WO2019109352A1

WO2019109352A1 - Method and device for sampling performance data of apparatus

Info

Publication number: WO2019109352A1
Application number: PCT/CN2017/115294
Authority: WO
Inventors: 刘金虎; 陈安伟
Original assignee: 华为技术有限公司
Priority date: 2017-12-08
Filing date: 2017-12-08
Publication date: 2019-06-13
Also published as: CN110140326A

Abstract

The present application provides a method and a device for sampling performance data of an apparatus. The method comprises: determining a first time range and a second time range in a first preset time period, a degree of change in values of performance data of the apparatus in the first time range being smaller than a degree of change in values of performance data of the apparatus in the second time range; sampling the performance data of the apparatus in the first time range at a first sampling interval in the first time range; and sampling the performance data of the apparatus in the second time range at a second sampling interval in the second time range, the first sampling interval being greater than the second sampling interval. In embodiments of the present application, a sampling interval used for sampling performance data of an apparatus is selected on the basis of a degree of change in values of performance data of the apparatus, thereby facilitating a reduction in a data amount of performance data obtained by sampling the performance data of the apparatus, while also reflecting changes of the performance of the reaction apparatus over time more accurately.

Description

Method and apparatus for sampling performance data of a device

Technical field

The present application relates to the field of data processing, and more particularly to a method and apparatus for sampling performance data of a device.

Background technique

By monitoring the performance of the device, you can monitor the device performance and discover the service trends and abnormalities of the device. This improves the service quality of the device and effectively prevents device faults. It is the most basic part of device performance analysis and operation and maintenance. Only accurate data collection can effectively guarantee the quality of monitoring equipment performance.

In the prior art, performance data of a device is typically sampled using a fixed sampling interval. If the sampling interval is set large, the sampled performance data may not accurately reflect the performance of the device over time. If the sampling interval is set small, the sampled performance data can accurately reflect the performance of the device over time, but the sampled performance data has a large amount of data and occupies more storage resources.

Summary of the invention

The present application provides a method and apparatus for sampling performance data of a device, reducing the amount of performance data obtained by sampling the performance data of the device, and accurately reflecting the performance of the device according to the performance data obtained after sampling. Changes over time.

A first aspect provides a method for sampling performance data of a device, including: determining a first time interval and a second time interval in a first preset time period, wherein the device is in the first time interval The degree of change in the value of the performance data is less than the degree of change in the value of the performance data of the device in the second time interval; in the first time interval, the first time is at the first sampling interval The performance data of the device in the interval is sampled; in the second time interval, the performance data of the device in the second time interval is sampled at a second sampling interval, wherein the first sampling interval is greater than The second sampling interval is described.

In the embodiment of the present application, the value of the value of the performance data of the device is different, and the performance data of the device is sampled by using different sampling intervals, and the first time interval of the value of the performance data of the device is small. The performance data of the device is sampled by using the first sampling interval, and the performance data of the device is sampled by using the second sampling interval for the second time interval in which the value of the performance data of the device changes greatly, and the performance data of the device is reduced. The amount of data of the performance data (ie, sampled data) obtained after sampling, and the performance data of the device can be accurately reflected according to the performance data obtained after sampling.

Optionally, the value of the performance data of the device may reflect the value of a certain type of performance parameter of the device.

With reference to the first aspect, in a possible implementation manner, the performance data of the device is data that changes with time in time, and the method further includes: performing performance data of the device according to the second preset time period. Determining the period according to a trend of the value, and determining the first time interval and the second time interval in the first preset time period, including: determining the first time interval and the second time in the period The interval, the first preset time period is the period.

The period of the change of the value of the performance data of the device with time is the first preset time period, and the arrangement of the first time interval and the second time interval in the period is determined, so as to be in the time period after the period The performance data of the device is sampled according to the arrangement of the first time interval and the second time interval in the cycle. It is avoided that the arrangement of the first time interval and the second time interval is determined in each period of the change of the value of the performance data of the device with time, which is advantageous for reducing the complexity of sampling the performance data of the device.

With reference to the first aspect, in a possible implementation manner, the determining the first time interval and the second time interval in the period includes: dividing the period into multiple time intervals, and determining the multiple The standard deviation of the performance data of the device in each time interval of each time interval; from the standard deviation set consisting of the standard deviations corresponding to the plurality of time intervals, the time interval in which the standard deviation is smaller than the standard deviation threshold is selected as The first time interval; a time interval in which a standard deviation is greater than or equal to the standard deviation threshold is selected as the second time interval from a standard deviation set consisting of standard deviations corresponding to the plurality of time intervals.

Optionally, the first time interval may be one or more time intervals within the period.

In the embodiment of the present application, the first time interval and the second time interval are determined by the relationship between the standard deviation of the value of the performance data in the time interval and the standard deviation threshold, which is beneficial to reduce the division of the period into the first time interval and the first time interval. The complexity of the two time intervals.

With reference to the first aspect, in a possible implementation manner, the determining the first time interval and the second time interval in the period includes: dividing the period into multiple time intervals, and determining the multiple a standard deviation of the performance data of the device in each time interval of each time interval; determining a time interval satisfying the selection rule of the initial interval as an initial interval based on the standard deviation corresponding to the plurality of time intervals respectively; The initial interval determines the first time interval, and the performance data in the first time interval includes performance data in the initial interval, and performance data of a device corresponding to the target sampling point, where the target sampling point is located in the a sampling point adjacent to the end point of the initial interval, the value of the performance data corresponding to the target sampling point and the mean value of the performance data in the initial interval satisfying |e ₁ -m ₀ | <3d ₀ , where e ₁ represents the value of the performance data of the device corresponding to the target sampling point, and m ₀ represents the mean value of the performance data in the initial interval. ;d ₀ represents the standard deviation of the values of the performance data in the initial interval.

In the embodiment of the present application, the performance data in the first time interval includes performance data in the initial interval, and performance data of the device corresponding to the target sampling point, where the target sampling point is outside the initial interval, and a sampling point adjacent to an end point of the initial interval, that is, an extension of an end point of the initial interval such that the first time interval contains as much performance data as possible, that is, a time period of the finally determined first time interval As large as possible, it is beneficial to further reduce the amount of data of the sampled data after sampling the performance data of the device.

Optionally, the selection rule of the initial interval may include any one of the following: the standard deviation of the value of the performance data of the corresponding device in the multiple time intervals, where the minimum standard deviation corresponds to the initial interval; Among the standard deviations of the values of the performance data of the corresponding devices in the time interval, the time interval corresponding to the standard deviation smaller than the standard deviation threshold is the initial interval.

With reference to the first aspect, in a possible implementation manner, the magnitude of the change in the value of the performance data of the device is inversely proportional to the length of the first sampling interval in the first time interval, and And or in the second time interval, the magnitude of the change in the value of the performance data of the device is inversely proportional to the length of the second sampling interval.

With reference to the first aspect, in a possible implementation manner, an inverse relationship between a magnitude of a change in the value of the performance data of the device and a length of the first sampling interval in the first time interval And an inverse relationship between the magnitude of the change in the value of the performance data of the device and the length of the second sampling interval in the second time interval

Wherein P represents the length of the first sampling interval or the second sampling interval, k is a preset number greater than 1, and σ represents the device in the first time interval or the second time interval The standard deviation of the performance data is used to indicate the degree of change of the value of the performance data in the first time interval or the second time interval.

In the embodiment of the present application, the time interval used for sampling the performance data of the device in the time interval is determined based on an inverse relationship between the degree of change of the value of the performance data of the device in the time interval and the length of the sampling interval. The value of the performance data of the reaction device that is more accurate for sampling data is more accurate with time.

With reference to the first aspect, in a possible implementation, the determining the period according to a trend of a value of performance data of the device in a second preset time period, including: acquiring the a trend of changing the value of the performance data of the device over time in a preset time period, and dividing the second preset time period into N sub-time segments, where N is a positive integer; and the N sub-periods The change trend of the value of the performance data of the device in the i-th sub-period is the reference, and the change trend of the value of the device performance data in the j-th sub-period within the N sub-periods is determined. The degree of similarity between the trends of the performance data of the device in the i-th sub-period, j ∈ [1, N], i ∈ [1, N], j ≠ i, and j and i is a positive integer; between the change trend of the value of the performance data of the device in the i-th sub-period and the change trend of the value of the performance data of the device in the j-th sub-period In the degree of similarity, selecting the performance of the device in the i-th sub-period Determining a target time period in which the change trend of the value is the same; determining a length of time between the first sampling point in the i-th sub-period and the second sampling point in the target time period, the first sampling point The position in the i-th sub-period is the same as the position of the second sampling point in the target time period; selecting any sampling point in the second preset time period as the start of the period The time is determined by the length of time between the first sampling point and the second sampling point being the length of the period.

With reference to the first aspect, in a possible implementation, the change trend of the value of the performance data of the device in the i-th sub-period and the performance data of the device in the j-th sub-period The degree of similarity between the trends of the values is

Wherein, Y represents the number of sampling points for sampling the performance data of the device included in the i-th sub-period and the j-th sub-period, and t _iy represents the _{yth in} the i-th sub-period The value of the performance data of the device, t _jy represents the value of the performance data of the _yth device in the jth sub-time period,

Means the mean value of the performance data of the device in the i-th sub-period,

Indicates the mean value of the performance data of the device in the jth sub-period.

With reference to the first aspect, in a possible implementation manner, the method further includes: determining a third time interval in the first preset time period, and performance data of the device in the third time interval The mean value is higher than the performance data threshold; in the third time interval, the performance data in the third time interval is sampled at a third sampling interval, the third sampling interval being smaller than the first sampling interval.

With reference to the first aspect, in a possible implementation manner, the performance data of the device is any one of the following data: an occupancy rate of the computing resource in the device, an occupancy rate of the transmission resource in the device, The occupancy rate of storage resources in the device.

The average value of the performance data of the device in the third time interval is higher than the performance data threshold, indicating that the performance data of the current device is at a peak period, and during the peak period, the probability of abnormal operation of the device is large, therefore, In the embodiment of the present application, sampling the performance data of the device in the third time interval by using the third sampling interval is beneficial to increasing the basis for subsequent analysis of the performance of the device.

In a second aspect, an apparatus for sampling performance data of a device is provided, the apparatus comprising various modules for performing the above method.

In a third aspect, an apparatus for sampling performance data of a device is provided, the apparatus having the functionality to implement the design of the above method. These functions can be implemented in hardware or in software by executing the corresponding software. The hardware or software includes one or more units corresponding to the functions described above.

In a fourth aspect, an apparatus for sampling performance data of a device, including a processor and a memory, is provided. The memory is for storing a computer program for calling and running the computer program from memory such that the apparatus performs the method of the above aspects.

In a fifth aspect, a computer program product is provided, the computer program product comprising: computer program code, causing a computer to perform the method of the above aspects when the computer program code is run on a computer.

In a sixth aspect, a computer readable medium storing program code for causing a computer to perform the method of the above aspects when the computer program code is run on a computer.

In a seventh aspect, a chip system is provided, the chip system comprising a processor, the means for sampling the performance data implementing the functions involved in the above aspects, for example, generating, receiving, transmitting, or processing the above method Data and/or information involved. In a possible design, the chip system further comprises a memory for storing necessary program instructions and data of the terminal device. The chip system can be composed of chips, and can also include chips and other discrete devices.

DRAWINGS

FIG. 1 is a schematic block diagram of a monitoring scenario to which the embodiment of the present application is applied.

FIG. 2 is a schematic flowchart of a method for sampling performance data of a device according to an embodiment of the present application.

FIG. 3 is a schematic diagram showing a distribution of a plurality of sub-periods in the second preset time period in the embodiment of the present application.

FIG. 4 is a schematic diagram showing a distribution of a plurality of sub-periods in a second preset time period according to another embodiment of the present application.

FIG. 5 is a schematic diagram showing the distribution of a plurality of time intervals in a period in the embodiment of the present application.

FIG. 6 is a schematic diagram showing the distribution of a plurality of time intervals in a cycle according to another embodiment of the present application.

FIG. 7 is a schematic flowchart of a method for expanding an initial interval according to an embodiment of the present application.

FIG. 8 is a schematic diagram showing the position of the initial interval in the cycle in the embodiment of the present application.

FIG. 9 is a schematic block diagram of an apparatus for sampling performance data of a device according to an embodiment of the present application.

FIG. 10 is a schematic block diagram of an apparatus for sampling performance data of a device according to another embodiment of the present application.

Detailed ways

The technical solutions in the present application will be described below with reference to the accompanying drawings.

For the sake of easy understanding, a monitoring scenario that can be applied in the embodiment of the present application is briefly introduced in conjunction with FIG. 1 . It should be noted that the embodiment of the present application is only described by using the monitoring scenario shown in FIG. 1 as an example. The embodiment of the present application can also be applied to other systems with monitoring requirements, such as storage devices in a storage system and The server accessing the storage device is monitored, or the terminal in the network communication system and the server serving the terminal are monitored, and the server serving the terminal may be a storage server. The storage device in the storage system can also be a storage server, which can be a computer, a computer, or other computing device that can be used for storage.

FIG. 1 is a schematic block diagram of a monitoring scenario to which the embodiment of the present application is applied. The devices included in the monitoring scenario shown in FIG. 1 include a terminal 110, a storage system 120, and a monitor 130.

The terminal 110 allows the user to access the storage space of the storage server through the client set on the terminal.

It should be understood that the foregoing terminals may include, but are not limited to, a computing device, a mobile station (MS), a mobile terminal (Mobile Terminal), a mobile phone (Mobile Telephone), a user equipment (User Equipment, UE), a handset (handset), and Portable equipment, etc.

The storage system 120 provides storage services and access services for the terminal. The control node is configured to control the storage node, such as monitoring the capacity of the storage node or responsible for load balancing of the storage node; the storage node is configured to provide storage space, for example, for storage terminal uploading. Data to the storage system.

The monitor 130 is configured to sample performance data of the storage system and/or the terminal, monitor performance changes of the storage system and/or the terminal over time, and upload the sampled performance data (also referred to as sampling data) to Storage server.

It should be noted that the foregoing monitor may be a third-party monitor independent of the storage system and the terminal, and the monitor may also be located in the storage system, and the monitor may also be located in the terminal. The specific implementation form of the monitor is not limited in this embodiment of the present application.

In the prior art, the monitor samples the performance data of the device by using a fixed sampling interval in the process of sampling the performance data in the cloud storage device. If the sampling interval is set large, the accuracy of the device performance reflected by the sampling data obtained after sampling is low, and it is not consistent with the fact that it cannot be determined that the performance of the device reflected by the sampling data changes with time. If the sampling interval is set small, the accuracy of the device performance reflected by the sampling data obtained after sampling is high, which is favorable to reflect that the performance of the device changes with time, which is consistent with the fact, but uses smaller sampling. Interval sampling the performance data of the device will increase the amount of data of the sampled data. Before analyzing the sampled data, it is necessary to provide sufficient storage space to store a large amount of sampled data and consume storage space. In addition, when the sampling data is analyzed by a storage server or other analysis device connected to the monitor, if the amount of data of the sampled data is large, more transmission resources are consumed when transmitting the sampled data.

Therefore, the above method of sampling the performance data of the device using a fixed sampling interval cannot simultaneously satisfy the requirement of reducing the storage space of the sampled data, and the requirement of accurately reflecting the performance of the device over time.

In order to accurately reflect the performance of the cloud storage device over time, and reduce the data amount of the sampled data, the embodiment of the present application provides a method for sampling performance data of the device, and determining the first preset time period. a time interval and a second time interval, wherein a change degree of the value of the performance data of the device in the first time interval is smaller than a change degree of the value of the performance data of the device in the second time interval Capturing performance data of the device in the first time interval at the first sampling interval in the first time interval; and in the second time interval, in the second time interval The performance data of the device in the interval is sampled, wherein the first sampling interval is greater than the second sampling interval.

The method of the present application is described in detail below in conjunction with FIG.

FIG. 2 is a schematic flowchart of a method for sampling performance data of a device according to an embodiment of the present application. It should be understood that the method illustrated in FIG. 2 may be performed by the monitor 130 of FIG. 1, and the method illustrated in FIG. 2 may include steps 210 through 250.

210: Determine a period, and determine, according to the period, a value of performance data of each sampling point device in the period.

Specifically, the above device may be a monitored object, such as a terminal and/or a storage system in the monitoring scenario shown in FIG. 1.

The value of the performance data of the above device may reflect the value of a certain type of performance parameter of the device. For example, the performance data of the device may be the occupancy rate of the computing resource of the device, for example, the occupancy rate of the processor of the storage server in the storage system, or the occupancy rate of the terminal processor; the performance data of the device may also be the storage of the device. The occupancy rate of the resource, for example, the storage space of the storage server in the storage system; the performance data of the device may also be the occupancy rate of the transmission resource of the device, for example, the storage server in the storage system responds to the read request sent by the terminal. The transmission resource occupied by the terminal when transmitting data to the terminal may be a read bandwidth, a write bandwidth, or the like, or a bandwidth occupied by the storage server in the storage system to transmit data to the terminal.

It should be understood that the above period may be obtained by calculation based on the variation of the value of the historical performance data of the device with time. The above cycle may also be preset, for example, the above period may be set to one day (24 hours) or one month. The above period may also be determined based on the variation of the value of the historical performance data of the device with time.

The method for determining the period based on the change rule of the historical performance data of the device with time is mainly described below, wherein the specific implementation manner of determining the period can be implemented by the following steps 211-213, and determining the performance data of the period in each sampling. The value of the point can be implemented by step 214. For specific implementations, refer to the following description.

211. Divide the second preset time period into multiple sub-time segments, and obtain a trend of the value of the performance data of the device in each of the plurality of sub-time segments over time.

Specifically, the performance data of the device in the second preset time period may be understood as historical performance data indicating the performance of the device in the second preset time period. That is, before performing step 211, the method further includes: acquiring performance data of the device in the second preset time period. For example, get performance data from the previous month or the previous year or the first 24 hours.

It should be understood that there are many ways to divide the second preset time period into multiple sub-time periods, which is not limited by the embodiment of the present application.

For example, one implementation of dividing the second preset time period into multiple sub-time segments based on FIG. 3 is to divide the second preset time period into consecutive multiple sub-time segments. Please refer to FIG. 3. FIG. 3 is a schematic diagram showing the distribution of multiple sub-time periods in the second preset time period in the embodiment of the present application. On the time axis shown in FIG. 3, comprising a second predetermined period of time the device performance data sampled n times, n sampling points obtained _{_{t 1, t 2, ... t}} n corresponding to each device The value of the performance data x ₁ , x ₂ , ..., x _n , n is a positive integer. The values of the performance data of the device corresponding to the n sampling points t ₁ , t ₂ , ... t _n respectively are x ₁ , x ₂ , ..., and the performance data sequence of the device is X={x ₁ . x ₂ , ..., x _n }, where x _n represents the value of the nth performance data in the performance data sequence. Then, the second preset time period is divided into m end-to-end sub-time segments T ₁ , T ₂ , . . . , T _m , m are positive integers. Each sub-period includes L sample points, that is, the value of the performance data of the device included in the sub-period T ₁ is {x ₁ , x ₂ , . . . , x _L }, and is included in the sub-period T ₂ The value of the performance data of the device is {x _L , x _L+1 ,..., x _2L-1 }, and the performance data of the device included in the sub-period T ₃ is {x _2L-1 , x _2L ,...,x _3L-2 }, the performance data of the device included in the sub-period T _m is {x _(m-1)(L-1)+1 ,...,x _n } .

It should be noted that the number of performance data included in the second preset time period may not be completely divided by multiple sub-time segments of equal length, that is, the performance data in the second preset time period is followed. When each sub-period includes L performance data partitions, the number of remaining performance data may be less than L. In this case, the remaining performance data may be divided into the last sub-time period, and the last one in the second preset time period. The number of performance data contained in the sub-period is less than L.

In the embodiment of the present application, another implementation manner of dividing the second preset time period into multiple sub-time segments may be implemented based on FIG. 4, that is, dividing the second preset time period into multiple sub-times by sliding the first sliding window segment. It should be noted that, according to the manner in which the second preset time period is divided into multiple consecutive sub-time segments as shown in FIG. 3, the second preset is compared by sliding the first sliding window based on FIG. 4 . The method in which the time segment is divided into multiple sub-time segments is relatively simple, and is advantageous for reducing the complexity of dividing the second preset time segment into multiple sub-time segments. Referring to FIG. 4, FIG. 4 is another schematic diagram showing distribution of multiple sub-time periods of the embodiment of the present application in a second preset time period. On the time axis shown in FIG. 4, the performance data of the device is sampled n times in the second preset time period, and the obtained n sample points t ₁ , t ₂ , . . . , t _n respectively correspond to The value of the performance data of the device x ₁ , x ₂ , ..., x _n , n is a positive integer. The values of the performance data of the device corresponding to the n sampling points t ₁ , t ₂ , ... t _n respectively x ₁ , x ₂ , ..., x _n constitute the performance data sequence of the device is X = {x ₁ , x ₂ , ..., x _n }, where x _n represents the value of the nth performance data in the performance data sequence. Then, the first sliding window with the preset time width L is set, and the first preset step size is used in the second preset time period, starting from the starting time of the second preset time period, and the second preset time is The direction of the end time of the segment (for example, the direction from left to right in FIG. 4) is slid to generate a plurality of sub-periods T ₁ , T ₂ , ..., T _n-L+1 having a time width L . Each time the first sliding window having the preset time width L is slid in the first preset step, the performance data sequence of the device is valued, and L adjacent performance data can be obtained, where L is a positive integer. There are multiple implementations of the first preset step size, which may be the length of time between two adjacent performance data in the above performance data sequence, that is, the sampling interval used when intensively sampling the performance data of the device. Thus, the sequentially generated (n-L+1) subsequences are: X ₁ = {x ₁ , x ₂ , ..., x _L }, X ₂ = {x ₂ , x ₃ , ..., x _L+1 }, ..., X _n-L+1 = {x _n-L+1 , x _n-L+2 , ..., x _n }. That is, the value of the performance data of the device included in the sub-period T ₁ is {x ₁ , x ₂ , . . . , x _L }, and the value of the performance data of the device included in the sub-period T ₂ is { x ₂ , x ₃ ,..., x _L+1 }, the performance data of the device included in the sub-period T ₃ is {x ₃ , x ₄ ,..., x _L+2 }, sub The performance data of the device included in the time period T _n-L+1 is taken as {x _n-L+1 , x _n-L+2 , . . . , x _n }.

It should be understood that FIG. 4 merely exemplarily lists the first sliding step is a time length between two sampling points, and the first sliding window is slid in the direction of increasing time, and the generated sub-period The specific value of the first preset step and the sliding direction of the first sliding window are not specifically limited. For example, the first preset step may also be a sampling point corresponding to t ₁ . The length of time between the sampling points corresponding to t ₃ . The sliding direction of the first sliding window may also be the direction from the end time of the second preset time period to the starting point of the second preset time period (the right-to-left direction in FIG. 4).

It should also be understood that the first preset step size may be a sampling interval in which the performance data of the n devices is sampled in the process of acquiring performance data of the n devices. That is to say, when the performance data of the history of the device is acquired at each sampling point, an integer multiple of the sampling interval used may be used as the first preset step size. For example, the first preset step size may be a sampling interval of 1 time, or a sampling interval of 2 times, which is not limited by the embodiment of the present application.

It should be understood that the foregoing second preset time period may be one day, or one month, or one year, and the specific length of the second preset time period is not limited in the embodiment of the present application. The longer the time length set by the second preset time period, the more the data amount of the historical performance data of the device used for determining the period is, the more advantageous it is to determine the above period.

It should also be understood that, in order to increase the success rate of the determination period, if the second preset time period is too short, for example, when a complete period is not included, the determination period may be failed. Therefore, the second preset time period is longer. Can improve the success rate of the determination cycle.

In the embodiment of the present application, the method for determining that the second preset time period is divided into multiple sub-time segments is determined by using the first sliding window, and the method for dividing the second preset time period into multiple sub-time segments in FIG. 3 The smaller granularity of the partitioning is beneficial to improve the accuracy of the subsequently determined period.

212: Determine two sub-time periods from the plurality of sub-time periods, wherein the values of the performance data of the devices change in the two sub-time periods.

The two sub-periods determined above may be the two sub-periods with the closest distance, or two sub-periods with relatively close distances.

Specifically, the trend of the value of the performance data of the device in the second preset time period is obtained, and the second preset time period is divided into N sub-time segments, where N is a positive integer; Determining a change trend of the value of the performance data of the device in the i-th sub-period of the N sub-periods as a reference, and determining the device performance data in the j-th sub-period within the N sub-periods The degree of similarity between the change trend of the value and the change trend of the value of the performance data of the device in the i-th sub-period, j ∈ [1, N], i ∈ [1, N], j ≠ i, and j and i are positive integers; a variation trend of the value of the performance data of the device from the i-th sub-period and a value of the performance data of the device in the j-th sub-period a degree of similarity between the change trends, selecting a target time period that is the same as a change trend of the value of the performance data of the device in the i-th sub-period; determining a first sampling point in the i-th sub-time period a length of time between a second sampling point in the target time period, The position of the first sampling point in the i-th sub-period is the same as the position of the second sampling point in the target time period; selecting any sampling point in the second preset time period is the The start time of the period is determined by the length of time between the first sampling point and the second sampling point being the length of the period.

Optionally, as an embodiment, a change trend of the value of the performance data of the device in the i-th sub-period and a change trend of the value of the performance data of the device in the j-th sub-period The degree of similarity between s _ij , where

Y represents the number of sampling points for sampling the performance data of the device included in the i-th sub-period and the j-th sub-period, and t _iy represents the _yth device in the i-th sub-period The value of the performance data, t _jy represents the value of the performance data of the _yth device in the jth sub-time period,

Incidentally, the degree of similarity between the change of s _ij values of performance data trend values represent the performance data of the i-th period and the j-th equipment devices period, but also Pearson correlation coefficients referred to, in the range of s _ij [-1,1], wherein the larger the value of s _ij, s _ij i.e. the value close to 1, indicates the i-th period of performance data device the more similar trend value change trend of performance data and the value j-th time period device; the smaller the value of s _ij, s _ij i.e. the value close to -1, denotes the i th period The change trend of the value of the performance data of the internal device is not the same as the change trend of the value of the performance data of the device in the jth time period.

Optionally, if the value of s _ij belongs to [0.8, 1], the change trend of the value of the performance data of the device in the i-th time period and the value of the performance data of the device in the j-th time period may be considered. The trend of change is the same.

In the embodiment of the present application, when the change trend of the value of the performance data of the device in the i-th time period is the same as the change trend of the value of the performance data of the device in the j-th time period, the value range of s _ij is not performed. Specifically, the value range of s _ij is approximately 1, indicating that the change trend of the value of the performance data of the device in the i-th time period is the same as the change trend of the value of the performance data of the device in the j-th time period. The stricter the conditions, the more accurate basis for the subsequent determination cycle.

Optionally, determining the two sub-time periods from the plurality of time periods may also be determined by the maximum value of the value of s _ij . That is to say, when the value of s _ij is a maximum value, the change trend of the value of the performance data of the device in the i-th time period and the change trend of the value of the performance data of the device in the j-th time period can be considered. the same.

For example, the change trend of the value of the performance data of the device in the sub-sequence X ₁ can be used as a reference, and the trend of the value of the performance data of the device in the other sub-sequences of the plurality of sub-sequences is compared one by one to determine the plurality of sub-sequences. A subsequence in which the value of the performance data of the device in the subsequence X ₁ is the same. Wherein the degree of similarity between the s value trend value change trend of performance data sequence X in the device ₁ with other sequences in the plurality of sub-sequences device _1j performance data may be determined by the Pearson correlation coefficient , which is

And

among them,

Means the mean of the i-th sub-sequence X _i ;

Represents the mean of the first subsequence X ₁ .

Wherein, the larger the value of s _1j is, the degree of similarity between the change trend of the value of the performance data of the device in the sub-sequence X ₁ and the change trend of the value of the performance data of the device in the other sub-sequences of the plurality of sub-sequences The higher the value in X ₁ , therefore, the degree of similarity between the change trend of the value of the performance data of the device in the sub-sequence X ₁ and the change in the value of the performance data of the j-th sub-sequence X _j device in the other sub-sequences value the same trend of a plurality of s _1j, select the maximum value of s _1j of two sequences as the change of the value of the performance data of the same device performance data of two sub-sequences, two sequences device The corresponding sub-periods are the two sub-periods that are finally determined.

Optionally, the two sub-time periods in which the value of the performance data of the device is the same may be the sub-time period corresponding to the similarity degree s _ij when the maximum value occurs twice.

For example, when comparing the trend of the performance data of the device in the first sub-period with the trend of the performance data of the device in the second sub-period, the devices in the first sub-period and the second sub-period The degree of similarity of the performance data shows the first maximum value. When comparing the trend of the performance data of the device in the first sub-period period with the trend of the performance data of the device in the third sub-period, the first sub- The second maximum value occurs when the time period is similar to the performance data of the device in the third sub-time period. In this case, the change trend of the performance data of the device in the second sub-time period and the third sub-time period can be considered to be the same.

It should be noted that, in order to reduce the probability that the two sub-time segments with the same change trend of the value of the foregoing performance data are located in the same period, in the process of dividing the second preset time period, multiple lengths may be used. The first sliding window repeatedly divides the second preset time period, and determines a plurality of reference periods based on the plurality of sub-time segments divided by the plurality of first sliding windows, and finally may select the reference period with the highest frequency among the plurality of reference periods As the above period, or the reference period in which the longest time among the plurality of reference periods is selected is taken as the above period.

213. Determine a period length according to two sub-time periods in which the value of the performance data of the device changes in the same trend.

Specifically, it may be determined that a length of time between a first sampling point in any time period and a second sampling point in another time period is the period length, and the first sampling point is in any one of the time periods The position in the middle is the same as the position at which the second sampling point is in the other time period. The performance data between the first sampling point of any one of the time periods and the first sampling point of the any one of the time periods and the second one of the other time periods is a piece of performance data satisfying the length of the period . Optionally, the first sampling point may be a starting moment of any one of the time periods, and the second sampling point is a starting moment of the another time period. Alternatively, the first sampling point may be an ending time of any one of the time periods, and the second sampling point is an ending time of the another time period.

For example, the length Q of the above period may mean that each period includes consecutive Q performance data in the above-mentioned performance data sequence, that is, the length of the above period Q=Q×T _s , where Q is a positive integer and T _{s is} represented in the second The sampling interval used to sample the performance data of the device during the preset time period to generate the performance data sequence. From the above performance data sequence, the sampling point corresponding to any performance data is selected as the starting time of the period, and all the performance data in the period length are selected to form a complete sequence sub-sequence X _T ={x ₁ ,x ₂ ,. ..,x _Q }.

214. Determine a value of each sampling point of the performance data of the device in the period.

Specifically, determining the performance data of the device in each period of the sampling point can be determined by the following two methods.

Method 1: Taking any sampling point in the second preset time period as the starting time of the period, determining the value of the performance data of the device corresponding to the consecutive sampling points in the period length.

Optionally, in the time period in which the value of the performance data of the device in the second preset time period is changed to a greater extent, any sampling point is selected as the starting time of the cycle. It is beneficial to ensure the continuity of the first time interval in one cycle, and avoid selecting the first time of the cycle when the start time of the cycle is selected in a time period in which the change of the value of the performance data of the device is small. The interval is divided into two time intervals.

In the second method, the actual value of the performance data of the device in each cycle is determined, and the average value of the actual value of the performance data of the device corresponding to each sampling point in the same position in multiple cycles is calculated, and each sample is calculated. The average value of the performance data of the device corresponding to the point is taken as the value of the performance data of the device corresponding to each sampling point in the above cycle.

For example, using the performance data of the device, the average value of the true values of each sampling point in two periods is taken as the value of each sampling point in the period of the performance data of the above device. If L performance data is included in each of the two periods, the two periods determined by the above method 1 are T ₁ = {x ₁ , x ₂ , ..., x _L } and T ₂ = {x, respectively. _L+1 , x _L+2 , ..., x _L+L }, where L is a positive integer. According to the second method, the average value of the performance data of the device in the period determined based on the periods T ₁ and T ₂ is

It should be noted that any two pieces of performance data in the multi-segment performance data in the foregoing manner 2 do not overlap at all, that is, the length of the interval between the first performance data in any two pieces of performance data may be the length of the period. Integer multiple. Certainly, any two pieces of performance data in the multi-segment performance data in the foregoing mode 2 may partially overlap, that is, the length of time between the first performance data in any two of the foregoing performance data may be less than or equal to the length of the period.

In the embodiment of the present application, if the value of the performance data of the device corresponding to each sampling point in the period is determined according to the method 2 in step 214, it is advantageous to improve the accuracy of determining the first time interval and the second time interval, thereby avoiding a single The performance data of the individual anomalies of the device appear in the cycle, and the division of the first time interval and the second time interval is inaccurate.

If the value of the foregoing performance data changes periodically with time, the first time interval and the second time interval in the cycle may be determined through step 220. The second preset time period in the following may be a period, for example, the foregoing steps may be 230 determined period. If the value of the foregoing performance data changes non-periodically with time, the first time interval and the second time interval in the second preset time period (arbitrary time period) may be determined by step 220 without performing the above step 210.

220. Determine a first time interval and a second time interval in the first preset time period, where the value of the performance data of the device changes less than the second time interval. The degree of change in the value of the performance data of the device.

Specifically, the degree of change of the value of the performance data of the device may be represented by a variance or a standard deviation of the value of the performance data of the device. That is, the degree of change in the value of the performance data of the device in the first time interval can be expressed by the variance or standard deviation of the value of the performance data of the device in the first time interval. The degree of change in the value of the performance data of the device in the second time interval can be expressed by the variance or standard deviation of the value of the performance data of the device in the second time interval.

The degree of change of the value of the performance data of the device in the first time interval is smaller than the change degree of the value of the performance data of the device in the second time interval, which may be understood as being the first The severity of the change in the value of the performance data of the device in the time interval is less than the severity of the change in the value of the performance data of the device in the second time interval, for example, the value of the performance data of the device in the first time interval. The variance of the device is less than the variance of the value of the performance data of the device in the second time interval, or the standard deviation of the value of the performance data of the device in the first time interval is less than the value of the performance data of the device in the second time interval Standard deviation.

If the first preset time period is an arbitrary time, the device performance data may be intensively sampled in the first preset time period to obtain the value of the device performance data, and then based on the device performance data. The degree of change of the value determines the first time interval and the second time interval within the first preset time period. If the change of the performance data of the device is greater than the preset change threshold, the time is the first time interval. If the value of the performance data of the device is less than or equal to the pre-determination The threshold of change is set, and the period of time is the second time interval.

In the embodiment of the present application, the first sampling of the performance data of the device in a dense sampling manner (ie, a small fixed sampling interval) is beneficial to improving the accuracy of the performance of the first sampling data reaction device over time. .

It should be noted that the lengths of the first time interval and the second time interval may be the same or different.

If the value of the performance data changes periodically with time, the first time interval and the second time interval are located in the cycle. Before determining the distribution of the first time interval and the second time interval in the period, the period may be divided into a plurality of time intervals, and there are many specific manners, which are not limited in this embodiment of the present application. Two of them are highlighted below in conjunction with FIGS. 5 to 6.

The first implementation manner of dividing the period into a plurality of time intervals is to divide the period into a plurality of consecutive time intervals, and the lengths of the plurality of time intervals may be the same or different. For example, FIG. 5 is a schematic diagram showing the distribution of a plurality of time intervals in a period in the embodiment of the present application. On the time axis shown in FIG. 5, the performance data of the device is sampled n times in one cycle, and the obtained n sampling points t ₁ , t ₂ , ... t _n respectively correspond to the device The value of the performance data is x ₁ , x ₂ ,..., x _n , and n is a positive integer. The period is equally divided into m end-to-end time intervals X ₁ , X ₂ , ..., X _m , m being positive integers. Each time interval contains K sampling points, and K is a positive integer, that is, the performance data of the device included in the time interval X ₁ is {x ₁ , x ₂ , . . . , x _K }, time interval X The performance data of the device included in ₂ is {x _K , x _K+1 ,..., x _2k-1 }, and the performance data of the device included in the time interval X _m is {x _{(m -1) K-(m-2)} ,..., x _n }.

It should be noted that FIG. 5 only shows a case where the performance time data of all n devices is included in the m time intervals when one cycle is equally divided into m time intervals. There may be another case where a period is divided into m time intervals, m time intervals contain only performance data of na devices, and a is a positive integer smaller than K. At this time, the remaining a The performance data of the device may be divided into the m+1th time interval, and the time length of the m+1th time interval is smaller than the mth time interval.

In addition, the second implementation manner of dividing the period into a plurality of time intervals is to divide the period into a plurality of time intervals having the same time length, and the sampling points included in the adjacent two time intervals in the plurality of time intervals are the same. .

In the second implementation manner of dividing the period into multiple time intervals, the period corresponding to a piece of performance data that satisfies the period length may be divided into multiple by sliding the second sliding window by the second preset step. Time interval. For example, FIG. 6 shows a schematic diagram of a distribution of a plurality of time intervals in a cycle of another embodiment of the present application. On the time axis shown in FIG. 6, the value of the performance data of the n devices in one cycle is included, that is, the n sampling points t ₁ , t ₂ , ... t _n respectively correspond to the performance data of the device. The values are x ₁ , x ₂ ,..., x _n , and n is a positive integer. a second sliding window having a preset time width of L, in a second preset step on the time axis, starting from the beginning of the cycle and in the direction of the end of the cycle (left to right in FIG. 6) Sliding to generate a plurality of time intervals X ₁ , X ₂ , . . . , X _n-K+1 , where time width K is a positive integer. The performance data of the device included in the time interval X ₁ is taken as {x ₁ , x ₂ , . . . , x _K }, and the value of the performance data of the device included in the time interval X ₂ is {x ₂ , x ₃ ,...,x _K+1 }, the performance data of the device included in the time interval X _n-K+1 is {x _n-K+1 , x _n-K+2 ,..., x _n }.

It should be noted that, when acquiring the performance data of the history of the device at each sampling point, an integer multiple of the sampling interval used may be used as the second preset step. For example, the second preset step size may be a sampling interval of 1 time, or a sampling interval of 2 times, which is not limited by the embodiment of the present application.

It should be understood that FIG. 6 merely exemplarily lists the second sliding window increasing along the time when the second preset step is the length of time between one sampling point (ie, one sampling interval). The sliding direction of the direction, the generated time interval, the sliding direction of the time window is not specifically limited in the embodiment of the present application. For example, the sliding direction of the second sliding window may be the starting point of the cycle as the starting point, and the first preset is The direction of the start time of the time zone (the direction from the right to the left in Fig. 6) slides.

It should be noted that, regardless of the manner in which the period is divided into a plurality of time intervals, when the number of sampling points included in the time interval is small, the number of sampling points included in the first time interval determined according to the time interval is also There will be less. At this time, the performance data of the device is sampled by using the first sampling interval in the first time interval, and the reduced amount of data after sampling is limited, and the sampling method using the embodiment of the present application is also increased. The complexity of sampling device performance data is not worth the cost. Therefore, by setting the minimum time length of the time interval to define the minimum value of the number of sampling points included in the time interval, it is advantageous to ensure the data amount of the performance data reduced using the sampling scheme of the embodiment of the present application.

The length of the minimum time interval may be determined based on the length of the period. For example, if the period is 24 hours, the length of the minimum time interval may be set to 15 minutes. If the period is 48 hours, the length of the minimum time interval may be set. It is 30 minutes. The length of the minimum time interval may also be determined based on a sampling interval used when sampling device performance data. For example, if the sampling interval is 5 seconds, the length of the minimum time interval may be set to 10 minutes, if the sampling interval is 30. In seconds, the length of the above minimum time interval can be set to 30 minutes. It is also possible to comprehensively consider the above period and the above sampling interval to determine the length of the minimum time interval. The specific determination manner of the length of the minimum time interval is not limited in the embodiment of the present application.

The length of the minimum time interval described above may be determined based on the length of the above-described period, and specifically may be determined by setting a minimum time length of the time interval and a ratio between the periods. The length of the minimum time interval described above may also be determined based on a sampling interval used when sampling device performance data, and specifically may be determined by setting a minimum time length of the time interval and a ratio between sampling intervals.

After dividing the period into a plurality of time intervals, the method of determining the first time interval is described in detail below. On the one hand, the time interval selected by the first time interval selection rule may be the first time interval, and on the other hand, the initial interval is selected according to the first time interval selection rule, and the extended time interval is determined by expanding the initial interval. For the first time interval.

The determining manner of the first time interval is as follows: the selecting rule of the first time interval comprises: selecting a preset number of time intervals from the plurality of time intervals as the first time interval, wherein the preset number of time intervals preferentially selects the plurality of time intervals The time interval in which the standard deviation of the performance data of the device is small. The preset number of values has a plurality of selection manners. For example, when the preset number of values is 1, the time interval corresponding to the preset standard time interval is the minimum standard deviation of the performance data of the devices in the multiple time intervals. . When the preset number of values is 2, the preset number of time intervals is two time intervals in which the standard deviation of the values of the performance data of the devices in the plurality of time intervals is the smallest.

The first time interval is determined by the second time interval. The selection rule of the first time interval includes selecting a time interval in which the standard deviation is smaller than the standard deviation threshold as the first time interval from the standard deviation set composed of the standard deviations corresponding to the plurality of time intervals.

Specifically, the time interval in which the above standard deviation is smaller than the standard deviation threshold may include one or more time intervals.

It should be noted that the standard deviation threshold may be determined according to the accuracy of the degree of change of the value of the performance data of the device to be sampled, and the accuracy of the change of the value of the performance data of the sampled device. The requirement is higher, and the value of the standard deviation threshold may be set lower, so that the value of the performance data of the device is less changed in the first selected time interval; if the performance data of the sampled device is The accuracy of the change of the value is low, and the value of the standard deviation threshold may be set higher, so that the value of the performance data of the device may be slightly changed in the first selected time interval. . That is to say, the above standard deviation threshold can indicate the degree of change in the value of the performance data of the largest device that the maintenance personnel can accept.

After the first time interval in the period is determined, the remaining time interval of the period may be used as the second time interval. Optionally, the standard deviation corresponding to each second time interval is greater than the standard deviation corresponding to all the first time intervals respectively. .

It should be noted that, if the standard deviation of the values of the performance data in the multiple time intervals is greater than the standard deviation threshold, the degree of change of the value of the performance data of the device is large, and the response of the method in the embodiment of the present application is used. The value of the performance data of the device may not be accurate enough. The performance data of the device may be sampled by using intensive sampling in the prior art.

The initial interval may be selected according to the first time interval selection rule, and the extended time interval is determined as the first time interval by expanding the initial interval.

Specifically, the time interval that satisfies the selection rule of the initial interval is determined as an initial interval based on the standard deviation corresponding to the plurality of time intervals respectively; and the first time interval is determined according to the initial interval, where the first time interval is The performance data includes performance data in the initial interval, and performance data of a device corresponding to the target sampling point, the target sampling point being a sample located outside the initial interval and adjacent to an endpoint of the initial interval Point, the value of the performance data corresponding to the target sampling point and the mean value of the performance data in the initial interval satisfy |e ₁ -m ₀ |<3d ₀ , where e ₁ represents a location corresponding to the target sampling point The value of the performance data of the device, m ₀ represents the mean value of the value of the performance data in the initial interval; d ₀ represents the standard deviation of the value of the performance data in the initial interval.

It should be noted that the selection rule of the initial interval may be the same as the selection rule of the first time interval in the above, that is, the initial interval may be determined in the same manner as the first time interval in the above, for the sake of brevity. No longer.

The following describes how to extend the initial interval, obtain an extended interval, and use the extended interval as the first time interval. The method for expanding the initial interval of the embodiment of the present application is described in detail below with reference to FIG. 7. FIG. 7 is a schematic flowchart of a method for expanding an initial interval according to an embodiment of the present application. After the initial interval is determined, the initial interval is expanded by the method shown in FIG. 7, and the extended interval is obtained as the first time interval. For details, refer to the description of steps 710 to 740 in the method shown in FIG. 7.

710. Determine a mean value m ₀ and a standard deviation d ₀ of the values of the performance data of the device in the initial interval.

For example, the mean value m ₀ and the standard deviation d ₀ of the values of the performance data of the device in the initial interval may be calculated using an Expectation Maximization (EM) algorithm.

720. Determine, according to an initial interval expansion rule, whether performance data of the device corresponding to the target sampling point can be extended to an initial interval to form an extended interval.

When step 720 is performed, it is first determined that the target sampling point is determined, and then based on the initial interval expansion rule, it is determined whether the performance data of the device corresponding to the target sampling point can be expanded to the initial interval.

The manner of determining the target sampling point will be described below with reference to FIG. 8 and the implementation of determining whether the performance data of the device corresponding to the target sampling point can be expanded to the initial interval based on the initial interval expansion rule will be described with reference to FIG.

First, the manner of determining the target sampling point will be described with reference to FIG. Please refer to FIG. 8. FIG. 8 is a schematic diagram showing the position of the initial interval in the cycle in the embodiment of the present application. Contained in the time axis shown in FIG. 8 the value of the performance data of the n devices within a cycle, i.e., n sampling points _{_{t 1, t 2, ... t}} n respectively correspond to the value of the performance data of the device Let x ₁ , x ₂ ,..., x _n , n be a positive integer. Exemplary, the initial starting time interval determined in the time period starting period, and the initial interval comprises L sampling points _{_{t 1, t 2, ... t}} L corresponding to each performance data device The values are x ₁ , x ₂ ,..., x _L .

If the initial interval is at position 1 shown in FIG. 8, the initial interval includes L sample points t ₁ , t ₂ , . . . , t _L , that is, the initial interval includes the start time t _{1 of the} cycle, and therefore, When the initial interval located at position 1 is expanded, the selection of the target sampling point may be selected from the left to the right along the time axis. For example, in the process of expanding the initial interval for the first time, the target sampling point may be select t _{L + 1,} if the value of t _{L + 1} corresponding to the performance data of the device satisfying the preset rule may be extended to the initial interval, extended acquisition interval includes t _{L + 1,} continuing in accordance with the above The method determines whether t _L+2 belongs to the initial interval, and when the value of the performance data of the device corresponding to t _L+2 satisfies the above preset rule, the previous extended interval is expanded, and the extended interval includes not only the previous extension. The added t _L+1 also includes the t _L+2 added by the current extension, and so on, until the value of the performance data of the device corresponding to the sampling point of the initial interval to be expanded does not satisfy the above preset condition, or until the period is to be End time Should be extended to the initial sample point interval, stopping the expansion of said initial section, to form an extended interval.

If the initial section 3 located at the position shown in FIG. 8, which comprises the initial sampling point interval t _{_n ',} ... t _n, that is, the end of the initial interval comprising the time period t _n, and therefore, in a position located 3 When the initial interval is expanded, the selection of the target sampling point may be selected from the right to the left along the time axis. For example, in the process of expanding the initial interval for the first time, the target sampling point may select t _{n '- 1,} when the performance data value t _n'-1 corresponding to the above-described device satisfies preset rule may be extended t _n'-1 to the initial interval, t _n'-2 continues to judge according to the methods described above Whether it belongs to the initial interval, and when the value of the performance data of the device corresponding to t n ' _-2 satisfies the above preset rule, t n ' _-2 may be extended to the initial interval, and so on, until the sampling of the initial interval to be extended The value of the performance data of the device corresponding to the point does not satisfy the above preset condition, or until the sampling point corresponding to the end time of the period is extended to the initial interval, and the expansion of the initial interval is stopped to form an extended interval.

If the initial interval is at position 2 shown in FIG. 8, the initial interval includes sampling points t _l′ , . . . , t _l , that is, the initial interval is located at the middle of the period, that is, the start time of the period is not included. Nor does it include the end of the cycle. Therefore, when the initial interval located at position 2 is expanded, the selection of the target sampling point may be selected from the right to the left direction along the time axis according to the initial interval at position 1 (for example, the target sampling point may be t _{l+ 1} ), it can also be selected from the right to left direction along the time axis according to the initial interval at position 3 (for example, the target sampling point can be t _l'-1 ), and the specific judgment manner is located at the position corresponding to the initial interval above. The judgment mode is the same as that of the position 3, and is not described here for brevity.

It should be noted that, in the embodiment of the present application, when the initial interval is located at the position 2, the time sequence between the selection manners of the target sampling points in the two directions is not limited. For example, the selection of the target sampling point may be first selected from the right to the left along the time axis, and then selected from the right to the left along the time axis; or the selection of the target sampling point may be from the left along the time axis. Select in the right direction and then from left to right along the time axis.

The implementation manner of determining whether the performance data of the device corresponding to the target sampling point belongs to the initial interval based on the initial interval expansion rule is described below with reference to FIG. 8 . It should be understood that the foregoing initial interval expansion rule is used to determine the value of the performance data of the device corresponding to the target sampling point, and the degree of change of the value of the performance data of the device in the initial interval. For example, the initial interval expansion rule may be The difference between the value of the performance data of the device corresponding to the target sampling point and the mean value of the performance data in the initial interval is less than a threshold; or the initial interval expansion may also be the value of the performance data of the device corresponding to the target sampling point. The difference between the maximum values of the performance data in the initial interval is smaller than the difference between the maximum value and the minimum value of the performance data in the initial interval. The specific content of the initial interval expansion is not limited in the embodiment of the present application.

Alternatively, the initial interval expansion described above may be determined based on a small probability event rule. That is, the value of the performance data corresponding to the target sampling point and the mean value of the performance data in the initial interval satisfy |e ₁ -m ₀ |<3d ₀ , where e ₁ represents the value of the performance data corresponding to the target sampling point.

It should be noted that the target sampling point is relative to the end point of the initial interval, and as the initial interval is continuously expanded, the endpoint of the initial interval also changes, and further, the target sampling point changes with the endpoint of the initial interval. Variety.

After determining two different results according to step 720, there are two different implementations. Specifically, if the performance data of the device corresponding to the target sampling point can be extended to the initial interval, step 730 is performed; if the performance data of the device corresponding to the target sampling point cannot be extended to the initial interval, step 740 is performed.

730: Extend performance data of the device corresponding to the target sampling point to an initial interval to form an extended interval.

That is to say, the endpoint of the extended interval changes to the target sampling point.

740. Stop expanding the initial interval, and determine that the current time interval is the first time interval.

It should be noted that the current time interval may be an unexpanded extended interval, that is, an initial interval; the current initial interval may also be an extended interval.

After determining the first preset time period or all the first time intervals in the period, the first preset time period or the remaining time interval in the period may be used as the second time interval.

Optionally, any two of the plurality of first time intervals in the period do not overlap each other.

230. Determine a first sampling interval when sampling performance data in the first time interval, and a second sampling interval when sampling performance data in the second time.

The first sampling interval and the second sampling interval may be preset, or may be determined based on the magnitude of the change of the value of the performance data in the first time interval or the second time interval. Make a limit. The following section focuses on how to determine the first time interval and the second time interval within a period.

Specifically, the length of the first sampling interval is inversely proportional to the magnitude of the change in the value of the performance data of the device in the first time interval, and/or the length of the second sampling interval is different from the first The magnitude of the change in the value of the performance data of the device in the two time interval is inversely proportional.

The first sampling interval and the second sampling interval may be preset, and it may be understood that the first sampling interval and the second sampling interval need not be changed according to the value of the performance data in the first time interval or the second time interval. The size is calculated. For example, after determining that the certain time period is the first time interval, the performance data of the device in the first time interval is directly sampled by using the preset first sampling interval; after determining that the certain time period is the second time interval, directly The performance data of the device in the second time interval is sampled using a preset second sampling interval.

The first sampling interval and the second sampling interval may be determined based on the magnitude of the change of the value of the performance data in the first time interval or the second time interval, and may be understood as required by the first sampling interval and the second sampling interval. The calculation is performed according to the magnitude of the change in the value of the performance data in the first time interval or the second time interval. Optionally, an inverse relationship between the length of the first sampling interval and the magnitude of the change in the value of the performance data of the device in the first time interval, and/or the length of the second sampling interval The inverse relationship between the magnitudes of changes in the values of the performance data of the device in the second time interval respectively satisfies

Specifically, the first time interval and the second time interval may be any time interval within the period, that is, the degree of change of the value of the performance data of the device in any time interval within the period, and The inverse relationship between the lengths of the sampling intervals corresponding to the time interval can be satisfied.

In the embodiment of the present application, based on an inverse relationship between the degree of change of the value of the performance data of the device in the time interval and the length of the sampling interval, the time used for sampling the performance data of the device in the time interval may be determined. interval. The sampling data obtained by the sampling interval determined by such a method is advantageous for more accurately reflecting the change of the value of the performance data of the device with time.

240. The performance data of the device in the first time interval is sampled at a first sampling interval in a first time interval.

250. The performance data of the device in the second time interval is sampled at a second sampling interval in a second time interval, wherein the first sampling interval is greater than the second sampling interval.

Optionally, as an embodiment, the method further includes: determining whether the method for sampling the performance data of the device needs to be re-executed by verifying the accuracy of the first time interval.

Specifically, the performance data of the device in the current first time interval is monitored in real time as the time interval to be verified. In the first time interval to be verified, the performance data obtained by real-time sampling the performance data of the device, the mean value m ₀ ' and the standard deviation d ₀ ' of the performance data of the device are determined, and based on the small probability time In principle, it is determined whether the value e _i ' of the performance data of the device corresponding to each sampling point in the first time interval to be verified satisfies the formula |e _i '-m ₀ '|<3d ₀ ', where i is positive Integer.

If the value of the performance data of the device corresponding to each sampling point in the first time interval to be verified satisfies the above formula, the degree of change of the performance data of the device in the current first time interval is small, and the judgment is small. The currently used sampling method for the performance data of the device does not need to be updated. If the value of the performance data of the device corresponding to each sampling point in the first time interval to be verified does not satisfy the above formula, the device in the current first time interval The value of the performance data varies greatly, and it is no longer suitable to use the first sampling interval for sampling.

At this time, by repeating the above steps 210 to 250, the cycle of the performance data of the device may be re-determined, or the distribution pattern of the first time interval and the second time interval in the cycle may be re-determined, or each of the periodic periods may be re-determined. The sampling interval used when sampling performance data in the time interval.

It should be noted that the embodiment of the present application may also omit the verification of the accuracy of the first time interval, but periodically re-execute steps 210 to 250 to implement a cycle of re-determining the performance data of the device, or re-determine the period. The distribution of a time interval and a second time interval, or the purpose of re-determining the sampling interval used to sample performance data for each time interval of the cycle.

Optionally, as an embodiment, the performance data of the device is any one of the following data: an occupancy rate of the computing resource in the device, an occupancy rate of the transmission resource in the device, and a storage resource in the device. The method further includes: determining a third time interval in the first preset time period, wherein an average value of the performance data of the device is higher than a performance data threshold in the third time interval; During the time interval, the performance data in the third time interval is sampled at a third sampling interval, the third sampling interval being smaller than the first sampling interval.

Specifically, the average value of the performance data of the device in the third time interval is higher than the performance data threshold, which may indicate that the performance data of the current device is at a peak period, and during the peak period, the probability of abnormal operation of the device is higher. Large, therefore, in order to increase the basis for subsequent analysis of the performance of the device, the performance data in the third time interval is used regardless of the degree of change in the value of the performance data in the third time interval. Sampling is performed, and the third sampling interval is equal to or greater than the second sampling interval, so that more performance data of the device is collected in the third time interval.

The method for sampling the performance data of the device in the embodiment of the present application is described in detail above with reference to FIG. 1 to FIG. 8. The following describes the performance data of the device in the embodiment of the present application in detail with reference to FIG. 9 and FIG. Device. It should be noted that the apparatus shown in FIG. 9 and FIG. 10 can implement various steps in the above method, for example, step 210-step 250 in FIG. For the sake of brevity, it will not be repeated here.

FIG. 9 is a schematic block diagram of an apparatus for transmitting data according to an embodiment of the present application. The apparatus 900 for transmitting data shown in FIG. 9 includes a first determining unit 910 and a sampling unit 920.

The first determining unit 910 is configured to determine a first time interval and a second time interval in the first preset time period, where the value of the performance data of the device is less than Determining the degree of change in the value of the performance data of the device in the second time interval;

The sampling unit 920 is configured to sample performance data of the device in the first time interval at the first sampling interval in the first time interval;

The sampling unit 920 is further configured to sample, in the second time interval, performance data of the device in the second time interval at a second sampling interval, where the first sampling interval is greater than The second sampling interval is described.

Optionally, as an embodiment, the performance data of the device is data that changes with time in time, and the first determining unit is further configured to: according to the performance data of the device in the second preset time period. Determining the period of the value as a function of time; and determining a first time interval and a second time interval within the period, the period being the first predetermined time period.

Optionally, as an embodiment, the first determining unit is further configured to: divide the period into multiple time intervals, and determine the device in each time interval of the multiple time intervals. a standard deviation of the performance data; a time interval in which the standard deviation is smaller than the standard deviation threshold is selected as the first time interval from the standard deviation set consisting of the standard deviations corresponding to the plurality of time intervals; In the standard deviation set composed of the standard deviations corresponding to the intervals, a time interval in which the standard deviation is greater than or equal to the standard deviation threshold is selected as the second time interval.

Optionally, as an embodiment, the first determining unit is further configured to: divide the period into multiple time intervals, and determine the device in each time interval of the multiple time intervals. a standard deviation of the performance data; determining, according to the standard deviation corresponding to the plurality of time intervals, a time interval that satisfies the selection rule of the initial interval as an initial interval; determining the first time interval according to the initial interval, the first The performance data in a time interval includes performance data in the initial interval, and performance data of a device corresponding to the target sampling point, the target sampling point being outside the initial interval and ending with the initial interval Adjacent sampling points, the value of the performance data corresponding to the target sampling point and the mean value of the performance data in the initial interval satisfy |e ₁ -m ₀ |<3d ₀ , where e ₁ represents the target sampling the value of the device performance data corresponding to the point, m ₀ represents the mean value of the initial interval performance data; d ₀ represents the initial value of the performance data section Standard deviation.

Optionally, as an embodiment, in the first time interval, the magnitude of the change in the value of the performance data of the device is inversely proportional to the length of the first sampling interval, and/or

During the second time interval, the magnitude of the change in the value of the performance data of the device is inversely proportional to the length of the second sampling interval.

Optionally, as an embodiment, an inverse relationship between a magnitude of a change in the value of the performance data of the device and a length of the first sampling interval in the first time interval, and

In the second time interval, an inverse relationship between the magnitude of the change in the value of the performance data of the device and the length of the second sampling interval is satisfied.

Optionally, as an embodiment, the device further includes: an acquiring unit, configured to acquire a trend of the value of the performance data of the device in the second preset time period, and the The second preset time period is divided into N sub-time segments, where N is a positive integer; the second determining unit is further configured to use the value of the performance data of the device in the i-th sub-period within the N sub-periods The change trend is a reference, determining a change trend of the value of the device performance data in the jth sub-period in the N sub-periods and a value of the performance data of the device in the i-th sub-period The degree of similarity between the trends of change, j ∈ [1, N], i ∈ [1, N], j ≠ i, and j and i are positive integers; a selection unit for the i-th sub-period Selecting the i-th sub-period in the degree of similarity between the change trend of the value of the performance data of the device and the change trend of the value of the performance data of the device in the j-th sub-period The target time period in which the value of the performance data of the device is the same The second determining unit is further configured to determine a length of time between the first sampling point in the ith sub-time period and a second sampling point in the target time period, where the first sampling point is in the The position in the i-th sub-period is the same as the position of the second sampling point in the target time period; the second determining unit is further configured to select any sampling point in the second preset time period For the start time of the period, the period is determined by the length of time between the first sampling point and the second sampling point being the length of the period.

Optionally, as an embodiment, a change trend of the value of the performance data of the device in the i-th sub-period and a change trend of the value of the performance data of the device in the j-th sub-period The degree of similarity between s _ij and where

Optionally, as an embodiment, the first determining unit is further configured to determine a third time interval in the first preset time period, where performance data of the device is in the third time interval. The mean is higher than the performance data threshold;

The sampling unit is further configured to sample performance data in the third time interval at a third sampling interval in a third time interval, where the third sampling interval is smaller than the first sampling interval.

Optionally, as an embodiment, the performance data of the device is any one of the following data: an occupancy rate of the computing resource in the device, an occupancy rate of the transmission resource in the device, and a storage resource in the device. Occupancy rate.

In an optional embodiment, the foregoing apparatus 900 for sampling performance data of the device may be the apparatus 1000 for sampling performance data of the device, where the first determining unit 910 and the sampling unit 920 may be processors in the device 1000. 1020, the device 1000 can further include an input/output interface 1030 and a memory 1010, as shown in FIG.

FIG. 10 is a schematic block diagram of an apparatus for sampling performance data of a device according to another embodiment of the present application. The apparatus 1000 for sampling performance data of a device shown in FIG. 10 may include a memory 1010, a processor 1020, and an input/output interface 1030. The memory 1010, the processor 1020, and the input/output interface 1030 are previously connected by an internal connection path for storing program instructions, and the processor 1020 is configured to execute program instructions stored by the memory 1020 to control input/output. The interface 1030 receives input data and information, and outputs data such as an operation result.

It should be understood that, in the embodiment of the present application, the processor 1020 may be a general-purpose central processing unit (CPU), a microprocessor, an application specific integrated circuit (ASIC), or one or more. An integrated circuit for performing related procedures to implement the technical solutions provided by the embodiments of the present application.

The memory 1010 can include read only memory and random access memory and provides instructions and data to the processor 1020. A portion of processor 1020 may also include a non-volatile random access memory. For example, the processor 1020 can also store information of the device type.

In the implementation process, each step of the above method may be completed by an integrated logic circuit of hardware in the processor 1020 or an instruction in a form of software. The method for sampling the performance data of the device disclosed in the embodiment of the present application may be directly implemented by the hardware processor, or may be performed by a combination of hardware and software modules in the processor. The software module can be located in a conventional storage medium such as random access memory, flash memory, read only memory, programmable read only memory or electrically erasable programmable memory, registers, and the like. The storage medium is located in the memory 1010, and the processor 1020 reads the information in the memory 1010 and completes the steps of the above method in combination with its hardware. To avoid repetition, it will not be described in detail here.

It should be understood that, in this embodiment of the present application, the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSPs), and dedicated integration. Application specific integrated circuit (ASIC), field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware component, etc. The general purpose processor may be a microprocessor or the processor or any conventional processor or the like.

It should be understood that B can be determined from A. However, it should also be understood that determining B from A does not mean that B is only determined based on A, and that B can also be determined based on A and/or other information.

It should be understood that, in the various embodiments of the present application, the size of the sequence numbers of the foregoing processes does not mean the order of execution sequence, and the order of execution of each process should be determined by its function and internal logic, and should not be applied to the embodiment of the present application. The implementation process constitutes any limitation.

In the several embodiments provided by the present application, it should be understood that the disclosed apparatus, apparatus, and method may be implemented in other manners. For example, the device embodiments described above are merely illustrative. For example, the division of the unit is only a logical function division. In actual implementation, there may be another division manner, for example, multiple units or components may be combined or Can be integrated into another device, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be in an electrical, mechanical or other form.

The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of the embodiment.

In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above integrated modules can be implemented in the form of hardware or in the form of hardware plus software function modules.

In the above embodiments, it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented in software, it may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in accordance with embodiments of the present application are generated in whole or in part. The computer can be a general purpose computer, a special purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer readable storage medium, which can be any available media that can be read by a computer or a data storage device such as a server, data center, or the like that includes one or more available media integrations. . The usable medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a Digital Video Disc (DVD)), or a semiconductor medium (eg, a Solid State Disk (SSD)). )Wait.

The foregoing is only a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto, and any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application. It should be covered by the scope of protection of this application. Therefore, the scope of protection of the present application should be determined by the scope of the claims.

Claims

A method for sampling performance data of a device, comprising:

Determining a first time interval and a second time interval in the first preset time period, wherein a degree of change in the value of the performance data of the device in the first time interval is less than that in the second time interval The degree of change in the value of the performance data of the device;

Capturing performance data of the device in the first time interval at the first sampling interval in the first time interval;

And during the second time interval, the performance data of the device in the second time interval is sampled at a second sampling interval, wherein the first sampling interval is greater than the second sampling interval.
The method of claim 1, wherein the performance data of the device is data that changes with time in a cycle, the method further comprising:

Determining the period according to a trend of the value of the performance data of the device in the second preset time period;

Determining the first time interval and the second time interval in the first preset time period, including:

Determining a first time interval and a second time interval within the period, the first predetermined time period being the period.
The method of claim 2, wherein the determining the first time interval and the second time interval in the period comprises:

Dividing the period into a plurality of time intervals, and determining a standard deviation of performance data of the device in each time interval of the plurality of time intervals;

Selecting, from the standard deviation set consisting of the standard deviations corresponding to the plurality of time intervals, a time interval in which the standard deviation is smaller than the standard deviation threshold as the first time interval;

A time interval in which a standard deviation is greater than or equal to the standard deviation threshold is selected as the second time interval from a standard deviation set consisting of standard deviations corresponding to the plurality of time intervals.
The method of claim 2, wherein the determining the first time interval and the second time interval in the period comprises:

Dividing the period into a plurality of time intervals, and determining a standard deviation of performance data of the device in each time interval of the plurality of time intervals;

Determining, according to a standard deviation corresponding to the plurality of time intervals, a time interval that satisfies a selection rule of the initial interval as an initial interval;

Determining, according to the initial interval, the first time interval, where the performance data in the first time interval includes performance data in the initial interval, and performance data of a device corresponding to the target sampling point, where the target sampling point is a sampling point located outside the initial interval and adjacent to an endpoint of the initial interval, where the value of the performance data corresponding to the target sampling point and the mean value of the performance data in the initial interval satisfy |e 1 - m 0 | <3d 0, wherein the value of the device performance data represents e 1 corresponding to the target sampling point, m 0 represents the mean value of the initial interval performance data; d 0 represents the The standard deviation of the value of the performance data in the initial interval.
The method according to any one of claims 1 to 4, characterized in that, in the first time interval, the magnitude of the change in the value of the performance data of the device and the length of the first sampling interval In inverse relationship, and / or

In the second time interval, the magnitude of the change in the value of the performance data of the device is inversely proportional to the length of the second sampling interval.
The method according to claim 5, wherein in the first time interval, an inverse relationship between the magnitude of the change in the value of the performance data of the device and the length of the first sampling interval, as well as

In the second time interval, an inverse relationship between the magnitude of the change in the value of the performance data of the device and the length of the second sampling interval is satisfied.
Wherein P represents the length of the first sampling interval or the second sampling interval, k is a preset number greater than 1, and σ represents the device in the first time interval or the second time interval The standard deviation of the performance data is used to indicate the degree of change of the value of the performance data in the first time interval or the second time interval.
The method according to any one of claims 2-6, wherein the determining the period according to a trend of the value of the performance data of the device in the second preset time period, including :

Obtaining a trend of the value of the performance data of the device in the second preset time period with time, and dividing the second preset time period into N sub-time segments, where N is a positive integer;

Determining, according to a change trend of the value of the performance data of the device in the i-th sub-period of the N sub-periods, determining the device performance data in the j-th sub-period within the N sub-periods The degree of similarity between the change trend of the value and the change trend of the value of the performance data of the device in the i-th sub-period, j ∈ [1, N], i ∈ [1, N], J≠i, and j and i are positive integers;

Selecting a degree of similarity between a change trend of the value of the performance data of the device in the i-th sub-period and a change trend of the value of the performance data of the device in the j-th sub-period a target time period that is the same as a change trend of the value of the performance data of the device in the i-th sub-period;

Determining a length of time between the first sampling point in the i-th sub-period and a second sampling point in the target time period, where the first sampling point is located in the i-th sub-period The second sampling point has the same position in the target time period;

Selecting any sampling point in the second preset time period as the start time of the period, and determining a length of the period between the first sampling point and the second sampling point as the length of the period, The period.
The method according to claim 7, wherein the change trend of the value of the performance data of the device in the i-th sub-period and the performance data of the device in the j-th sub-period The degree of similarity between the trends of the values is s ij , where
Y represents the number of sampling points for sampling the performance data of the device included in the i-th sub-period and the j-th sub-period, and t iy represents the yth device in the i-th sub-period The value of the performance data, t jy represents the value of the performance data of the yth device in the jth sub-time period,
Means the mean value of the performance data of the device in the i-th sub-period,
Indicates the mean value of the performance data of the device in the jth sub-period.
The method of any of claims 1-8, wherein the method further comprises:

Determining a third time interval in the first preset time period, wherein an average value of the performance data of the device is higher than a performance data threshold in the third time interval;

In a third time interval, performance data in the third time interval is sampled at a third sampling interval, the third sampling interval being less than the first sampling interval.
The method according to any one of claims 1 to 9, wherein the performance data of the device is any one of the following data: an occupancy rate of a computing resource in the device, and a resource in the device Occupancy, the occupancy rate of storage resources in the device.
An apparatus for sampling performance data of a device, comprising:

a first determining unit, configured to determine a first time interval and a second time interval in the first preset time period, where the value of the performance data of the device changes less than The degree of change in the value of the performance data of the device in the second time interval;

a sampling unit, configured to sample, in the first time interval, performance data of the device in the first time interval at a first sampling interval, and in the second time interval, at a second sampling interval The performance data of the device in the second time interval is sampled, wherein the first sampling interval is greater than the second sampling interval.
The device according to claim 11, wherein the performance data of the device is data that changes with time in time.

The first determining unit is further configured to determine the period according to a trend of the value of the performance data of the device in the second preset time period; and

Determining a first time interval and a second time interval within the period, the period being the first predetermined time period.
The device according to claim 12, wherein the first determining unit is further configured to:

Dividing the period into a plurality of time intervals, and determining a standard deviation of performance data of the device in each time interval of the plurality of time intervals;

Selecting, from the standard deviation set consisting of the standard deviations corresponding to the plurality of time intervals, a time interval in which the standard deviation is smaller than the standard deviation threshold as the first time interval;

A time interval in which a standard deviation is greater than or equal to the standard deviation threshold is selected as the second time interval from a standard deviation set consisting of standard deviations corresponding to the plurality of time intervals.
The device according to claim 12, wherein the first determining unit is further configured to:

Dividing the period into a plurality of time intervals, and determining a standard deviation of performance data of the device in each time interval of the plurality of time intervals;

Determining, according to a standard deviation corresponding to the plurality of time intervals, a time interval that satisfies a selection rule of the initial interval as an initial interval;

Determining, according to the initial interval, the first time interval, where the performance data in the first time interval includes performance data in the initial interval, and performance data of a device corresponding to the target sampling point, where the target sampling point is a sampling point located outside the initial interval and adjacent to an endpoint of the initial interval, where the value of the performance data corresponding to the target sampling point and the mean value of the performance data in the initial interval satisfy |e 1 - m 0 | <3d 0, wherein the value of the device performance data represents e 1 corresponding to the target sampling point, m 0 represents the mean value of the initial interval performance data; d 0 represents the The standard deviation of the value of the performance data in the initial interval.
The apparatus according to any one of claims 11 to 14, wherein the magnitude of the change in the value of the performance data of the device and the length of the first sampling interval in the first time interval In inverse relationship, and / or

During the second time interval, the magnitude of the change in the value of the performance data of the device is inversely proportional to the length of the second sampling interval.
The apparatus according to claim 15, wherein in the first time interval, an inverse relationship between a magnitude of a change in the value of the performance data of the device and a length of the first sampling interval, as well as

In the second time interval, an inverse relationship between the magnitude of the change in the value of the performance data of the device and the length of the second sampling interval is satisfied.
Wherein P represents the length of the first sampling interval or the second sampling interval, k is a preset number greater than 1, and σ represents the device in the first time interval or the second time interval The standard deviation of the performance data is used to indicate the degree of change of the value of the performance data in the first time interval or the second time interval.
The device of any of claims 12-16, wherein the device further comprises:

An acquiring unit, configured to acquire a trend of the value of the performance data of the device in the second preset time period, and divide the second preset time segment into N sub-time segments, where N is positive Integer

a second determining unit, configured to determine, according to a change trend of the value of the performance data of the device in the i-th sub-period within the N sub-periods, the j-th sub-time in the N sub-time segments The degree of similarity between the change trend of the value of the device performance data in the segment and the change trend of the value of the performance data of the device in the i-th sub-period, j∈[1,N], I∈[1,N], j≠i, and j and i are positive integers;

a selecting unit, configured to change a trend between a change trend of the value of the performance data of the device in the i-th sub-period and a change trend in a value of the performance data of the device in the j-th sub-period a degree of similarity, selecting a target time period that is the same as a change trend of the value of the performance data of the device in the i-th sub-period;

The second determining unit is further configured to determine a length of time between the first sampling point in the i-th sub-period and the second sampling point in the target time period,

Selecting any sampling point in the second preset time period as the start time of the period, and determining a length of the period between the first sampling point and the second sampling point as the length of the period, The period, the position of the first sampling point in the ith sub-period is the same as the position of the second sampling point in the target time period.
The apparatus according to claim 17, wherein the change trend of the value of the performance data of the device in the i-th sub-period and the performance data of the device in the j-th sub-period The degree of similarity between the trends of the values is s ij , where
Y represents the number of sampling points for sampling the performance data of the device included in the i-th sub-period and the j-th sub-period, and t iy represents the yth device in the i-th sub-period The value of the performance data, t jy represents the value of the performance data of the yth device in the jth sub-time period,
Means the mean value of the performance data of the device in the i-th sub-period,
Indicates the mean value of the performance data of the device in the jth sub-period.
A device according to any of claims 11-18, wherein

The first determining unit is further configured to determine a third time interval in the first preset time period, wherein an average value of the performance data of the device is higher than a performance data threshold in the third time interval;

The sampling unit is further configured to sample performance data in the third time interval at a third sampling interval in a third time interval, where the third sampling interval is smaller than the first sampling interval.
The device according to any one of claims 11 to 19, wherein the performance data of the device is any one of the following data: an occupancy rate of a computing resource in the device, and a resource in the device Occupancy, the occupancy rate of storage resources in the device.
An apparatus for sampling performance data of a device, comprising: an input/output interface, a processor, and a memory, wherein the processor is configured to control the input/output interface to transmit and receive information, and the memory is used to store a computer A program for calling and running the computer program from the memory, such that the apparatus performs the method of any one of claims 1 to 10.
A computer readable medium, wherein the computer readable medium stores program code, when the computer program code is run on a computer, causing the computer to perform the method of any one of claims 1 to 10. Methods.