connection end to notify the first connection end that a connection can be established, and includes a management datagram (SMI MAD) corresponding to the subnet management interface in the unlimited bandwidth network. For example, the situation where the current connection state of the link to be detected switches from an abnormal state to a normal state may include, but is not limited to, the restart of a node or process in the link to be detected.

[0133] S940, according to the connection notification instruction, controls the first connection end to establish a connection with the second connection end.

[0134] Optionally, combined Figure 10 As shown in step S1003, after the first connection end receives the connection notification instruction sent by the second connection end, the computer device 142 can control the first connection end to send the connection establishment instruction corresponding to the connection link to be detected to the second connection end at a preset rate according to the connection notification instruction, thereby controlling the first connection end to establish a connection with the second connection end.

[0135] In this embodiment, for a probed connection link whose current connection state changes from an abnormal state to a normal state, the second connection end in the probed connection link can actively notify the first connection end in the probed connection link to establish a connection. This not only ensures that the connection is established promptly after the fault in the probed connection link is resolved, but also avoids the problem of the first connection end in the probed connection link frequently attempting to establish a connection while the fault in the probed connection link has not yet been resolved. This would prevent the subnet manager from receiving too many path query requests in a short period, leading to excessive load on the subnet manager, which could cause IB network anomalies or even paralysis. Therefore, having the second connection end actively notify the first connection end to establish a connection improves the security of the IB network.

[0136] In one embodiment, a network management method is also provided, in which the method is applied to Figure 1 Taking computer device 142 as an example, the following steps are included:

[0137] If the fault detection result indicates that the current connection status of the link to be detected is normal, then the first connection end is controlled to send a reservation connection instruction to the second connection end. The reservation connection instruction carries the reservation connection information corresponding to the link to be detected, and the reservation connection instruction includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network. The fault detection result is the result of fault detection on the link to be detected in the unlimited bandwidth network.

[0138] The control second connection end sends a connection response instruction to the first connection end according to the reservation connection instruction; the connection response instruction carries the invitation connection information corresponding to the connection link to be probed, and the connection response instruction includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network.

[0139] Based on the invitation connection information, control the first connection end to establish a connection with the second connection end.

[0140] For a detailed description of this embodiment, please refer to the above embodiments, which will not be repeated here.

[0141] In the aforementioned network management method, for a probed connection link whose connection status remains normal, the reserved connection instruction can carry the reserved connection information corresponding to the probed connection link, and the connection response instruction can carry the invitation connection information corresponding to the probed connection link. This establishes a connection between the first and second connection ends of the probed connection link through a "reservation-registration-invitation-connection" process. Furthermore, this embodiment can determine a preset rate based on the subnet manager's processing efficiency and the number of links to be connected, thus controlling the frequency of connection establishment. This avoids the problem of excessive load on the subnet manager caused by receiving a large number of path query requests in a short period during traditional connection establishment processes, which could lead to IB network anomalies or even IB network paralysis. Therefore, establishing a connection between the first and second connection ends of the probed connection link through the "reservation-registration-invitation-connection" process improves the security of the IB network.

[0142] In an optional embodiment, such as Figure 11 As shown, a network management method is provided, applied to computer device 142, including:

[0143] S1102, Detect whether the initial connection status of the link to be probed in the unlimited bandwidth network is in a normal state; the link to be probed includes a first connection end and a second connection end; if yes, then execute S1104; if no, then execute S1114.

[0144] S1104, Obtain the management data packet corresponding to the subnet management interface, and generate a fault detection command based on the management data packet corresponding to the subnet management interface;

[0145] S1106, the first connection end is controlled to send a fault detection command to the second connection end through a local identifier; the fault detection command includes a management datagram corresponding to the subnet management interface in the unlimited bandwidth network.

[0146] S1108, Detect whether the first connection end receives a fault response instruction sent by the second connection end within a first preset time period after sending the fault detection instruction; the fault response instruction is generated in response to the fault detection instruction, and the fault response instruction includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network; if yes, then execute S1110; if no, then execute S1112.

[0147] S1110, Determine that the fault detection result is that the current connection status of the link to be detected is in a normal state; Execute S1118 or S1128;

[0148] S1112, determine that the fault detection result is that the current connection status of the link to be detected is not in a normal state; repeat S1104 until the fault detection result is determined to be that the current connection status of the link to be detected is in a normal state;

[0149] S1114, check whether the current connection status of the link to be probed is in the probe state; not in the normal state includes abnormal state and state of not receiving results; if yes, then execute S1116; if no, then return to execute S1104.

[0150] S1116, detect whether a fault response command sent by the second connection end is received within the second preset time period; if yes, execute S1110; if no, return to execute S1104.

[0151] S1118, if the fault detection result is that the current connection status of the link to be detected is always in a normal state, then control the first connection end to send a reservation connection instruction to the second connection end; the reservation connection instruction carries the reservation connection information corresponding to the link to be detected, and the reservation connection instruction includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network.

[0152] S1120, Control the second connection end to register the reservation connection information on the second connection end according to the reservation connection instruction;

[0153] S1122, Update the priority of the link to be probed in the list of links to be connected according to the reservation connection information, and obtain the target priority of the link to be probed in the list of links to be connected;

[0154] S1124, according to the target priority, control the second connection end to send a connection response instruction to the first connection end at a preset rate; the preset rate is determined based on the processing efficiency of the subnet manager and the number of links to be connected; the connection response instruction carries the invitation connection information corresponding to the link to be probed, and the connection response instruction includes the management datagram corresponding to the subnet management interface in the unlimited bandwidth network;

[0155] S1126, Based on the invitation connection information, control the first connection end to establish a connection with the second connection end;

[0156] S1128, if the fault detection result is that the current connection status of the link to be detected has switched from an abnormal state to a normal state, then control the second connection end to send a connection notification instruction to the first connection end; the connection notification instruction includes the management datagram corresponding to the subnet management interface in the unlimited bandwidth network;

[0157] S1130, according to the connection notification instruction, control the first connection end to establish a connection with the second connection end.

[0158] In the aforementioned network management method, since the fault detection command of this application includes the management data packet corresponding to the subnet management interface, it is possible to directly control the first connection end of the link to be detected to send the fault detection command to the second connection end of the link to be detected. This avoids the need to send ICMP packets or path query requests to the subnet manager, thus preventing the excessive load on the subnet manager caused by link fault detection through the subnet manager in traditional methods, which could lead to IB network anomalies or even IB network paralysis. Subsequently, fault detection is directly performed on the link to be detected based on the fault detection command to generate fault detection results, and network management can be performed on the unlimited bandwidth network based on the fault detection results, thereby improving the security of the IB network.

[0159] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0160] Based on the same inventive concept, this application also provides a network management device for implementing the network management method described above. The solution provided by this device is similar to the implementation described in the above method; therefore, the specific limitations in one or more network management device embodiments provided below can be found in the limitations of the network management method described above, and will not be repeated here.

[0161] In one embodiment, such as Figure 12 As shown, a network management device 1200 is provided, including: a detection module 1220, a transmission module 1240, and a network management module 1260, wherein:

[0162] The detection module 1220 is used to detect whether the initial connection status of the connection link to be detected in the unlimited bandwidth network is in a normal state; the connection link to be detected includes a first connection end and a second connection end.

[0163] The sending module 1240 is used to control the first connection end of the connection to be detected to send a fault detection command to the second connection end of the connection to be detected if the initial connection state of the connection to be detected is in a normal state; the fault detection command includes a management data packet corresponding to the subnet management interface in the unlimited bandwidth network.

[0164] The network management module 1260 is used to perform fault detection on the connection link to be detected according to the fault detection command, generate fault detection results, and perform network management on the unlimited bandwidth network based on the fault detection results.

[0165] In one embodiment, the sending module 1240 includes:

[0166] The acquisition unit is used to acquire the management data packets corresponding to the subnet management interface and generate fault detection instructions based on the management data packets corresponding to the subnet management interface.

[0167] The sending unit is used to control the first connection end to send a fault detection command to the second connection end through a local identifier.

[0168] In one embodiment, the network management module 1260 includes:

[0169] The detection unit is used to detect whether the first connection end receives a fault response instruction sent by the second connection end within a first preset time period after sending the fault detection instruction; the fault response instruction is generated in response to the fault detection instruction, and the fault response instruction includes a management data packet corresponding to the subnet management interface in the unlimited bandwidth network;

[0170] The determining unit is configured to determine that the current connection status of the detected connection link is normal if the first connection end receives a fault response instruction sent by the second connection end within a first preset time period.

[0171] In one embodiment, the network management device 1200 further includes:

[0172] The normal state determination module is used to determine that the current connection state of the connection link to be detected is not in a normal state if the first connection end does not receive a fault response instruction sent by the second connection end within a first preset time period.

[0173] The first loop module is used to repeatedly execute the command to send a fault detection instruction from the first connection end of the connection to be detected to the second connection end of the connection to be detected; and to detect whether the first connection end receives a fault response instruction from the second connection end within a first preset time period after sending the fault detection instruction, until it is determined that the fault detection result is that the current connection status of the connection to be detected is normal.

[0174] In one embodiment, the network management device 1200 further includes:

[0175] The probe status detection module is used to detect whether the current connection status of the link to be probed is in the probe state if the initial connection status of the link to be probed is not in the normal state; not in the normal state includes abnormal state and state of not receiving results;

[0176] The fault response command detection module is used to detect whether a fault response command sent by the second connection end is received within a second preset time period if the current connection status of the connection link to be detected is in the detection state.

[0177] The second loop module is used to return to the execution control if the current connection state of the connection to be detected is not in the detection state, and to send a fault detection command from the first connection end of the connection to be detected to the second connection end of the connection to be detected; to perform fault detection on the connection to be detected according to the fault detection command, generate fault detection results, and perform network management on the unlimited bandwidth network based on the fault detection results.

[0178] In one embodiment, the network management device 1200 further includes:

[0179] The reservation connection instruction sending module is used to control the first connection end to send a reservation connection instruction to the second connection end if the fault detection result indicates that the current connection status of the connection link to be detected is in a normal state. The reservation connection instruction carries the reservation connection information corresponding to the connection link to be detected, and includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network.

[0180] The connection response instruction sending module is used to control the second connection end to send a connection response instruction to the first connection end according to the reservation connection instruction; the connection response instruction carries the invitation connection information corresponding to the connection link to be probed, and the connection response instruction includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network;

[0181] The first connection establishment module is used to control the first connection end to establish a connection with the second connection end based on the invitation connection information.

[0182] In one embodiment, the connection response instruction sending module includes:

[0183] The registration unit is used to control the second connection end to register the reservation connection information on the second connection end according to the reservation connection instruction.

[0184] The target priority determination unit is used to update the priority of the link to be detected in the list of links to be connected according to the reservation connection information, so as to obtain the target priority of the link to be detected in the list of links to be connected.

[0185] The connection response instruction sending unit is used to control the second connection end to send a connection response instruction to the first connection end at a preset rate according to the target priority; the preset rate is determined based on the processing efficiency of the subnet manager and the number of links to be connected.

[0186] In one embodiment, the network management device 1200 further includes:

[0187] The connection notification instruction sending module is used to control the second connection end to send a connection notification instruction to the first connection end if the fault detection result shows that the current connection status of the connection link to be detected has switched from an abnormal state to a normal state; the connection notification instruction includes a management datagram corresponding to the subnet management interface in the unlimited bandwidth network;

[0188] The second connection establishment module is used to control the first connection end and the second connection end to establish a connection according to the connection notification instruction.

[0189] In one embodiment, a network management device is also provided, comprising: a reservation connection instruction sending module, a connection response instruction sending module, and a first connection establishment module, wherein:

[0190] The reservation connection instruction sending module is used to control the first connection end to send a reservation connection instruction to the second connection end if the fault detection result indicates that the current connection status of the link to be detected is normal. The reservation connection instruction carries the reservation connection information corresponding to the link to be detected, and includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network. The fault detection result is the result of fault detection on the link to be detected in the unlimited bandwidth network.

[0191] The connection response instruction sending module is used to control the second connection end to send a connection response instruction to the first connection end according to the reservation connection instruction; the connection response instruction carries the invitation connection information corresponding to the connection link to be probed, and the connection response instruction includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network;

[0192] The first connection establishment module is used to control the first connection end to establish a connection with the second connection end based on the invitation connection information.

[0193] Each module in the aforementioned network management device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0194] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 13As shown, this computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores network management data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a network management method.

[0195] Those skilled in the art will understand that Figure 13 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0196] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0197] The initial connection status of the link to be probed in the unlimited bandwidth network is detected to ensure it is in a normal state; the link to be probed includes a first connection end and a second connection end.

[0198] If the initial connection state of the link to be probed is normal, the first connection end of the link to be probed is controlled to send a fault detection command to the second connection end of the link to be probed; the fault detection command includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network;

[0199] Based on the fault detection instructions, fault detection is performed on the connection links to be detected, fault detection results are generated, and network management is performed on the unlimited bandwidth network based on the fault detection results.

[0200] In one embodiment, the first connection end of the connection to be detected sends a fault detection command to the second connection end of the connection to be detected, and the processor, when executing the computer program, also performs the following steps:

[0201] Obtain the management data packet corresponding to the subnet management interface, and generate a fault detection command based on the management data packet corresponding to the subnet management interface;

[0202] The first connection end sends fault detection commands to the second connection end through a local identifier.

[0203] In one embodiment, fault detection is performed on the connection link to be detected according to the fault detection instruction, and fault detection results are generated. When the processor executes the computer program, the following steps are also implemented:

[0204] The system detects whether the first connection end receives a fault response command sent by the second connection end within a first preset time period after sending the fault detection command; the fault response command is generated in response to the fault detection command and includes a management datagram corresponding to the subnet management interface in the unlimited bandwidth network.

[0205] If the first connection end receives a fault response command sent by the second connection end within the first preset time period, the fault detection result is determined to be that the current connection status of the connection link to be detected is in a normal state.

[0206] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0207] If the first connection end does not receive a fault response instruction sent by the second connection end within the first preset time period, the fault detection result is determined to be that the current connection status of the connection link to be detected is not in a normal state.

[0208] The system continuously executes a cycle to send a fault detection command from the first connection end of the link to be detected to the second connection end of the link to be detected; it then checks whether the first connection end receives a fault response command from the second connection end within a first preset time period after sending the fault detection command, until it is determined that the fault detection result indicates that the current connection status of the link to be detected is normal.

[0209] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0210] If the initial connection state of the link to be probed is not in a normal state, then check whether the current connection state of the link to be probed is in a probe state; not in a normal state includes abnormal state and state where no result has been received.

[0211] If the current connection status of the link to be probed is in the probe state, then detect whether a fault response command sent by the second connection end is received within the second preset time period.

[0212] If the current connection state of the link to be probed is not in the probe state, the execution control will return to send a fault detection command from the first connection end of the link to be probed to the second connection end of the link to be probed; according to the fault detection command, the link to be probed will be probed for faults, generate fault detection results, and perform network management on the unlimited bandwidth network based on the fault detection results.

[0213] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0214] If the fault detection result indicates that the current connection status of the link to be detected is normal, then the first connection end is controlled to send a reservation connection instruction to the second connection end. The reservation connection instruction carries the reservation connection information corresponding to the link to be detected, and includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network.

[0215] The control second connection end sends a connection response instruction to the first connection end according to the reservation connection instruction; the connection response instruction carries the invitation connection information corresponding to the connection link to be probed, and the connection response instruction includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network;

[0216] Based on the invitation connection information, control the first connection end to establish a connection with the second connection end.

[0217] In one embodiment, the second connection end is controlled to send a connection response instruction to the first connection end according to the reserved connection instruction. When the processor executes the computer program, it also performs the following steps:

[0218] The control unit registers the reserved connection information on the second connection end according to the reserved connection instruction.

[0219] Update the priority of the link to be probed in the list of links to be connected based on the reservation connection information, and obtain the target priority of the link to be probed in the list of links to be connected;

[0220] Based on the target priority, the second connection end is controlled to send a connection response command to the first connection end at a preset rate; the preset rate is determined based on the processing efficiency of the subnet manager and the number of links to be connected.

[0221] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0222] If the fault detection result indicates that the current connection status of the link to be detected has switched from an abnormal state to a normal state, then the second connection end is controlled to send a connection notification instruction to the first connection end; the connection notification instruction includes the management datagram corresponding to the subnet management interface in the unlimited bandwidth network;

[0223] According to the connection notification instruction, control the first connection end to establish a connection with the second connection end.

[0224] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0225] If the fault detection result indicates that the current connection status of the link to be detected is normal, then the first connection end is controlled to send a reservation connection instruction to the second connection end. The reservation connection instruction carries the reservation connection information corresponding to the link to be detected, and the reservation connection instruction includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network. The fault detection result is the result of fault detection on the link to be detected in the unlimited bandwidth network.

[0226] The control second connection end sends a connection response instruction to the first connection end according to the reservation connection instruction; the connection response instruction carries the invitation connection information corresponding to the connection link to be probed, and the connection response instruction includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network;

[0227] Based on the invitation connection information, control the first connection end to establish a connection with the second connection end.

[0228] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0229] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0230] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0231] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0232] The above embodiments are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A network management method, characterized in that, The method, applicable to any computer device in a cluster that is communicatively connected to a subnet manager, includes: The initial connection status of the link to be probed in the unlimited bandwidth network is detected to determine whether it is in a normal state; the link to be probed includes a first connection end and a second connection end; If the initial connection state of the link to be probed is normal, then the first connection end of the link to be probed is controlled to send a fault detection command to the second connection end of the link to be probed; the fault detection command includes a management data packet corresponding to the subnet management interface in the unlimited bandwidth network; The fault detection command is used to detect faults in the connection link to be detected, and a fault detection result is generated. Based on the fault detection result, network management is performed on the unlimited bandwidth network. The network management process includes link fault detection of the unlimited bandwidth network and connection management of the unlimited bandwidth network. The network management of the unlimited bandwidth network based on the fault detection results includes: If the fault detection result indicates that the current connection status of the connection link to be detected is normal within a preset time period, then the first connection end of the connection link to be detected is controlled to establish a connection with the second connection end of the connection link to be detected through a scheduled connection command. If the fault detection result indicates that the current connection status of the link to be detected has switched from an abnormal state to a normal state, then the second connection end in the link to be detected will actively notify the first connection end in the link to be detected to establish a connection; the second connection end has stored the link information of the link to be detected before the link to be detected experienced a fault. If the fault detection result indicates that the current connection status of the link to be detected is not in a normal state, then the invalid connection information in the link to be detected is cleared.

2. The method according to claim 1, characterized in that, The step of controlling the first connection end of the connection to be detected to send a fault detection command to the second connection end of the connection to be detected includes: Obtain the management data packet corresponding to the subnet management interface, and generate the fault detection command based on the management data packet corresponding to the subnet management interface; The fault detection command is sent from the first connection end to the second connection end by controlling the local identifier.

3. The method according to claim 1 or 2, characterized in that, The step of performing fault detection on the connection link to be detected according to the fault detection command and generating fault detection results includes: The system detects whether the first connection end receives a fault response instruction sent by the second connection end within a first preset time period after sending the fault detection instruction; the fault response instruction is generated in response to the fault detection instruction, and the fault response instruction includes a management data packet corresponding to the subnet management interface in the unlimited bandwidth network. If the first connection end receives a fault response command sent by the second connection end within a first preset time period, then the fault detection result is determined to be that the current connection status of the connection link to be detected is in a normal state.

4. The method according to claim 3, characterized in that, The method further includes: If the first connection end does not receive a fault response instruction sent by the second connection end within a first preset time period, then the fault detection result is determined to be that the current connection status of the connection link to be detected is not in a normal state. The process involves repeatedly sending a fault detection command from the first connection end of the link to be detected to the second connection end of the link to be detected; detecting whether the first connection end receives a fault response command from the second connection end within a first preset time period after sending the fault detection command, until it is determined that the fault detection result indicates that the current connection status of the link to be detected is normal.

5. The method according to claim 1, characterized in that, The method further includes: If the initial connection state of the link to be probed is not in a normal state, then it is detected whether the current connection state of the link to be probed is in a probe state; the state not in a normal state includes abnormal state and state of not receiving results; If the current connection state of the connection to be detected is in the detection state, then it is detected whether a fault response command sent by the second connection end is received within the second preset time period. If the current connection state of the link to be probed is not in the probe state, then return to the execution of sending a fault detection command from the first connection end of the link to be probed to the second connection end of the link to be probed; perform fault detection on the link to be probed according to the fault detection command, generate fault detection results, and perform network management on the unlimited bandwidth network based on the fault detection results.

6. The method according to claim 3, characterized in that, The method further includes: If the fault detection result indicates that the current connection status of the connection link to be detected is normal, then the first connection end is controlled to send a reservation connection instruction to the second connection end; the reservation connection instruction carries the reservation connection information corresponding to the connection link to be detected, and the reservation connection instruction includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network; The second connection terminal is controlled to send a connection response instruction to the first connection terminal according to the reservation connection instruction; the connection response instruction carries the invitation connection information corresponding to the connection link to be probed, and the connection response instruction includes the management data packet corresponding to the subnet management interface in the unlimited bandwidth network; Based on the invitation connection information, control the first connection end to establish a connection with the second connection end.

7. The method according to claim 6, characterized in that, The control of the second connection end to send a connection response instruction to the first connection end according to the reserved connection instruction includes: The second connection terminal is controlled to register the reservation connection information on the second connection terminal according to the reservation connection instruction; Update the priority of the link to be probed in the list of links to be connected based on the reservation connection information, and obtain the target priority of the link to be probed in the list of links to be connected; Based on the target priority, the second connection end is controlled to send a connection response command to the first connection end at a preset rate; the preset rate is determined based on the processing efficiency of the subnet manager and the number of links to be connected.

8. The method according to claim 6, characterized in that, The method further includes: If the fault detection result indicates that the current connection status of the link to be detected has switched from an abnormal state to a normal state, then the second connection end is controlled to send a connection notification instruction to the first connection end; the connection notification instruction includes a management datagram corresponding to the subnet management interface in the unlimited bandwidth network; According to the connection notification instruction, control the first connection end to establish a connection with the second connection end.

9. A network management device, characterized in that, The apparatus is applicable to any computer device in a cluster that is communicatively connected to a subnet manager, and includes: The detection module is used to detect whether the initial connection status of the link to be probed in the unlimited bandwidth network is in a normal state; the link to be probed includes a first connection end and a second connection end; The sending module is configured to, if the initial connection state of the connection to be probed is in a normal state, control the first connection end of the connection to be probed to send a fault detection command to the second connection end of the connection to be probed; the fault detection command includes a management data packet corresponding to the subnet management interface in the unlimited bandwidth network; The network management module is used to perform fault detection on the connection link to be detected according to the fault detection instruction, generate fault detection results, and perform network management on the unlimited bandwidth network based on the fault detection results; the network management process includes link fault detection of the unlimited bandwidth network and connection management of the unlimited bandwidth network. The network management module includes: if the fault detection result indicates that the current connection status of the link to be detected is normal within a preset time period, then controlling the first connection end of the link to be detected to establish a connection with the second connection end of the link to be detected via a scheduled connection command; if the fault detection result indicates that the current connection status of the link to be detected switches from an abnormal state to a normal state, then the second connection end in the link to be detected actively notifies the first connection end in the link to be detected to establish a connection; the second connection end stores the link information of the link to be detected before the fault occurs; if the fault detection result indicates that the current connection status of the link to be detected is not normal, then clearing the invalid connection information in the link to be detected.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 8.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 8.

12. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 8.

Images