CN-122001782-A - Method for analyzing predictability of metrics from monitoring agents in a computer network

Abstract

The present disclosure relates to a method for analyzing predictability of metrics from monitoring agents in a computer network. Systems, methods, apparatuses, and computer program products for predicting network metrics are disclosed. A method may include a computing device receiving metric data, extracting at least one time series attribute from the metric data, calculating at least one statistical measure based on the at least one time series attribute, performing an integration score based on the at least one statistical measure, generating a predictability score based on the integration score, and determining predictability of the metric data based on a comparison of the predictability score to a preset threshold.

Inventors

• K .sai
• V. D. Papu
• KINI ANANTH
• S. M.N. Ugandan

Assignees

• 太阳风环球有限责任公司

Dates

Publication Date

20260508

Application Date

20251107

Priority Date

20241108

Claims (20)

1. 1. A method, comprising: receiving, by the computing device, metric data; Extracting, by the computing device, at least one time-series attribute from the metric data; calculating, by the computing device, at least one statistical measure based on the at least one time-series attribute; performing, by the computing device, an integration score based on the at least one statistical measure; Generating, by the computing device, a predictability score based on the integrated score, and Predictability of the metric data is determined by the computing device based on a comparison of the predictability score to a preset threshold.
2. 2. The method of claim 1, further comprising: Metric prediction is prevented by the computing device upon determining that the metric data is unpredictable.
3. 3. The method of claim 1, further comprising: Metric prediction is enabled by the computing device upon determining that the metric data is predictable.
4. 4. The method of claim 1, further comprising: At least one periodic predictability check is performed by the computing device.
5. 5. The method of claim 1, wherein the metric data comprises at least one of: output utilization percentage; Average load; Average percentage of memory usage; average response time; Average total input/output operations per second; Average read input/output operations per second; average write input/output operations per second; average total delay; average read delay; Average write latency; Percentage of disk usage, and Any other type of metric data.
6. 6. The method of claim 1, wherein the at least one time series attribute comprises at least one of: Trend; Seasonal; Circularity; Peak and Low valleys.
7. 7. The method of claim 1, wherein the at least one statistical measure comprises at least one of: Degree of dispersion; Spectral density; Residual variability and Omega score.
8. 8. The method of claim 1, wherein calculating the at least one statistical measure further comprises: Calculating, by the computing device, a spectrum of at least one signal; Calculating, by the computing device, a power spectral density of the at least one signal via squaring an amplitude of the at least one signal; normalizing, by the computing device, the at least one signal by a number of bins; Normalizing, by the computing device, the calculated power spectral density to a probability density function; Calculating, by the computing device, a power spectral entropy according to a standard entropy calculation formula, and Determining, by the computing device, whether the normalized power spectral density indicates a predictable sequence or an unpredictable noise sequence.
9. 9. The method of claim 7, further comprising: Calculating by the computing device the Omega score from a subtraction of the power spectral entropy, Wherein the Omega score is calculated to be less than an indication of predictability.
10. 10. An apparatus, comprising: At least one processor, and At least one memory storing instructions that, when executed by the at least one processor, cause the apparatus at least to: receiving metric data; extracting at least one time series attribute from the metric data; Calculating at least one statistical measure based on the at least one time series attribute; performing an integration score based on the at least one statistical measure; Generating a predictability score based on the integrated score, and Predictability of the metric data is determined based on a comparison of the predictability score to a preset threshold.
11. 11. The apparatus of claim 10, wherein the at least one memory and the instructions, when executed by the at least one processor, further cause the apparatus at least to: Metric prediction is prevented upon determining that the metric data is unpredictable.
12. 12. The apparatus of claim 10, wherein the at least one memory and the instructions, when executed by the at least one processor, further cause the apparatus at least to: Metric prediction is enabled upon determining that the metric data is predictable.
13. 13. The apparatus of claim 10, wherein the at least one memory and the instructions, when executed by the at least one processor, further cause the apparatus at least to: At least one periodic predictability check is performed.
14. 14. The apparatus of claim 10, wherein the metric data comprises at least one of: output utilization percentage; Average load; Average percentage of memory usage; average response time; average total input/output operands per second; average read input/output operands per second; An average per second write input/output operand; average total delay; average read delay; Average write latency; Percentage of disk usage, and Any other type of metric data.
15. 15. The apparatus of claim 10, wherein the at least one time series attribute comprises at least one of: Trend; Seasonal; Circularity; Peak and Low valleys.
16. 16. The apparatus of claim 10, wherein the at least one statistical measure comprises at least one of: Degree of dispersion; Spectral density; Residual variability and Omega score.
17. 17. The apparatus of claim 10, wherein calculating the at least one statistical measure further comprises: Calculating a frequency spectrum of at least one signal; Calculating a power spectral density of the at least one signal via squaring an amplitude of the at least one signal; Normalizing the at least one signal by the number of bins; Normalizing the calculated power spectral density to a probability density function; calculating the power spectrum entropy according to the standard entropy calculation formula and It is determined whether the normalized power spectral density indicates a predictable sequence or an unpredictable noise sequence.
18. 18. The apparatus of claim 16, wherein the at least one memory and the instructions, when executed by the at least one processor, further cause the apparatus at least to: The Omega score is calculated from a subtraction of the power spectral entropy, Wherein the Omega score is calculated to be less than an indication of predictability.
19. 19. A non-transitory computer readable medium comprising program instructions that, when executed by an apparatus, cause the apparatus to perform at least one method comprising: receiving metric data; extracting at least one time series attribute from the metric data; Calculating at least one statistical measure based on the at least one time series attribute; performing an integration score based on the at least one statistical measure; Generating a predictability score based on the integrated score, and Predictability of the metric data is determined based on a comparison of the predictability score to a preset threshold.
20. 20. The non-transitory computer-readable medium of claim 19, wherein the apparatus is further caused to perform: Metric prediction is prevented upon determining that the metric data is unpredictable.

Description

Method for analyzing predictability of metrics from monitoring agents in a computer network Technical Field Some example embodiments may generally relate to predicting network metrics between monitoring agents in a computer network. Background Metric prediction is an important aspect of modern Information Technology (IT) environments. Such metric predictions enable IT systems to take proactive measures for future consumption patterns of computer resources. However, due to various factors (including uncertainty in behavior patterns and external influences), predictions may be unreliable. Disclosure of Invention According to some example embodiments, a method may include receiving, by a computing device, metric data. The method may also include extracting, by the computing device, at least one time-series attribute from the metric data. The method may also include calculating, by the computing device, at least one statistical measure based on the at least one time-series attribute. The method may also include performing, by the computing device, an integration score based on the at least one statistical measurement. The method may also include generating, by the computing device, a predictability score based on the integrated score. The method may also include determining, by the computing device, predictability of the metric data based on a comparison of the predictability score to a preset threshold. According to certain example embodiments, an apparatus may include means for receiving metrology data. The apparatus may further comprise means for extracting at least one time series attribute from the metrology data. The apparatus may further include means for calculating at least one statistical measure based on the at least one time series attribute. The apparatus may also include means for performing an integration score based on the at least one statistical measure. The apparatus may also include means for generating a predictability score based on the integrated score. The apparatus may also include means for determining predictability of the metric data based on a comparison of the predictability score to a preset threshold. According to various example embodiments, a non-transitory computer readable medium may include program instructions that, when executed by an apparatus, cause the apparatus to perform at least one method. The method may include receiving metric data. The method may further include extracting at least one time series attribute from the metric data. The method may further include calculating at least one statistical measure based on the at least one time series attribute. The method may further include performing an integration score based on the at least one statistical measure. The method may further include generating a predictability score based on the integrated score. The method may further include determining predictability of the metric data based on a comparison of the predictability score to a preset threshold. According to some example embodiments, a computer program product may perform a method. The method may include receiving metric data. The method may further include extracting at least one time series attribute from the metric data. The method may further include calculating at least one statistical measure based on the at least one time series attribute. The method may further include performing an integration score based on the at least one statistical measure. The method may further include generating a predictability score based on the integrated score. The method may further include determining predictability of the metric data based on a comparison of the predictability score to a preset threshold. According to some example embodiments, an apparatus may include at least one processor and at least one memory storing instructions that, when executed by the at least one processor, may also cause the apparatus to at least receive metrology data. The at least one memory and the instructions, when executed by the at least one processor, may also cause the apparatus to at least receive metrology data. The at least one memory and the instructions, when executed by the at least one processor, may also cause the apparatus to calculate at least one statistical measure based at least on the at least one time series attribute. The at least one memory and the instructions, when executed by the at least one processor, may also cause the apparatus to calculate at least one statistical measure based at least on the at least one time series attribute. The at least one memory and the instructions, when executed by the at least one processor, may also cause the apparatus to perform integration scoring based at least on the at least one statistical measure. The at least one memory and the instructions, when executed by the at least one processor, may also cause the apparatus to generate a predictability score based at least on the integrated score. The at least one memory and the i