Computer fault monitoring method and system based on Internet

Abstract

The invention provides a computer fault monitoring method and system based on the Internet, and belongs to the technical field of fault monitoring, and the method comprises the steps: sending a current load state calling command to a to-be-monitored computer, obtaining a current load proportion, and sending a high-load task instruction to the to-be-monitored computer; sending a current load state calling command to the to-be-monitored computer according to equal time intervals, acquiring current load state data of the to-be-monitored computer, endowing time sequence information, and fitting a load change curve; determining a load peak value ratio time, and determining the number of time periods of the load change of the maximum peak-peak value and a time interval between the time periods; and comparing the number of the time periods and the maximum time interval with a preset fluctuation abnormity judgment threshold value, and determining whether a fault occurs or not. According to the method, fault monitoring of the computer cluster can be efficiently carried out through the Internet, the current state and the operation level of the computer are obtained in time, and the overall performance requirement of the computer cluster is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of fault monitoring systems, and in particular to a computer fault monitoring method and system based on the Internet. Background Art

[0002] Computer fault monitoring methods refer to the use of different technical means and tools to monitor the operation of computer systems in real time or regularly in order to detect, identify and respond to possible faults or abnormalities in a timely manner. These methods are designed to improve the reliability, stability and performance of the system, while reducing downtime and production losses caused by faults. Common computer fault monitoring includes performance monitoring and resource utilization monitoring, such as monitoring the CPU usage of the computer to identify whether there is an overload or abnormality; tracking the use of computer memory to ensure that there is no memory leak or insufficient problem; monitoring the use of disk space to prevent disk overflow or low space problems; monitoring network traffic and bandwidth usage to detect network bottlenecks or abnormal traffic. Secondly, for the computer's own hardware, once a fault occurs, it greatly affects the processing efficiency of the task or process.

[0003] In the prior art, computer fault monitoring is performed using built-in software to determine whether the current computer has a fault based on the real-time performance data and resource changes of the current system. Therefore, the existing fault monitoring usually only performs fault determination when a problem occurs during the operation of the computer. However, in the face of a multi-computer cluster, fault monitoring needs to be performed regularly to ensure the overall operating performance requirements. A single computer local monitoring method cannot achieve efficient monitoring purposes. Summary of the invention

[0004] To solve the above problems, the present invention provides an Internet-based computer fault monitoring method and system, which can efficiently monitor the faults of computer clusters through the Internet, timely obtain the current status and operating level of the computer, and ensure the overall performance requirements of the computer cluster.

[0005] To achieve the above object, the present invention provides the following technical solutions.

[0006] A computer fault monitoring method based on the Internet comprises the following steps:

[0007] Multiple computers to be monitored call fault monitoring instructions to authorize access requests from the monitoring end;

[0008] Send a current load status call command to the computer to be monitored via the Internet to obtain the computer's current load status data and determine the current load ratio;

[0009] According to the current load ratio and the length of time that the computer allows for a short-term load peak, determine the type of high-load task instruction and adjust the actual number of processes of the high-load task instruction; send the high-load task instruction to the computer to be monitored and request execution to increase the current load of the computer;

[0010] Sending a current load status call command to the computer to be monitored at equal time intervals, obtaining the current load status data of the computer to be monitored, assigning timing information, and fitting a load change curve;

[0011] Extract the characteristic segment of the current load change curve, obtain the peak duration segment of the load change curve, and the time period when the maximum peak-to-peak value and the maximum peak-to-peak value are generated; determine the load peak proportion time according to the peak duration segment of the load change curve; obtain the time period when all load changes higher than the maximum peak-to-peak value under the preset ratio are generated according to the time period when the maximum peak-to-peak value and the maximum peak-to-peak value are generated, and determine the number of each time period and the time interval between each time period;

[0012] The number of time periods and the maximum time interval are compared with the preset fluctuation abnormality judgment threshold, and the fluctuation state of the load change curve is obtained;

[0013] Determine whether a fault occurs based on the request receiving status, load peak ratio time, and load change curve fluctuation status of the current monitored computer.

[0014] Preferably, the access request of the authorization monitoring terminal comprises the following steps:

[0015] Configure the access control list that allows access to the network device configured by the computer to be monitored, and configure the relevant information on the monitoring device, including the source IP address, target IP address, specific port number and protocol;

[0016] Set the fault monitoring instruction. When the computer to be monitored calls the fault diagnosis instruction, open the specific port to allow the access request of the monitoring end.

[0017] Preferably, determining the high-load task instruction type according to the current load ratio and the duration of the short-term load peak allowed by the computer comprises the following steps:

[0018] Determine the number of processes allowed to execute tasks and the length of time allowed for short-term load peaks of the computer to be monitored based on the hardware configuration of the computer to be monitored and the expected load bearing capacity; the hardware configuration includes the number of CPU cores and memory capacity;

[0019] Randomly generate high-load task instructions, test the number of processes required to obtain each high-load task instruction and the duration of the short-term load peak under the same hardware configuration;

[0020] Select high-load task instructions based on the current load percentage, the number of processes allowed to execute tasks, and the length of time the computer allows short-term load peaks.

[0021] Preferably, the high-load task instructions include scientific computing tasks and big data processing tasks.

[0022] Preferably, the extracting of the characteristic segment of the current load change curve, obtaining the peak duration segment of the load change curve, and the time period of the maximum peak-to-peak value and the maximum peak-to-peak value, comprises the following steps:

[0023] Obtain the load status data of the computer to be monitored in equal time periods, assign timing information, and draw a time series diagram of the load change curve based on the load status data;

[0024] According to the preset maximum peak value threshold of the load status data, the initial point and the end point of the maximum peak segment in the load change curve are searched along the time series, the peak duration segments of multiple load change curves are determined, and the load peak proportion time is calculated;

[0025] Determine a sliding window of fixed length, check each data point and the load state data points around it along the time series, to find out the load state data point that is lower than all the surrounding load state data points, and determine the duration of the load state data point reaching the initial point of the maximum peak segment, and then determine the time period in which the maximum peak-to-peak value and the maximum peak-to-peak value are generated;

[0026] The time periods in which the maximum peak-to-peak load change occurs under a preset ratio are obtained, and the number of time periods and the time intervals between the time periods are determined.

[0027] Preferably, the time period for obtaining the maximum peak-to-peak load change under the preset ratio is a load change time period of 70% of all the maximum peak-to-peak values.

[0028] Preferably, determining whether a fault occurs according to the request receiving state, load peak proportion time and load change curve fluctuation state of the current computer to be monitored includes the following steps:

[0029] Compare the number of time periods and the maximum time interval with the preset abnormal fluctuation judgment threshold, and thereby obtain the fluctuation state of the load change curve; wherein the abnormal fluctuation judgment threshold includes the number of time periods threshold and the time interval threshold between time periods;

[0030] When the computer to be monitored refuses to receive a request, or the load peak ratio is higher than a preset value, or exceeds a fluctuation abnormality judgment threshold, the computer to be monitored generates a fault or an abnormality.

[0031] A computer fault monitoring system based on the Internet, comprising:

[0032] An access request module is used to send diagnostic instructions to multiple computers to be monitored, and multiple computers to be monitored call fault monitoring instructions to authorize access requests from the monitoring end;

[0033] The load status monitoring module is used to send a current load status call command to the computer to be monitored, obtain the current load status data of the computer, and determine the current load ratio;

[0034] The load loading module is used to determine the type of high-load task instruction according to the current load ratio and the length of time that the computer allows for a short-term load peak, adjust the actual number of processes of the high-load task instruction, send the high-load task instruction to the computer to be monitored, and request execution, thereby increasing the current load of the computer;

[0035] A load change curve fitting module is used to assign timing information to the current load state data of the computer to be monitored at equal time intervals obtained by the load state monitoring module, and to obtain a load change curve by fitting;

[0036] The load feature extraction module is used to extract the feature segment of the current load change curve, obtain the peak duration segment of the load change curve, and the time period when the maximum peak-to-peak value and the maximum peak-to-peak value are generated; it is also used to determine the load peak proportion time according to the peak duration segment of the load change curve; it is also used to obtain the time period when all load changes higher than the maximum peak-to-peak value under a preset ratio are generated according to the time period when the maximum peak-to-peak value and the maximum peak-to-peak value are generated, and determine the number of each time period and the time interval between each time period;

[0037] The fault diagnosis module is used to compare the number of time periods and the maximum time interval with the preset fluctuation abnormality judgment threshold, and thereby obtain the fluctuation state of the load change curve; it is also used to determine whether a fault occurs based on the request reception state of the current monitored computer, the load peak proportion time and the fluctuation state of the load change curve obtained by the access request module.

[0038] Beneficial effects of the present invention:

[0039] The present invention proposes a computer fault monitoring method and system based on the Internet. The method sends a load state call instruction and a high-load task instruction to a computer cluster to be monitored through the Internet, obtains the load state change of the current computer running the high-load task instruction, and judges whether the monitoring target has a fault according to the load change peak duration period and the relevant characteristics of the maximum peak-peak value and the maximum peak-peak value generation time period that characterize the load task processing state, and compares with a preset threshold. The method accurately judges the current state of the computer by obtaining the state characteristics of the computer processing high-load tasks, realizes remote and efficient fault monitoring of the computer cluster, has no restrictions on the monitoring time and location, can achieve early warning and fault identification, and then timely handle the fault to ensure the stable operation of the computer cluster. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 is a specific flow chart of an embodiment of the present invention;

[0041] Figure 2 is a flow chart of the access request authorization steps of an embodiment of the present invention;

[0042] Figure 3 is a flow chart of the steps of determining and adjusting high-load task instructions according to an embodiment of the present invention;

[0043] Figure 4 is a flow chart of a feature acquisition step in an embodiment of the present invention;

[0044] Figure 5 It is an architecture diagram of an Internet-based computer fault monitoring system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0046] Example 1

[0047] In the prior art, computer fault monitoring is performed using built-in software to determine whether the current computer has a fault based on the real-time performance data and resource changes of the current system. Therefore, the existing fault monitoring usually performs fault determination when a problem occurs during the operation of the computer. However, in the face of a multi-computer cluster, fault monitoring needs to be performed regularly to ensure the overall operating performance requirements. A single local computer monitoring method cannot achieve efficient monitoring purposes. To this end, this embodiment proposes a computer fault monitoring method based on the Internet, such as Figure 1 As shown, Figure 1The flowchart of the computer fault monitoring method based on the Internet specifically includes the following steps:

[0048] S1: Multiple computers to be monitored call fault monitoring instructions to authorize access requests from the monitoring end.

[0049] S2: Send a current load status call command to the computer to be monitored via the Internet, obtain the computer's current load status data, and determine the current load ratio.

[0050] S3: Determine the type of high-load task instruction based on the current load ratio and the length of time the computer allows for a short-term load peak, and adjust the actual number of processes of the high-load task instruction; send the high-load task instruction to the computer to be monitored and request execution to increase the current load of the computer.

[0051] S4: Sending a current load status call command to the computer to be monitored at equal time intervals, obtaining the current load status data of the computer to be monitored, assigning timing information, and fitting a load change curve.

[0052] S5: Extract the characteristic segment of the current load change curve, obtain the peak duration segment of the load change curve, and the time period when the maximum peak-to-peak value and the maximum peak-to-peak value are generated; determine the load peak proportion time according to the peak duration segment of the load change curve; obtain the time period when all load changes higher than the maximum peak-to-peak value under the preset ratio are generated according to the time period when the maximum peak-to-peak value and the maximum peak-to-peak value are generated, and determine the number of each time period and the time interval between each time period.

[0053] S6: Compare the number of time periods and the maximum time interval with a preset fluctuation abnormality judgment threshold, and thereby obtain the fluctuation state of the load change curve.

[0054] S7: Determine whether a fault occurs according to the request receiving state, load peak ratio time, and load change curve fluctuation state of the current monitored computer.

[0055] Furthermore, in the access request authorization step in S1, if Figure 2 As shown, the specific steps include:

[0056] S1.1: Configure the access control list that allows access to the network device configured by the monitored computer, and configure the relevant information on the monitoring device, including the source IP address, target IP address, specific port number and protocol.

[0057] S1.2: Set the fault monitoring instruction. When the computer to be monitored calls the fault diagnosis instruction, open the specific port to allow the access request of the monitoring end.

[0058] Among them, the access control list (ACL) is a network security control mechanism used to limit or allow devices or users to access resources. ACL is usually configured based on conditions such as source IP address, destination IP address, port number, protocol, etc. to determine which requests can pass. By configuring ACL on routers, firewalls, or network devices, it is possible to control whether requests from specific devices can pass through the network.

[0059] Specifically, the load status calling command in S2 may be:

[0060] Use the tasklist command to view a list of currently running processes and their detailed information.

[0061] Use the wmic command to query various system information, including CPU and memory usage.

[0062] For example: wmic cpu get loadpercentage gets the CPU usage.

[0063] For example: wmic OS get FreePhysicalMemory gets the available physical memory.

[0064] Furthermore, the high-load task instructions in S3 are determined and adjusted, such as Figure 3 As shown, the following steps are included:

[0065] S3.1: Determine the number of processes allowed to execute tasks and the length of time allowed for short-term load peaks of the computer to be monitored based on the hardware configuration of the computer to be monitored and the expected load bearing capacity; the hardware configuration includes the number of CPU cores and memory capacity.

[0066] S3.2: Randomly generate high-load task instructions, test the number of processes required to obtain each high-load task instruction and the duration of the short-term load peak under the same hardware configuration.

[0067] S3.3: Select high-load task instructions based on the current load ratio, the number of processes allowed to execute tasks, and the length of time the computer allows for short-term load peaks.

[0068] Further, high-load task instructions include:

[0069] Scientific computing tasks:

[0070] Numerical computing: involves a large number of mathematical operations, such as matrix calculations, solving differential equations, etc. Tools and libraries: usually use Python's scientific computing libraries (such as NumPy, SciPy), MATLAB, R language, etc.

[0071] Big data processing tasks:

[0072] Data Analysis and Processing: Analyze, aggregate, clean, and perform other operations on large-scale data sets. Tools and frameworks: Big data processing frameworks such as Hadoop and Spark; Python's Pandas, Dask, and other data processing libraries. Deep learning model training tasks: Artificial intelligence and machine learning (AI / ML): Train complex deep learning models, such as neural networks. Frameworks and libraries: Deep learning frameworks such as TensorFlow and PyTorch; CUDA / GPU programming (such as using CUDA C++). Parallel computing tasks: Parallel programs and distributed computing: Use multiple processing units or multiple computing nodes to perform parallel processing of tasks. Tools and libraries: MPI (Message Passing Interface) programming, OpenMP, CUDA, and other parallel computing tools and frameworks. Computation-intensive tasks: High Performance Computing (HPC): Tasks that require a large amount of CPU computing resources to process. Tools and frameworks: Usually written in programming languages ​​such as C / C++ and Fortran; use optimized algorithms and data structures.

[0073] Furthermore, the curve fitting in S4 and the feature acquisition in S5 are as follows. Figure 4 As shown, the following steps are included:

[0074] S5.1: Obtain the load status data of the computer to be monitored in equal time periods, assign timing information, and draw a time series diagram of the load change curve according to the load status data, so as to intuitively observe the change trend of the load.

[0075] S5.2: According to the preset maximum peak threshold of the load status data, search for the initial point and the end point of the maximum peak segment in the load change curve along the time series, determine the peak duration segments of multiple load change curves, and calculate the load peak proportion time.

[0076] S5.3: Determine a sliding window of fixed length, and check each data point and its surrounding load status data points along the time series to find a load status data point that is lower than all surrounding load status data points, and determine the duration of the load status data point reaching the initial point of the maximum peak segment, thereby determining the time period in which the maximum peak-to-peak value and the maximum peak-to-peak value are generated.

[0077] S5.4: Obtain the time period in which the maximum peak-to-peak load change occurs under a preset ratio, and determine the number of time periods and the time interval between time periods. The preset ratio in this embodiment is 70%.

[0078] The method proposed in this embodiment sends a load state call instruction and a high-load task instruction to the computer cluster to be monitored through the Internet, obtains the load state change of the current computer running the high-load task instruction, and judges whether the monitoring target has a fault according to the load change peak duration period and the relevant characteristics of the maximum peak-peak value and the maximum peak-peak value generation time period that characterize the load task processing state, and compares with the preset threshold. This method accurately judges the current state of the computer by obtaining the state characteristics of the computer processing high-load tasks, realizes remote and efficient fault monitoring of the computer cluster, has no restrictions on the monitoring time and location, can achieve early warning and fault identification, and then timely handle the fault to ensure the stable operation of the computer cluster.

[0079] The above is an Internet-based computer fault monitoring method provided by an embodiment of this embodiment. Based on the same idea, this embodiment also provides a corresponding Internet-based computer fault monitoring system, such as Figure 5 As shown, specifically including:

[0080] The access request module is used to send diagnostic instructions to multiple computers to be monitored, and multiple computers to be monitored call fault monitoring instructions to authorize access requests from the monitoring end.

[0081] The load status monitoring module is used to send a current load status call command to the computer to be monitored, obtain the computer's current load status data, and determine the current load ratio.

[0082] The load loading module is used to determine the type of high-load task instructions according to the current load ratio and the length of time the computer allows for short-term load peaks, adjust the actual number of processes of the high-load task instructions, send high-load task instructions to the computer to be monitored, and request execution to increase the current load of the computer.

[0083] The load change curve fitting module is used to assign timing information according to the current load state data of the monitored computer at equal time intervals obtained by the load state monitoring module, and to obtain the load change curve through fitting.

[0084] The load feature extraction module is used to extract the characteristic segment of the current load change curve, obtain the peak duration segment of the load change curve, and the time period when the maximum peak-to-peak value and the maximum peak-to-peak value are generated; it is also used to determine the load peak proportion time according to the peak duration segment of the load change curve; it is also used to obtain the time period when all load changes higher than the maximum peak-to-peak value under a preset ratio are generated according to the time period when the maximum peak-to-peak value and the maximum peak-to-peak value are generated, and determine the number of each time period and the time interval between each time period.

[0085] The fault diagnosis module is used to compare the number of time periods and the maximum time interval with the preset fluctuation abnormality judgment threshold, and thereby obtain the fluctuation state of the load change curve; it is also used to determine whether a fault occurs based on the request reception state of the current monitored computer, the load peak proportion time and the fluctuation state of the load change curve obtained by the access request module.

[0086] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A computer fault monitoring method based on the Internet, characterized in that: The following steps are involved: Multiple computers to be monitored call fault monitoring instructions to authorize access requests from the monitoring end; Send a current load status call command to the computer to be monitored via the Internet to obtain the computer's current load status data and determine the current load ratio; According to the current load ratio and the length of time that the computer allows for a short-term load peak, determine the type of high-load task instruction and adjust the actual number of processes of the high-load task instruction; send the high-load task instruction to the computer to be monitored and request execution to increase the current load of the computer; Sending a current load status call command to the computer to be monitored at equal time intervals to obtain the current load status data of the computer to be monitored, assigning timing information, and fitting a load change curve; Extract the characteristic segment of the current load change curve, obtain the peak duration segment of the load change curve, and the time period when the maximum peak-to-peak value and the maximum peak-to-peak value are generated; determine the load peak proportion time according to the peak duration segment of the load change curve; obtain the time period when all load changes higher than the maximum peak-to-peak value under the preset ratio are generated according to the time period when the maximum peak-to-peak value and the maximum peak-to-peak value are generated, and determine the number of each time period and the time interval between each time period; The number of time periods and the maximum time interval are compared with the preset fluctuation abnormality judgment threshold, and the fluctuation state of the load change curve is obtained; Determine whether a fault occurs based on the request receiving status, load peak ratio time, and load change curve fluctuation status of the current monitored computer.

2. The computer fault monitoring method based on the Internet according to claim 1, characterized in that: The access request of the authorization monitoring terminal includes the following steps: Configure the access control list that allows access to the network device configured by the computer to be monitored, and configure the relevant information on the monitoring device, including the source IP address, target IP address, specific port number and protocol; Set the fault monitoring instruction. When the computer to be monitored calls the fault diagnosis instruction, open the specific port to allow the access request of the monitoring end.

3. The computer fault monitoring method based on the Internet according to claim 1, characterized in that: Determining the high-load task instruction type according to the current load ratio and the length of time the computer allows for a short-term load peak includes the following steps: Determine the number of processes allowed to execute tasks and the length of time allowed for short-term load peaks of the computer to be monitored based on the hardware configuration of the computer to be monitored and the expected load bearing capacity; the hardware configuration includes the number of CPU cores and memory capacity; Randomly generate high-load task instructions, test the number of processes required to obtain each high-load task instruction and the duration of the short-term load peak under the same hardware configuration; Select high-load task instructions based on the current load percentage, the number of processes allowed to execute tasks, and the length of time the computer allows short-term load peaks.

4. The Internet-based computer fault monitoring method according to claim 3, characterized in that: The high-load task instructions include scientific computing tasks and big data processing tasks.

5. The computer fault monitoring method based on the Internet according to claim 1, characterized in that: The extracting of the characteristic segment of the current load change curve to obtain the peak duration segment of the load change curve and the time period in which the maximum peak-to-peak value and the maximum peak-to-peak value are generated comprises the following steps: Obtain the load status data of the computer to be monitored in equal time periods, assign timing information, and draw a time series diagram of the load change curve according to the load status data; According to the preset maximum peak value threshold of the load status data, the initial point and the end point of the maximum peak segment in the load change curve are searched along the time series, the peak duration segments of multiple load change curves are determined, and the load peak proportion time is calculated; Determine a sliding window of fixed length, check each data point and the load state data points around it along the time series, to find out the load state data point that is lower than all the surrounding load state data points, and determine the duration of the load state data point reaching the initial point of the maximum peak segment, and then determine the time period in which the maximum peak-to-peak value and the maximum peak-to-peak value are generated; The time periods in which the maximum peak-to-peak load change occurs under a preset ratio are obtained, and the number of time periods and the time intervals between the time periods are determined.

6. The computer fault monitoring method based on the Internet according to claim 5, characterized in that: The time period for obtaining the maximum peak-to-peak value load change under the preset ratio is a load change time period based on 70% of all the maximum peak-to-peak values.

7. The computer fault monitoring method based on the Internet according to claim 5, characterized in that: The method of determining whether a fault occurs according to the request receiving state, load peak proportion time and load change curve fluctuation state of the current computer to be monitored includes the following steps: Compare the number of time periods and the maximum time interval with the preset abnormal fluctuation judgment threshold, and thereby obtain the fluctuation state of the load change curve; wherein the abnormal fluctuation judgment threshold includes the number of time periods threshold and the time interval threshold between time periods; When the computer to be monitored refuses to receive a request, or the load peak ratio is higher than a preset value, or exceeds a fluctuation abnormality judgment threshold, the computer to be monitored generates a fault or an abnormality.

8. A computer fault monitoring system based on the Internet, characterized in that: include: An access request module is used to send diagnostic instructions to multiple computers to be monitored, and multiple computers to be monitored call fault monitoring instructions to authorize access requests from the monitoring end; The load status monitoring module is used to send a current load status call command to the computer to be monitored, obtain the current load status data of the computer, and determine the current load ratio; The load loading module is used to determine the type of high-load task instruction according to the current load ratio and the length of time that the computer allows for a short-term load peak, adjust the actual number of processes of the high-load task instruction, send the high-load task instruction to the computer to be monitored, and request execution, thereby increasing the current load of the computer; A load change curve fitting module is used to assign timing information to the current load state data of the monitored computer at equal time intervals obtained by the load state monitoring module, and to obtain a load change curve by fitting; The load feature extraction module is used to extract the feature segment of the current load change curve, obtain the peak duration segment of the load change curve, and the time period when the maximum peak-to-peak value and the maximum peak-to-peak value are generated; it is also used to determine the load peak proportion time according to the peak duration segment of the load change curve; it is also used to obtain the time period when all load changes higher than the maximum peak-to-peak value under a preset ratio are generated according to the time period when the maximum peak-to-peak value and the maximum peak-to-peak value are generated, and determine the number of each time period and the time interval between each time period; The fault diagnosis module is used to compare the number of time periods and the maximum time interval with the preset fluctuation abnormality judgment threshold, and thereby obtain the fluctuation state of the load change curve; it is also used to determine whether a fault occurs based on the request reception state of the current monitored computer, the load peak proportion time and the fluctuation state of the load change curve obtained by the access request module.

Images