CN112799932B - Method, electronic device, and storage medium for predicting health level of application - Google Patents
Method, electronic device, and storage medium for predicting health level of application Download PDFInfo
- Publication number
- CN112799932B CN112799932B CN202110329739.7A CN202110329739A CN112799932B CN 112799932 B CN112799932 B CN 112799932B CN 202110329739 A CN202110329739 A CN 202110329739A CN 112799932 B CN112799932 B CN 112799932B
- Authority
- CN
- China
- Prior art keywords
- values
- application
- current
- historical
- resources
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
- G06F11/3409—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment for performance assessment
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Abstract
Embodiments of the present disclosure relate to methods, electronic devices, and computer storage media for predicting a health level of an application, and relate to the field of information processing. According to the method, a plurality of current use values, a plurality of current initial values and a plurality of current limit values of the application in a plurality of time periods in a current period for a plurality of resources are obtained; determining a plurality of current discrete coefficient characteristic values of the application in a current period for a plurality of resources; determining a predicted abnormal characteristic value applied in a next period of the current period based on the machine learning model and the plurality of current discrete coefficient characteristic values; and determining a predicted health level of the application within a next period based on a predetermined association between the range of anomaly characteristic values and the health level and the predicted anomaly characteristic values. Therefore, the predicted value of the machine learning model can reflect the actual scene better.
Description
Technical Field
Embodiments of the present disclosure relate generally to the field of information processing, and more particularly, to methods, electronic devices, and computer storage media for predicting a health level of an application.
Background
A set of service system under the micro-service architecture consists of a plurality of applications, and the applications are mutually interacted due to the richness of the services. The complexity of the application processing function logic cannot be measured, and the external access amount cannot be predicted, so that the response of the application is quick and slow, and even an abnormal return is caused in some cases. This results in part of the service interaction not being successfully completed, which brings a poor experience to the service. Therefore, it is desirable to be able to accurately predict the health of an application, so as to provide early warning and correction in time.
Disclosure of Invention
A method, an electronic device, and a computer storage medium for predicting a health level of an application are provided, which can make a predicted value of a machine learning model more reflective of an actual scene.
According to a first aspect of the present disclosure, a method for predicting a health level of an application is provided. The method comprises the following steps: acquiring a plurality of current use values, a plurality of current initial values and a plurality of current limit values of the application for a plurality of resources in a plurality of time periods in a current period; determining a plurality of current discrete coefficient characteristic values of the application for a plurality of resources within a current period based on the plurality of current usage values, the plurality of current initial values, and the plurality of current limit values; determining a predicted abnormal feature value applied in a next period of the current period based on a machine learning model and a plurality of current discrete coefficient feature values, wherein the machine learning model is trained based on a plurality of historical discrete coefficient feature values applied in a plurality of first historical periods for a plurality of resources and a plurality of abnormal feature values applied in a plurality of second historical periods, each second historical period in the plurality of second historical periods is the next period corresponding to the first historical period, and the plurality of abnormal feature values are determined based on a plurality of abnormal access amounts and a plurality of total access amounts applied in the plurality of second historical periods; and determining a predicted health level of the application within a next period based on the predicted anomaly characteristic value and a predetermined association between the range of anomaly characteristic values and the health level.
According to a second aspect of the present disclosure, an electronic device is provided. The electronic device includes: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, the instructions being executable by the at least one processor to enable the at least one processor to perform the method according to the first aspect.
In a third aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored which, when executed by a processor, implements a method according to the first aspect of the present disclosure.
It should be understood that the statements in this section do not necessarily identify key or critical features of the embodiments of the present disclosure, nor do they limit the scope of the present disclosure. Other features of the present disclosure will become apparent from the following description.
Drawings
The above and other features, advantages and aspects of various embodiments of the present disclosure will become more apparent by referring to the following detailed description when taken in conjunction with the accompanying drawings. In the drawings, like or similar reference characters designate like or similar elements.
FIG. 1 is a schematic diagram of an information handling environment 100 according to an embodiment of the present disclosure.
Fig. 2 is a schematic diagram of a method 200 for predicting a health level of an application, in accordance with an embodiment of the present disclosure.
Fig. 3 is a schematic diagram of a method 300 for predicting a health level of an application, in accordance with an embodiment of the present disclosure.
Fig. 4 is a schematic diagram of a method 400 for determining a plurality of current discrete coefficient feature values according to an embodiment of the present disclosure.
Fig. 5 is a schematic diagram of a method 500 for training a machine learning model, in accordance with an embodiment of the present disclosure.
FIG. 6 is a block diagram of an electronic device used to implement a method for predicting a health level of an application of an embodiment of the present disclosure.
Detailed Description
Exemplary embodiments of the present disclosure are described below with reference to the accompanying drawings, in which various details of the embodiments of the disclosure are included to assist understanding, and which are to be considered as merely exemplary. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the present disclosure. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.
The term "include" and variations thereof as used herein is meant to be inclusive in an open-ended manner, i.e., "including but not limited to". Unless specifically stated otherwise, the term "or" means "and/or". The term "based on" means "based at least in part on". The terms "one example embodiment" and "one embodiment" mean "at least one example embodiment". The term "another embodiment" means "at least one additional embodiment". The terms "first," "second," and the like may refer to different or the same object. Other explicit and implicit definitions are also possible below.
As mentioned above, it is desirable to be able to accurately predict the health of an application, so that early warning and correction can be performed in time.
To address, at least in part, one or more of the above issues and other potential issues, an example embodiment of the present disclosure proposes a scheme for predicting a health level of an application. In this approach, a computing device obtains a plurality of current usage values, a plurality of current initial values, and a plurality of current limit values for a plurality of resources by an application over a plurality of time periods within a current period, and determines a plurality of current discrete coefficient feature values for the application over the current period for the plurality of resources based on the plurality of current usage values, the plurality of current initial values, and the plurality of current limit values. The computing device determines a predicted anomaly characteristic value for the application during a next period of the current period based on a machine learning model and a plurality of current discrete coefficient characteristic values, the machine learning model being trained based on a plurality of historical discrete coefficient characteristic values for the plurality of resources during a plurality of first historical periods and a plurality of anomaly characteristic values for the plurality of second historical periods, each of the plurality of second historical periods being the next period corresponding to the first historical period, the plurality of anomaly characteristic values being determined based on a plurality of anomaly visitors and a plurality of total visitors applied during the plurality of second historical periods. The computing device determines a predicted health level for the application within a next period based on a predetermined association between the range of anomalous feature values and the health level and the predicted anomalous feature values. In this way, the predicted value of the machine learning model can be made to reflect the actual scene better.
Hereinafter, specific examples of the present scheme will be described in more detail with reference to the accompanying drawings.
FIG. 1 shows a schematic diagram of an example of an information processing environment 100, according to an embodiment of the present disclosure. The information processing environment 100 may include a computing device 110, applications 120-1 through 120-n (hereinafter collectively referred to as applications 120), an application monitoring system 130, an application access log system 140, and a server 150 running the applications 120.
The computing device 110 includes, for example, but is not limited to, a server computer, a multiprocessor system, a mainframe computer, a distributed computing environment including any of the above systems or devices, and the like. In some embodiments, the computing device 110 may have one or more processing units, including special purpose processing units such as image processing units GPU, field programmable gate arrays FPGA, and application specific integrated circuits ASIC, and general purpose processing units such as central processing units CPU.
The application monitoring system 130 is used to monitor the running status of the application 120. Various status information of the application 120 may be stored in the application monitoring system 130, including, but not limited to, a plurality of historical usage values, a plurality of historical initial values, and a plurality of historical limit values for a plurality of resources over a plurality of time periods within a plurality of first historical periods for the application 120, a plurality of current usage values, a plurality of current initial values, and a plurality of current limit values for a plurality of resources over a plurality of time periods within a current period for the application 120.
The period includes, for example, but is not limited to, one day. The plurality of resources include, for example, but are not limited to, computing resources, storage resources, and bandwidth resources. Examples of computing resources include, but are not limited to, CPUs, GPUs, and the like, and examples of memory resources include, but are not limited to, internal memory resources and external memory resources. For example, in the micro service mode, the application runs on the container platform, the container platform isolates the resource usage, and the running resource of each application is provided with an initial value, a limit value and a usage value, such as a CPU initial value, a CPU limit value and a CPU usage value, and an internal memory initial value, an internal memory limit value and an internal memory usage value, wherein the usage value is between the initial value and the limit value.
Various access data of the application 120, such as a plurality of abnormal access amounts and a plurality of total access amounts of the application 120 during a plurality of second history periods, may be stored in the application access log system 140.
The computing device 110 may obtain, from the application monitoring system 130, a plurality of historical usage values, a plurality of historical initial values, and a plurality of historical limit values for the plurality of resources by the application 120 over a plurality of time periods within a plurality of first historical periods, a plurality of current usage values, a plurality of current initial values, and a plurality of current limit values for the plurality of resources over a plurality of time periods within a current period by the application 120, and a plurality of abnormal access amounts and a plurality of total access amounts for the application 120 over a plurality of second historical periods from the application access log system 140.
The computing device 110 is configured to obtain a plurality of current usage values, a plurality of current initial values, and a plurality of current limit values for the plurality of resources by the application 120 over a plurality of time periods within a current period; determining a plurality of current discrete coefficient feature values for the application 120 for the plurality of resources over a current period based on the plurality of current usage values, the plurality of current initial values, and the plurality of current limit values; determining a predicted abnormal feature value of the application 120 in a next period of the current period based on a plurality of historical discrete coefficient feature values of the application 120 for a plurality of resources in a plurality of first historical periods and a plurality of abnormal feature values of the application 120 in a plurality of second historical periods, each of the plurality of second historical periods being the next period of the corresponding first historical period, and a plurality of current discrete coefficient feature values determined based on a plurality of abnormal access amounts and a plurality of total access amounts of the application 120 in the plurality of second historical periods; and determining a predicted health level for the application 120 during a next period based on the predicted anomaly characteristic value and a predetermined association between the range of anomaly characteristic values and the health level.
Therefore, the predicted value of the machine learning model can reflect the actual scene better.
Fig. 2 shows a flow diagram of a method 200 for predicting a health level of an application in accordance with an embodiment of the present disclosure. For example, the method 200 may be performed by the computing device 110 as shown in FIG. 1. It should be understood that method 200 may also include additional blocks not shown and/or may omit blocks shown, as the scope of the present disclosure is not limited in this respect.
At block 202, the computing device 110 obtains a plurality of current usage values, a plurality of current initial values, and a plurality of current limit values for the plurality of resources by the application 120 over a plurality of time periods within a current period. The plurality of current usage values, the plurality of current initial values, and the plurality of current limit values may have been previously or immediately obtained by the computing device 110 from the application monitoring system 130.
For example, including but not limited to 1 day. The current period is, for example, the current day. Time periods include, for example, but are not limited to, minutes, hours, and the like.
For example, the computing device 110 may obtain a CPU current usage value, a CPU current initial value, and a CPU current limit value, an internal memory (MEM) current usage value, an internal memory current initial value, and an internal memory current limit value, a bandwidth current usage value, a bandwidth current initial value, and a bandwidth current limit value for a plurality of minutes of the day. In addition, the computing device 110 may also obtain the number of disk reads and writes Per Second (IOPS). Examples may be as shown in the following table.
At block 204, the computing device 110 determines a plurality of current discrete coefficient feature values for the application 120 for the plurality of resources within the current period based on the plurality of current usage values, the plurality of current initial values, and the plurality of current limit values. The method for determining a plurality of current discrete coefficient characteristic values is described in detail below in conjunction with fig. 4.
At block 206, the computing device 110 determines a predicted anomaly feature value for the application 120 in a period next to the current period based on the machine learning model and the plurality of current discrete coefficient feature values. The machine learning model is trained based on a plurality of historical discrete coefficient feature values of the application 120 for a plurality of resources over a plurality of first historical periods and a plurality of anomaly feature values of the application 120 over a plurality of second historical periods. Each of the plurality of second history periods is a next period corresponding to the first history period. The plurality of exceptional characteristic values are determined based on a plurality of exceptional accesses and a plurality of total accesses by the application 120 during a plurality of second history periods.
The next period is tomorrow, for example.
At block 208, the computing device 110 determines a predicted health level for the application 120 in the next period based on the predicted anomalous feature values and the predetermined associations between the range of anomalous feature values and the health levels.
Thus, by characterizing the actual situation based on the discrete coefficient feature values generated using the values, the initial values, and the limit values, the predicted values of the machine learning model are made to more reflect the actual scenario with resource constraints.
Fig. 3 shows a flow diagram of a method 300 for predicting a health level of an application, in accordance with an embodiment of the present disclosure. For example, the method 300 may be performed by the computing device 110 as shown in FIG. 1. It should be understood that method 300 may also include additional blocks not shown and/or may omit blocks shown, as the scope of the disclosure is not limited in this respect.
At block 302, the computing device 110 obtains a plurality of historical usage values, a plurality of historical initial values, and a plurality of historical limit values for a plurality of periods of time within a plurality of first historical periods for a plurality of resources by the application 120. The plurality of historical usage values, the plurality of historical initial values, and the plurality of historical limit values may have been previously or immediately obtained by the computing device 110 from the application monitoring system 130.
The plurality of first history periods are, for example, a plurality of days in history. The obtaining of the plurality of historical use values, the plurality of historical initial values and the plurality of historical limit values is similar to the obtaining of the current use value, the current initial value and the current limit value, and is not repeated.
At block 304, the computing device 110 obtains a plurality of abnormal access volumes and a plurality of total access volumes for the application 120 over a plurality of second history periods, each of the plurality of second history periods being a next period of the corresponding first history period. The plurality of exceptional accesses and the plurality of total accesses may be previously or immediately obtained by the computing device 110 from the application access log system 140.
At block 306, the computing device 110 determines a plurality of anomalous feature values for the application 120 over a plurality of second history periods based on the plurality of anomalous visions and the plurality of total visions.
The anomaly characteristic value in the second history period may be generated, for example, by determining a quotient between an anomaly access amount and a total access amount in the same second history period. Examples of the abnormal access amount, the total access amount, and the abnormal feature value are shown in the following tables.
At block 308, the computing device 110 determines a plurality of historical discrete coefficient feature values for the application 120 for the plurality of resources over a plurality of first historical periods based on the plurality of historical usage values, the plurality of historical initial values, and the plurality of historical limit values.
At block 310, the computing device 110 trains a machine learning model based on the plurality of historical discrete coefficient feature values and the plurality of anomaly feature values.
At block 312, computing device 110 obtains a plurality of current usage values, a plurality of current initial values, and a plurality of current limit values for a plurality of periods within a current period for a plurality of resources by application 120.
At block 314, the computing device 110 determines a plurality of current discrete coefficient feature values for the application 120 for the plurality of resources within the current period based on the plurality of current usage values, the plurality of current initial values, and the plurality of current limit values.
At block 316, the computing device 110 determines a predicted anomaly feature value for the application 120 in a next period of the current period based on the trained machine learning model and the plurality of current discrete coefficient feature values.
At block 318, the computing device 110 determines a predicted health level for the application 120 in the next period based on the predicted anomalous feature values and the predetermined associations between the range of anomalous feature values and the health levels.
The health levels may include, for example, 3 levels, healthy, sub-healthy, and sick. The predetermined associations are, for example, health associated with the anomaly characteristic value range [0, 0.01), sub-health associated with the anomaly characteristic value range [0.01, 0.1), and morbidity associated with the anomaly characteristic value range [0.1, 1). It should be understood that this is by way of example only and that the predetermined association may take other forms.
Thus, the actual situation can be characterized by the discrete coefficient feature value generated based on the use value, the initial value, and the limit value, so that the predicted value of the machine learning model can reflect the actual scene with resource constraints more. Further, the health level of the application in the next period can be predicted based on the trained machine learning model and the usage values, initial values, and limit values of the application for the plurality of resources in the current period.
In some embodiments, computing device 110 may also determine whether the predicted health level of application 120 during the next period is below a predetermined health level. If the computing device 110 determines that the predicted health level of the application 120 during the next period is below the predetermined health level, the initial values and limit values of the application 120 for the plurality of resources during the next period are adjusted. The predetermined health level includes, for example, but is not limited to, sub-health. Adjusting the initial value and the limit value may include decreasing the initial value by a predetermined magnitude and increasing the limit value, thereby enabling the application to gain more space for resource usage and increasing the actual health level of the application during the next period.
Thus, the initial value and the limit value in the next period can be adjusted in time based on the predicted health level, thereby improving the actual health level of the application in the next period.
Fig. 4 shows a flow diagram of a method 400 for determining a plurality of current discrete coefficient feature values according to an embodiment of the present disclosure. For example, the method 400 may be performed by the computing device 110 as shown in FIG. 1. It should be understood that method 400 may also include additional blocks not shown and/or may omit blocks shown, as the scope of the disclosure is not limited in this respect. The computing device 110 performs the following steps for each of the plurality of resources.
At block 402, the computing device 110 determines a plurality of first differences between a plurality of current usage values and a plurality of current initial values and a plurality of second differences between the plurality of current usage values and a plurality of current limit values for the resource over a plurality of time periods within a current period for the application 120.
At block 404, the computing device 110 determines a covariance between the first plurality of differences and the second plurality of differences.
At block 406, the computing device 110 determines a first standard deviation of the plurality of first differences and a second standard deviation of the plurality of second differences.
At block 408, the computing device 110 determines a current discrete coefficient feature value for the resource within a current period for the application 120 based on the covariance, the first standard deviation, and the second standard deviation.
For example, the following formula may be employed to determine the current discrete coefficient characteristic value.
Wherein the content of the first and second substances,
is the ith sequence value of a plurality of differences between the plurality of current usage values and the plurality of current initial values,
is composed of
Is measured.
Is the ith sequence value of a plurality of differences between the plurality of current usage values and the plurality of current limit values,
is composed of
Is measured.
Therefore, by using the difference of the value relative to the initial value and the limiting value, the discrete coefficient characteristic value of the application relative to the resource is determined, and the actual operation condition of the application under the resource limited condition is reflected more accurately.
Further, the computing device 110 may also determine, for example, bandwidth discrete coefficient characteristic values, read IO discrete coefficient characteristic values, and write IO discrete coefficient characteristic values from the standard deviation and the usage value. The specific formula can be as follows.
Wherein the content of the first and second substances,
is the ith bandwidth usage value, read IOPS usage value or write IOPS usage value on a certain day,
is the average of n bandwidth usage values, read IOPS usage values, or write IOPS usage values.
Fig. 5 shows a flow diagram of a method 500 for training a machine learning model according to an embodiment of the present disclosure. For example, the method 500 may be performed by the computing device 110 as shown in fig. 1. It should be understood that method 500 may also include additional blocks not shown and/or may omit blocks shown, as the scope of the disclosure is not limited in this respect.
At block 502, the computing device 110 generates a plurality of samples based on a plurality of historical discrete coefficient feature values and a plurality of anomaly feature values, each sample of the plurality of samples including applying the plurality of historical discrete coefficient feature values for the plurality of resources over a first historical period associated with the sample and applying the anomaly feature value over a corresponding second historical period for the first historical period.
An example of multiple samples may be as follows, where bandwidth, read/write IO discrete coefficient characteristics are optional.
For each of the plurality of samples, the following steps are performed in a loop.
At block 504, the computing device 110 generates an intermediate predicted anomaly feature value based on the linear regression model and a plurality of historical discrete coefficient feature values included in the sample.
The linear regression model includes a plurality of weights corresponding to a plurality of resources. The linear regression model can be expressed by the following formula.
Wherein the content of the first and second substances,
is the intermediate prediction abnormal characteristic value corresponding to the ith sample, n is the number of samples, k is the number of kinds of resources,
is the weight of the j-th resource,
and the characteristic value of the historical discrete coefficient corresponding to the j resource in the ith sample is shown, wherein b is a bias term.
At block 506, the computing device 110 determines a plurality of updated values for a plurality of weights associated with a plurality of resources in the linear regression model based on the predetermined loss function, the intermediate predicted anomaly characteristic values, and the anomaly characteristic values included in the samples.
The predetermined loss function includes, for example, but is not limited to, a maximum likelihood estimate. The predetermined loss function can be expressed by the following equation.
Wherein the content of the first and second substances,
is the intermediate prediction abnormal characteristic value corresponding to the ith sample,
is the true predicted eigenvalue corresponding to the ith sample.
For example, a partial derivative of the predetermined loss function with respect to the weight may be calculated to determine the update value.
At block 508, the computing device 110 updates a plurality of weights based on the plurality of updated values.
At block 510, the computing device 110 determines whether the plurality of update values are each less than a predetermined value. The predetermined value includes, for example, but is not limited to, 0.01.
If, at block 510, the computing device 110 determines that the plurality of update values are each less than the predetermined value, then the loop is ended resulting in the trained linear regression model as the trained machine learning model, else the loop continues for the next sample.
Therefore, the linear regression model can be trained on the basis of the multiple historical discrete coefficient characteristic values of the multiple resources in the first historical period and the abnormal characteristic values in the second historical period corresponding to the first historical period, and the abnormal characteristic values are predicted, so that the predicted abnormal characteristic values are closer to the actual situation under the constraint condition.
Fig. 6 illustrates a schematic block diagram of an example device 600 that can be used to implement embodiments of the present disclosure. For example, computing device 110 as shown in FIG. 1 may be implemented by device 600. As shown, device 600 includes a Central Processing Unit (CPU) 601 that may perform various appropriate actions and processes in accordance with computer program instructions stored in a Read Only Memory (ROM) 602 or loaded from a storage unit 608 into a Random Access Memory (RAM) 603. In the random access memory 603, various programs and data required for the operation of the device 600 can also be stored. The central processing unit 601, the read only memory 602, and the random access memory 603 are connected to each other via a bus 604. An input/output (I/O) interface 605 is also connected to bus 604.
A number of components in the device 600 are connected to the input/output interface 605, including: an input unit 606 such as a keyboard, a mouse, a microphone, and the like; an output unit 607 such as various types of displays, speakers, and the like; a storage unit 608, such as a magnetic disk, optical disk, or the like; and a communication unit 609 such as a network card, modem, wireless communication transceiver, etc. The communication unit 609 allows the device 600 to exchange information/data with other devices via a computer network such as the internet and/or various telecommunication networks.
The various processes and processes described above, such as the method 200 and 500, may be performed by the central processing unit 601. For example, in some embodiments, the method 200-500 may be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 608. In some embodiments, part or all of the computer program may be loaded and/or installed onto the device 600 via the read only memory 602 and/or the communication unit 609. When the computer program is loaded into the random access memory 603 and executed by the central processing unit 601, one or more of the actions of the method 200 and 500 described above may be performed.
The present disclosure relates to methods, apparatuses, systems, electronic devices, computer-readable storage media and/or computer program products. The computer program product may include computer-readable program instructions for performing various aspects of the present disclosure.
The computer readable storage medium may be a tangible device that can hold and store the instructions for use by the instruction execution device. The computer readable storage medium may be, for example, but not limited to, an electronic memory device, a magnetic memory device, an optical memory device, an electromagnetic memory device, a semiconductor memory device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a Static Random Access Memory (SRAM), a portable compact disc read-only memory (CD-ROM), a Digital Versatile Disc (DVD), a memory stick, a floppy disk, a mechanical coding device, such as punch cards or in-groove projection structures having instructions stored thereon, and any suitable combination of the foregoing. Computer-readable storage media as used herein is not to be construed as transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission medium (e.g., optical pulses through a fiber optic cable), or electrical signals transmitted through electrical wires.
The computer-readable program instructions described herein may be downloaded from a computer-readable storage medium to a respective computing/processing device, or to an external computer or external storage device via a network, such as the internet, a local area network, a wide area network, and/or a wireless network. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers and/or edge servers. The network adapter card or network interface in each computing/processing device receives computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in the respective computing/processing device.
The computer program instructions for carrying out operations of the present disclosure may be assembler instructions, Instruction Set Architecture (ISA) instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source or object code written in any combination of one or more programming languages, including an object oriented programming language such as Smalltalk, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider). In some embodiments, the electronic circuitry that can execute the computer-readable program instructions implements aspects of the present disclosure by utilizing the state information of the computer-readable program instructions to personalize the electronic circuitry, such as a programmable logic circuit, a Field Programmable Gate Array (FPGA), or a Programmable Logic Array (PLA).
Various aspects of the present disclosure are described herein with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the disclosure. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer-readable program instructions.
These computer-readable program instructions may be provided to a processing unit of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processing unit of the computer or other programmable data processing apparatus, create means for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks. These computer-readable program instructions may also be stored in a computer-readable storage medium that can direct a computer, programmable data processing apparatus, and/or other devices to function in a particular manner, such that the computer-readable medium storing the instructions comprises an article of manufacture including instructions which implement the function/act specified in the flowchart and/or block diagram block or blocks.
The computer readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other devices to produce a computer implemented process such that the instructions which execute on the computer, other programmable apparatus or other devices implement the functions/acts specified in the flowchart and/or block diagram block or blocks.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
Having described embodiments of the present disclosure, the foregoing description is intended to be exemplary, not exhaustive, and not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terms used herein were chosen in order to best explain the principles of the embodiments, the practical application, or technical improvements to the techniques in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Claims (7)
1. A method for predicting a health level of an application, comprising:
acquiring a plurality of current use values, a plurality of current initial values and a plurality of current limit values of the application for a plurality of resources in a plurality of time periods in a current period;
determining a plurality of current discrete coefficient feature values for the application for the plurality of resources over the current period based on the plurality of current usage values, the plurality of current initial values, and the plurality of current limit values;
determining a predicted anomaly characteristic value for the application during a next period of the current period based on a machine learning model and the plurality of current discrete coefficient characteristic values, the machine learning model being trained based on a plurality of historical discrete coefficient characteristic values for the application for the plurality of resources during a plurality of first historical periods and a plurality of anomaly characteristic values for the application during a plurality of second historical periods, each of the plurality of second historical periods being a next period corresponding to a first historical period, the plurality of anomaly characteristic values being determined based on a plurality of anomaly visitors and a plurality of total visitors for the application during the plurality of second historical periods; and
determining a predicted health level of the application within the next period based on a predetermined association between a range of anomaly characteristic values and a health level and the predicted anomaly characteristic values;
wherein determining that the application is to apply the plurality of current discrete coefficient feature values for the plurality of resources within the current period comprises, for each resource of the plurality of resources:
determining a plurality of first differences between a plurality of current usage values and a plurality of current initial values and a plurality of second differences between a plurality of current usage values and a plurality of current limit values for the resource by the application over the plurality of time periods within the current period;
determining a covariance between the plurality of first differences and the plurality of second differences;
determining a first standard deviation of the plurality of first differences and a second standard deviation of the plurality of second differences; and
determining a current discrete coefficient feature value of the application for the resource within the current period based on the covariance, the first standard deviation, and the second standard deviation.
2. The method of claim 1, further comprising:
obtaining a plurality of historical usage values, a plurality of historical initial values and a plurality of historical limit values of the application for the plurality of resources in the plurality of time periods in the plurality of first historical periods;
obtaining a plurality of abnormal access quantities and a plurality of total access quantities of the application in the plurality of second history periods;
determining a plurality of anomaly feature values of the application during the plurality of second history periods based on the plurality of anomaly access volumes and the plurality of total access volumes;
determining a plurality of historical discrete coefficient feature values of the application for the plurality of resources over the plurality of first historical periods based on the plurality of historical usage values, the plurality of historical initial values, and the plurality of historical limit values; and
training the machine learning model based on the plurality of historical discrete coefficient feature values and the plurality of abnormal feature values.
3. The method of claim 2, wherein training the machine learning model comprises:
generating a plurality of samples based on the plurality of historical discrete coefficient feature values and the plurality of anomaly feature values, each sample of the plurality of samples including a plurality of historical discrete coefficient feature values of the application for the plurality of resources over a first historical period associated with the sample and an anomaly feature value of the application over a corresponding second historical period during the first historical period; and
for each sample of the plurality of samples, cyclically performing the steps of:
generating an intermediate prediction abnormal characteristic value based on a linear regression model and a plurality of historical discrete coefficient characteristic values included in the sample;
determining a plurality of updated values of a plurality of weights associated with the plurality of resources in the linear regression model based on a predetermined loss function, the intermediate predicted abnormal feature values, and abnormal feature values included in the samples;
updating the plurality of weights based on the plurality of updated values; and
if it is determined that the plurality of update values are all less than the predetermined value, the loop ends.
4. The method of claim 1, further comprising:
adjusting the initial values and the limit values of the application for the plurality of resources during the next period if it is determined that the application's predicted health level during the next period is below a predetermined health level.
5. The method of claim 1, wherein the plurality of resources comprises computing resources, storage resources, and bandwidth resources.
6. An electronic device, comprising:
at least one processor; and
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-5.
7. A non-transitory computer readable storage medium having stored thereon computer instructions for causing the computer to perform the method of any one of claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110329739.7A CN112799932B (en) | 2021-03-29 | 2021-03-29 | Method, electronic device, and storage medium for predicting health level of application |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110329739.7A CN112799932B (en) | 2021-03-29 | 2021-03-29 | Method, electronic device, and storage medium for predicting health level of application |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112799932A CN112799932A (en) | 2021-05-14 |
| CN112799932B true CN112799932B (en) | 2021-07-06 |
Family
ID=75815797
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110329739.7A Active CN112799932B (en) | 2021-03-29 | 2021-03-29 | Method, electronic device, and storage medium for predicting health level of application |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112799932B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190213099A1 (en) * | 2018-01-05 | 2019-07-11 | NEC Laboratories Europe GmbH | Methods and systems for machine-learning-based resource prediction for resource allocation and anomaly detection |
| CN110674009A (en) * | 2019-09-10 | 2020-01-10 | 平安普惠企业管理有限公司 | Application server performance monitoring method and device, storage medium and electronic equipment |
| CN111124844A (en) * | 2018-10-30 | 2020-05-08 | 安碁资讯股份有限公司 | Method and apparatus for detecting abnormal operation of operating system |
| CN111527478A (en) * | 2017-10-13 | 2020-08-11 | 华为技术有限公司 | System and method for detecting abnormal user experience and performance in cooperation with cloud equipment |
-
2021
- 2021-03-29 CN CN202110329739.7A patent/CN112799932B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111527478A (en) * | 2017-10-13 | 2020-08-11 | 华为技术有限公司 | System and method for detecting abnormal user experience and performance in cooperation with cloud equipment |
| US20190213099A1 (en) * | 2018-01-05 | 2019-07-11 | NEC Laboratories Europe GmbH | Methods and systems for machine-learning-based resource prediction for resource allocation and anomaly detection |
| CN111124844A (en) * | 2018-10-30 | 2020-05-08 | 安碁资讯股份有限公司 | Method and apparatus for detecting abnormal operation of operating system |
| CN110674009A (en) * | 2019-09-10 | 2020-01-10 | 平安普惠企业管理有限公司 | Application server performance monitoring method and device, storage medium and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112799932A (en) | 2021-05-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9946633B2 (en) | Assessing risk of software commits to prioritize verification resources | |
| CN110852421B (en) | Model generation method and device | |
| US10956001B2 (en) | Progress bar updated based on crowd sourced statistics | |
| US11004012B2 (en) | Assessment of machine learning performance with limited test data | |
| US10769551B2 (en) | Training data set determination | |
| US10832150B2 (en) | Optimized re-training for analytic models | |
| JP7320053B2 (en) | Refinement of predictive models | |
| US10579371B2 (en) | Recommendations for custom software upgrade by cognitive technologies | |
| CN110766185A (en) | User quantity determination method and system, and computer system | |
| US11182689B2 (en) | Adaptive learning rate schedule in distributed stochastic gradient descent | |
| CN111353601A (en) | Method and apparatus for predicting delay of model structure | |
| US9734450B2 (en) | Data loss prevention to remove false positives | |
| US10572827B2 (en) | Prediction quality assessment | |
| CN112799932B (en) | Method, electronic device, and storage medium for predicting health level of application | |
| US10332234B2 (en) | Matrix factorization with approximate computing | |
| CN107229487B (en) | Code processing method and device | |
| US11823083B2 (en) | N-steps-ahead prediction based on discounted sum of m-th order differences | |
| US11410077B2 (en) | Implementing a computer system task involving nonstationary streaming time-series data by removing biased gradients from memory | |
| US9336140B1 (en) | Efficient management of hierarchically-linked data storage spaces | |
| CN111857558B (en) | Method, apparatus and computer program product for storage management | |
| CN112988455A (en) | Method, apparatus and computer program product for data backup | |
| US11151221B2 (en) | Probability density ratio estimation | |
| CN113361701A (en) | Quantification method and device of neural network model | |
| US20180232292A1 (en) | Error checking of a multi-threaded computer processor design under test | |
| CN115061722B (en) | Method, apparatus, and medium for configuring resources for a new version of an application |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |