Performance data analysis method for storage device and performance data analysis tool thereof

Abstract

The invention discloses a performance data analysis method and a performance data analysis tool for storage equipment, and relates to the technical field of performance analysis of the storage equipment. The method comprises the steps of obtaining performance data and a preset performance fluctuation table of the storage device; when the performance data is inconsistent with the preset data format, storing target performance data obtained by performing keyword screening on the performance data into a preset format data table, and automatically drawing a performance curve according to the obtained target data table; performing performance fluctuation analysis on the performance curve according to a target balancing algorithm determined by the type of the performance test and a preset performance fluctuation table to obtain a performance fluctuation analysis condition; when the performance fluctuation analysis condition is that abnormal fluctuation data exists, marking the abnormal fluctuation data on the performance curve; and when the performance fluctuation analysis condition is that data with abnormal fluctuation does not exist, prompting that the performance curve is normal. The performance curve of the storage device can be automatically drawn and analyzed, and the analysis efficiency of the performance data of the storage device is improved.

Description

Technical Field

[0001] The present application relates to the field of flash memory technology, and in particular to a method for analyzing performance data of a storage device and a performance data analysis tool thereof. Background Art

[0002] Currently, the market demand for mobile phones is growing, and consumers have higher and higher performance requirements for mobile phones. Therefore, firmware performance has also become a hard indicator that the storage industry needs to consider. How to quickly analyze whether there are defects in firmware performance has become a technical issue that all original equipment manufacturers in the industry pay attention to.

[0003] In the related art, most of the performance curves are obtained by manual drawing through Excel. The manual drawing method of performance curves is very time-consuming, and after obtaining the performance curve, it is still necessary to manually analyze it to see whether the curve meets the standards, and the analysis efficiency is low. Although some current curve drawing tools can draw curves based on input data, it is necessary to manually screen out appropriate data and input it into the curve drawing tool before the curve can be drawn. This is inconvenient to use and the drawing efficiency is low. The curve drawing tool only provides a drawing function and does not have an analysis function. Manual analysis is still required, and it is difficult to quickly analyze whether the firmware performance has defects. Therefore, how to improve the efficiency of performance data analysis is a technical problem that needs to be solved urgently. Summary of the Invention

[0004] This application aims to solve at least one of the technical problems existing in the prior art. To this end, this application proposes a method for analyzing performance data of a storage device and a performance data analysis tool thereof, which can automatically draw and analyze the performance curve of the storage device, thereby improving the efficiency of analyzing the performance data of the storage device.

[0005] In a first aspect, an embodiment of the present application provides a method for analyzing performance data of a storage device, which is applied to a performance data analysis tool. The method includes:

[0006] Acquire performance data and a preset performance fluctuation table of a storage device; wherein different performance data are obtained by performing different types of performance tests on the storage device;

[0007] When the performance data is inconsistent with the preset data format, the target performance data obtained after keyword screening of the performance data is stored in the preset format data table to obtain a target data table;

[0008] Automatically draw a performance curve based on the target data table;

[0009] Determining a target trade-off algorithm according to the type of the performance test, performing a performance fluctuation analysis on the performance curve according to the target trade-off algorithm and the preset performance fluctuation table to obtain a performance fluctuation analysis result; wherein the performance fluctuation analysis result includes: data with abnormal fluctuations and data without abnormal fluctuations;

[0010] When the performance fluctuation analysis result shows that abnormal fluctuation data exists, the abnormal fluctuation data is marked on the performance curve; when the performance fluctuation analysis result shows that abnormal fluctuation data does not exist, it is indicated that the performance curve is normal.

[0011] In a second aspect, an embodiment of the present application provides a performance data analysis tool for a storage device, which is used to execute a performance data analysis method for a storage device as described in any one of the embodiments of the first aspect.

[0012] The embodiment of the present application includes: when analyzing the performance data of a storage device, first, obtaining the performance data of the storage device and a preset performance fluctuation table through a performance data analysis tool for the storage device; wherein, different performance data are obtained by performing different types of performance tests on the storage device; secondly, when the performance data is inconsistent with the preset data format, storing the target performance data obtained after keyword screening of the performance data into a preset format data table to obtain a target data table; then, automatically drawing a performance curve according to the target data table; obtaining a performance curve by automatically screening the performance data and automatically drawing the performance curve; reducing the time cost of manual analysis and improving drawing efficiency; then, determining the target performance data according to the type of the performance test The target trade-off algorithm performs a performance fluctuation analysis on the performance curve according to the target trade-off algorithm and the preset performance fluctuation table to obtain a performance fluctuation analysis situation; wherein, the performance fluctuation analysis situation includes: data with abnormal fluctuations and data without abnormal fluctuations; finally, when the performance fluctuation analysis situation is: data with abnormal fluctuations, the data with abnormal fluctuations is marked on the performance curve; when the performance fluctuation analysis situation is: data without abnormal fluctuations, it is indicated that the performance curve is normal; the target trade-off algorithm is used to automatically analyze the performance curve of the storage device, thereby improving the efficiency of analyzing the performance data of the storage device; and the performance fluctuation analysis situation is intuitively displayed in the performance curve, making it easy to view the analysis results of the performance data of the storage device. That is to say, the embodiment of the present application can automatically draw and analyze the performance curve of the storage device, thereby improving the efficiency of analyzing the performance data of the storage device. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 This is a schematic diagram of the functional modules of a performance data analysis tool for storage devices provided by an embodiment of the present application;

[0014] Figure 2 This is a flowchart of the steps of a method for analyzing performance data of a storage device provided by an embodiment of the present application;

[0015] Figure 3 When the target trade-off algorithm is the first trade-off algorithm, Figure 2 Schematic diagram of the specific steps of step S140;

[0016] Figure 4 When the target trade-off algorithm is the second trade-off algorithm, Figure 2 Schematic diagram of the specific steps of step S140;

[0017] Figure 5 When the target trade-off algorithm is the third trade-off algorithm, Figure 2 Schematic diagram of the specific steps of step S140;

[0018] Figure 6 This is a schematic diagram of a performance curve with abnormal fluctuations marked in one embodiment of the present application. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solutions and advantages of this application more clear, this application is further described in detail below with reference to the accompanying drawings and embodiments.

[0020] It should be noted that although a logical order is shown in the flowchart in the description of this application, in some cases, the steps shown or described may be performed in an order different from that in the flowchart. In the description of this application, "several" means one or more, and "more" means two or more. The description of "first" and "second" is only used to distinguish technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application pertains. The terms used herein are for the purpose of describing the embodiments of this application only and are not intended to limit this application.

[0022] The present application provides a method for analyzing the performance data of a storage device and a performance data analysis tool thereof, and relates to the technical field of storage device performance analysis. The method includes: obtaining the performance data of the storage device and a preset performance fluctuation table; when the performance data is inconsistent with the preset data format, storing the target performance data obtained after keyword screening of the performance data in a preset format data table, and automatically drawing a performance curve based on the obtained target data table; performing a performance fluctuation analysis on the performance curve based on the target trade-off algorithm determined by the type of performance test and the preset performance fluctuation table to obtain a performance fluctuation analysis situation; when the performance fluctuation analysis situation shows that the data has abnormal fluctuations, marking the data with abnormal fluctuations on the performance curve; when the performance fluctuation analysis situation shows that the data does not have abnormal fluctuations, indicating that the performance curve is normal. It can automatically draw and analyze the performance curve of the storage device, thereby improving the efficiency of analyzing the performance data of the storage device.

[0023] The embodiments of the present application are further described below with reference to the accompanying drawings.

[0024] like Figure 1 As shown, a performance data analysis tool 100 for storage devices provided by an embodiment of the present application includes: a data acquisition module 110, a format conversion module 120, an automatic drawing module 130, a trade-off analysis module 140, and an identification prompt module 150.

[0025] Specifically, the data acquisition module 110 is used to obtain performance data of the storage device and a preset performance fluctuation table; wherein different performance data are obtained by performing different types of performance tests on the storage device.

[0026] Specifically, the format conversion module 120 is used to: when the performance data is inconsistent with the preset data format, store the target performance data obtained after keyword screening of the performance data into the preset format data table to obtain a target data table.

[0027] Specifically, the automatic drawing module 130 is used to automatically draw a performance curve according to the target data table.

[0028] Specifically, the trade-off analysis module 140 is used to: determine the target trade-off algorithm according to the type of performance test, perform performance fluctuation analysis on the performance curve according to the target trade-off algorithm and the preset performance fluctuation table, and obtain a performance fluctuation analysis situation; wherein, the performance fluctuation analysis situation includes: data with abnormal fluctuations and data without abnormal fluctuations.

[0029] Specifically, the identification prompt module 150 is used to: when the performance fluctuation analysis result shows that there is abnormal fluctuation data, mark the abnormal fluctuation data on the performance curve; when the performance fluctuation analysis result shows that there is no abnormal fluctuation data, prompt that the performance curve is normal.

[0030] Specifically, the performance data analysis tool 100 for storage devices is mainly developed and implemented using Python language.

[0031] The performance data analysis tool provided in the embodiment of the present application realizes the performance data analysis method of the storage device in the embodiment of the present application through the mutual cooperation of the data acquisition module 110, the format conversion module 120, the automatic drawing module 130, the trade-off analysis module 140, and the identification prompt module 150. It can automatically draw and analyze the performance curve of the storage device, thereby improving the analysis efficiency of the performance data of the storage device.

[0032] Those skilled in the art will understand that the functional structure of the performance data analysis tool shown in the figure does not constitute a limitation on the embodiments of the present application, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0033] Those skilled in the art will understand that the functional structure and application scenarios of the performance data analysis tool described in the embodiments of the present application are intended to more clearly illustrate the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided in the embodiments of the present application. Those skilled in the art will know that with the evolution of the system architecture and the emergence of new application scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0034] Based on the functional structure of the above-mentioned performance data analysis tool, various embodiments of the performance data analysis method for a storage device of the present application are proposed below.

[0035] like Figure 2 As shown, the method for analyzing the performance data of a storage device may include but is not limited to steps S110 to S150.

[0036] Step S110: Acquire performance data of the storage device and a preset performance fluctuation table; wherein different performance data are obtained by performing different types of performance tests on the storage device.

[0037] Step S120: When the performance data is inconsistent with the preset data format, target performance data obtained after keyword screening of the performance data is stored in a preset format data table to obtain a target data table.

[0038] Step S130: automatically drawing a performance curve according to the target data table.

[0039] Step S140: Determine a target tradeoff algorithm based on the type of performance test, and perform performance fluctuation analysis on the performance curve based on the target tradeoff algorithm and a preset performance fluctuation table to obtain a performance fluctuation analysis result, wherein the performance fluctuation analysis result includes data with abnormal fluctuations and data without abnormal fluctuations.

[0040] Step S150: When the performance fluctuation analysis result is: there is abnormal fluctuation data, the abnormal fluctuation data is marked on the performance curve; when the performance fluctuation analysis result is: there is no abnormal fluctuation data, it is indicated that the performance curve is normal.

[0041] It can be understood that the performance data analysis object of the embodiment of the present application is: storage devices; specifically, storage devices include but are not limited to: UFS devices, eMMC devices.

[0042] It's understandable that before analyzing storage device performance data, it's necessary to conduct different types of performance tests on the storage device to obtain different performance data, laying the data foundation for performance analysis. Performance test types include: full-capacity sequential write data, full-capacity sequential read data, and random read and write data.

[0043] It is understandable that after obtaining the performance fluctuation analysis, abnormal fluctuations and prompts will be marked in the performance curve based on the performance fluctuation analysis. When the analysis shows abnormal fluctuations in the data, the abnormal fluctuations will be marked in red on the performance curve. Figure 6 shown; Figure 6 The horizontal axis is the serial number and the vertical axis is the performance value. If it is determined that there is no fluctuation, it indicates that the performance curve is normal. Through step S150, the performance fluctuation analysis is intuitively displayed in the performance curve, making it easy to view the analysis results of the performance data of the storage device.

[0044] Further explaining step S110, the performance data of the storage device is stored in a database, and the performance data can be called according to the model of the storage device to be analyzed.

[0045] Further explaining step S110, wherein the format of the preset performance fluctuation table is Excel format, the preset performance fluctuation table is used to store: performance threshold, the first preset difference threshold of the target analysis area, the preset mean threshold of the target analysis area, the difference threshold of the maximum performance value, the second preset difference threshold of the target analysis area and other data; so that when performing data analysis on the storage device, the content in the preset performance fluctuation table can be obtained through Python to execute the target trade-off algorithm to analyze the performance curve.

[0046] It is understood that the preset performance fluctuation table is derived from long-term experiments on storage devices. Different preset performance fluctuation tables are used for storage devices of different capacities. Preset performance fluctuation tables for different storage devices are stored separately in a database and can be called up based on the model of the storage device to be analyzed.

[0047] Further describing step S120, step S120 includes: when the performance data is inconsistent with the preset data format, determining a keyword; extracting target performance data from the performance data based on the keyword; and storing the target performance data in a data table in the preset format to obtain a target data table. This allows performance data in different formats to be unified into the same preset format, laying the foundation for subsequent automatic curve drawing.

[0048] Specifically, keywords are some specified characters of the printing performance data, and specific types of keywords can be pre-configured. Therefore, this application does not impose specific restrictions on the types of keywords.

[0049] In some embodiments, when the performance data is consistent with a preset data format, it is directly converted into a target data table, laying the foundation for subsequent automatic curve drawing.

[0050] Further describing step S130, automatically drawing a performance curve based on the target data table improves drawing efficiency and lays a foundation for subsequent performance analysis based on the performance curve, thereby improving the efficiency of analyzing performance data of the storage device. In some embodiments, the obtained performance curve includes: an SLC area, a mixed SLC and TLC area, and a TLC area.

[0051] Through steps S110 to S150, when analyzing the performance data of the storage device, first, the performance data of the storage device and the preset performance fluctuation table are obtained through the performance data analysis tool of the storage device; wherein, different performance data are obtained by performing different types of performance tests on the storage device; secondly, when the performance data is inconsistent with the preset data format, the target performance data obtained after keyword screening of the performance data is stored in the preset format data table to obtain the target data table; then, a performance curve is automatically drawn according to the target data table; the performance curve is obtained by automatically screening the performance data and automatically drawing the performance curve; the time cost of manual analysis is reduced and the drawing efficiency is improved; then, according to the performance test The target trade-off algorithm is determined based on the type of the target trade-off algorithm, and the performance fluctuation analysis is performed on the performance curve according to the target trade-off algorithm and the preset performance fluctuation table to obtain the performance fluctuation analysis situation; wherein, the performance fluctuation analysis situation includes: data with abnormal fluctuations and data without abnormal fluctuations; finally, when the performance fluctuation analysis situation is: data with abnormal fluctuations, the data with abnormal fluctuations is marked on the performance curve; when the performance fluctuation analysis situation is: data without abnormal fluctuations, it is indicated that the performance curve is normal; the performance curve of the storage device is automatically analyzed through the target trade-off algorithm to improve the analysis efficiency of the performance data of the storage device; and the performance fluctuation analysis situation is intuitively displayed in the performance curve to facilitate viewing the analysis results of the performance data of the storage device. That is to say, the embodiment of the present application can automatically draw and analyze the performance curve of the storage device, thereby improving the analysis efficiency of the performance data of the storage device.

[0052] Further explaining step S140, specifically, the target trade-off algorithm includes: a first trade-off algorithm, a second trade-off algorithm, and a third trade-off algorithm. Specifically, when the performance test type is: full-capacity sequential data write, the target trade-off algorithm is the first trade-off algorithm; when the performance test type is: full-capacity sequential data read, the target trade-off algorithm is the second trade-off algorithm; when the performance test type is: random read and write data, the target trade-off algorithm is the third trade-off algorithm.

[0053] According to some embodiments of the present application, Figure 3 , further illustrating step S140, when the type of performance test is: full-capacity sequential data writing; the target trade-off algorithm is the first trade-off algorithm; step S140 includes but is not limited to steps S210 to S230.

[0054] Step S210: According to the first trade-off algorithm and the preset performance fluctuation table, the performance curve is subjected to a first fluctuation analysis based on the first judgment condition to obtain a first fluctuation analysis result, the performance curve is subjected to a second fluctuation analysis based on the second judgment condition to obtain a second fluctuation analysis result, and the performance curve is subjected to a third fluctuation analysis based on the third judgment condition to obtain a third fluctuation analysis result.

[0055] Step S220: When one of the first fluctuation analysis result, the second fluctuation analysis result, and the third fluctuation analysis result indicates that abnormal fluctuation exists, the fluctuation analysis situation is determined as: data with abnormal fluctuation exists.

[0056] Step S230: When the first fluctuation analysis result, the second fluctuation analysis result, and the third fluctuation analysis result all indicate that there is no abnormal fluctuation, the performance fluctuation analysis condition is determined to be: there is no abnormal fluctuation data.

[0057] Specifically, the first trade-off algorithm includes: a preconfigured first judgment condition, a second judgment condition, and a third judgment condition.

[0058] Specifically, the first judgment condition is: the maximum value of the performance data is not greater than the performance threshold and the minimum value of the performance data is not greater than the performance threshold.

[0059] Specifically, the second judgment condition is: the absolute value of the difference between two adjacent data is not greater than a first preset difference threshold.

[0060] Specifically, the third judgment condition is: the mean value between two adjacent data is not less than a preset mean value threshold.

[0061] Through steps S210 to S230, the performance curve corresponding to the performance data obtained based on the performance test of full-capacity sequential data writing is automatically analyzed for performance fluctuations based on the first trade-off algorithm. This allows the performance curve to be analyzed more quickly to obtain performance fluctuation analysis results, thereby improving the efficiency of analyzing the performance data of the storage device.

[0062] According to some embodiments of the present application, step S210 is further described, wherein a first fluctuation analysis is performed on the performance curve based on a first judgment condition to obtain a first fluctuation analysis result, including but not limited to steps S211 to S213.

[0063] Step S211: Obtain a performance threshold from a preset performance fluctuation table.

[0064] Step S212: Compare the maximum value and the minimum value of the performance data of the target analysis area in the performance curve with the performance threshold respectively; wherein the target analysis area is: the sub-area obtained by dividing the performance curve, and there are at least two sub-areas.

[0065] Step S213: When the maximum value of the performance data is greater than the performance threshold or the minimum value of the performance data is greater than the performance threshold, it is judged that the first judgment condition is not met, and the first fluctuation analysis result is determined to be: there is abnormal fluctuation; when the maximum value of the performance data is not greater than the performance threshold and the minimum value of the performance data is not greater than the performance threshold, it is judged that the first judgment condition is met, and the first fluctuation analysis result is determined to be: there is no abnormal fluctuation.

[0066] It is understandable that different storage devices have different performance thresholds, and this application does not impose any specific restrictions on the value of the performance threshold.

[0067] Specifically, the maximum value of performance data refers to the maximum value of all data in the performance curve, and the minimum value of performance data refers to the minimum value of all data in the performance curve.

[0068] In one embodiment, when the performance curve is divided into three sub-regions: an SLC region, a mixed region of SLC and TLC, and a TLC region; all three sub-regions are target analysis regions.

[0069] It can be understood that steps S212 to S213 are respectively performed in the SLC area, the mixed area, and the TLC area to complete the first fluctuation analysis processing based on the first judgment condition and obtain the first fluctuation analysis results of each sub-area.

[0070] Through steps S211 to S213 , the first fluctuation analysis process based on the first judgment condition is automatically completed to obtain the first fluctuation analysis result, which is beneficial to improving the analysis efficiency of the performance curve.

[0071] Two examples are given to further illustrate the specific process of the first fluctuation analysis processing based on the first judgment condition provided in the embodiment of the present application.

[0072] Example 1: The performance threshold obtained from the preset performance fluctuation table is 100, the maximum performance data in the performance curve is 80, and the minimum performance data is 50. If the maximum performance data is not greater than the performance threshold 100, and the minimum performance data is not greater than the performance threshold 100, then the first judgment condition is met, and the first fluctuation analysis result is determined to be: there is no abnormal fluctuation.

[0073] Example 2: The performance threshold obtained from the preset performance fluctuation table is 100, the maximum value of the performance data in the performance curve is 110, and the minimum value of the performance data is 50. If the maximum value of the performance data is greater than the performance threshold 100, and the minimum value of the performance data is not greater than the performance threshold 100, then the first judgment condition is not met, and the first fluctuation analysis result is determined to be: abnormal fluctuation exists.

[0074] According to some embodiments of the present application, step S210 is further described, wherein a second fluctuation analysis is performed on the performance curve based on a second judgment condition to obtain a second fluctuation analysis result, including but not limited to steps S214 to S216.

[0075] Step S214: obtaining a first preset difference threshold of a target analysis area from a preset performance fluctuation table; wherein the target analysis area is: a sub-area obtained by dividing the performance curve, and there are at least two sub-areas.

[0076] Step S215: performing a difference analysis process on the target analysis area in the performance curve, and comparing the absolute value of the difference between two adjacent data in the target analysis area with a first preset difference threshold;

[0077] Step S216: When the absolute value of the difference between two adjacent data is not greater than the first preset difference threshold, it is judged that the second judgment condition is met, and the result of the second fluctuation analysis is determined to be: there is no abnormal fluctuation; when the absolute value of the difference between two adjacent data is greater than the first preset difference threshold, it is judged that the second judgment condition is not met, and the result of the second fluctuation analysis is determined to be: there is abnormal fluctuation.

[0078] Specifically, different manufacturers may have different curve analysis strategies. During the second fluctuation analysis based on the second judgment condition, the performance curve is divided into two, three, or four sub-regions. The second fluctuation analysis based on the second judgment condition is then performed on each of the resulting sub-regions. The resulting sub-regions are all target analysis regions.

[0079] Specifically, the first preset difference thresholds for different target analysis areas are different. For example, if the performance curve is divided into three sub-areas: an SLC area, a mixed area of SLC and TLC, and a TLC area; all three sub-areas are target analysis areas; then the first preset difference thresholds corresponding to the SLC area, the mixed area, and the TLC area are different. Therefore, this application does not impose specific restrictions on the value of the first preset difference threshold for the target analysis area.

[0080] It can be understood that steps S214 to S216 are respectively executed in the SLC area, the mixed area, and the TLC area to complete the second fluctuation analysis processing based on the second judgment condition and obtain the second fluctuation analysis results of each sub-area.

[0081] Through steps S214 to S216 , the second fluctuation analysis process based on the second judgment condition is automatically completed, which is beneficial to improving the analysis efficiency of the performance curve.

[0082] An example is given to further illustrate the specific process of the second fluctuation analysis processing based on the second judgment condition provided in the embodiment of the present application.

[0083] Example 3:

[0084] First, when the performance curve includes: SLC area, mixed area of SLC and TLC, and TLC area; the first preset difference threshold corresponding to the SLC area is obtained from the preset performance fluctuation table as 60, the first preset difference threshold corresponding to the mixed area is 80, and the first preset difference threshold corresponding to the TLC area is 100.

[0085] Next, in the SLC area, the absolute value of the difference between two adjacent data is compared with the first preset difference threshold 60. When the absolute value of the difference between two adjacent data is 80 (greater than 60), it is judged that the second judgment condition is not met, and the second fluctuation analysis result of the SLC area is determined to be: there is abnormal fluctuation.

[0086] Then, in the mixed area of SLC and TLC, the absolute value of the difference between two adjacent data is compared with the first preset difference threshold 80. When the absolute value of the difference between all two adjacent data is not greater than the first preset difference threshold, it is judged that the second judgment condition is met, and the second fluctuation analysis result of the mixed area is determined to be: there is no abnormal fluctuation.

[0087] Finally, in the TLC region, the absolute value of the difference between two adjacent data points is compared with a first preset difference threshold of 100. If the absolute value of the difference between all adjacent data points is not greater than the first preset difference threshold, the second judgment condition is determined to be met, and the second fluctuation analysis result for the TLC region is determined to be: no abnormal fluctuation exists. This completes the second fluctuation analysis process based on the second judgment condition.

[0088] According to some embodiments of the present application, step S210 is further described, wherein a third fluctuation analysis is performed on the performance curve based on a third judgment condition to obtain a third fluctuation analysis result, including but not limited to steps S217 to S219.

[0089] Step S217: obtaining a preset mean threshold of a target analysis area from a preset performance fluctuation table; the target analysis area is: a sub-area obtained by dividing the performance curve, and there are at least two sub-areas.

[0090] Step S218: performing mean analysis processing on the target analysis area in the performance curve, and comparing the mean between two adjacent data in the target analysis area with a preset mean threshold of the target analysis area.

[0091] Step S219: When the mean between two adjacent data is not less than the preset mean threshold, it is judged that the third judgment condition is met, and the third fluctuation analysis result of the target analysis area is determined to be: there is no abnormal fluctuation; when the mean between two adjacent data is less than the preset mean threshold, it is judged that the third judgment condition is not met, and the third fluctuation analysis result of the target analysis area is determined to be: there is abnormal fluctuation.

[0092] Specifically, different manufacturers may have different curve analysis strategies. During the third fluctuation analysis based on the third judgment condition, the performance curve is divided into two, three, or four sub-regions. The third fluctuation analysis based on the third judgment condition is then performed on each of the resulting sub-regions. The resulting sub-regions are all target analysis regions.

[0093] Specifically, the preset mean thresholds for different target analysis areas differ. For example, if the performance curve is divided into three sub-areas: an SLC area, a mixed SLC and TLC area, and a TLC area, and all three sub-areas are target analysis areas, then the preset mean thresholds for the SLC area, the mixed SLC and TLC area, and the TLC area are different. Therefore, this application does not impose specific restrictions on the preset mean thresholds for the target analysis areas.

[0094] It can be understood that steps S217 to S219 are respectively executed in the SLC area, the mixed area, and the TLC area to complete the third fluctuation analysis processing based on the third judgment condition and obtain the third fluctuation analysis results of each sub-area.

[0095] Through steps S217 to S219, the third fluctuation analysis process based on the third judgment condition is automatically completed, which is beneficial to improving the analysis efficiency of the performance curve.

[0096] An example is given to further illustrate the specific process of the third fluctuation analysis processing based on the third judgment condition provided in the embodiment of the present application.

[0097] Example 4:

[0098] First, when the performance curve includes an SLC area, a mixed SLC and TLC area, and a TLC area, the preset mean threshold for the SLC area is obtained from the preset performance fluctuation table as 50, 70, and 100 for the mixed SLC and TLC area. Next, a mean analysis is performed on the SLC area. If the mean between two adjacent data points is 30 (less than 50), it is determined that the third judgment condition is not met. The third fluctuation analysis result for the SLC area is: abnormal fluctuation exists, indicating abnormal fluctuation exists in the SLC area. Then, the mean analysis is performed similarly on the mixed and TLC areas to obtain the third fluctuation analysis result. This completes the third fluctuation analysis based on the third judgment condition.

[0099] According to some embodiments of the present application, Figure 4 , further illustrating step S140, when the type of performance test is: full-capacity sequential data reading; the target trade-off algorithm is the second trade-off algorithm; step S140 includes but is not limited to steps S310 to S330.

[0100] Step S310: According to the second trade-off algorithm and the preset performance fluctuation table, the performance curve is subjected to a fourth fluctuation analysis based on the fourth judgment condition to obtain a fourth fluctuation analysis result, and the performance curve is subjected to a fifth fluctuation analysis based on the fifth judgment condition to obtain a fifth fluctuation analysis result.

[0101] Step S320: When one of the fourth fluctuation analysis result and the fifth fluctuation analysis result indicates that abnormal fluctuation exists, the fluctuation analysis situation is determined to be: data with abnormal fluctuation exists.

[0102] Step S330: When both the fourth fluctuation analysis result and the fifth fluctuation analysis result indicate that there is no abnormal fluctuation, the performance fluctuation analysis condition is determined to be: there is no abnormal fluctuation data.

[0103] Specifically, the second trade-off algorithm includes: a preconfigured fourth judgment condition and a fifth judgment condition.

[0104] Specifically, the fourth judgment condition is: the difference between the maximum value of the performance data and the minimum value of the performance data is not greater than the difference threshold of the performance values.

[0105] Specifically, the fifth judgment condition is: the absolute value of the difference between two adjacent data is not greater than the second preset difference threshold.

[0106] Through steps S310 to S330, the performance curve corresponding to the performance data obtained based on the performance test of full-capacity sequential reading data is automatically analyzed for performance fluctuations based on the second trade-off algorithm. This allows the performance curve to be analyzed more quickly to obtain performance fluctuation analysis results, thereby improving the efficiency of analyzing the performance data of the storage device.

[0107] According to some embodiments of the present application, step S310 is further described, wherein a fourth fluctuation analysis is performed on the performance curve based on a fourth judgment condition to obtain a fourth fluctuation analysis result, including but not limited to steps S311 to S313.

[0108] Step S311: obtaining a performance maximum difference threshold from a preset performance fluctuation table.

[0109] Step S312: Obtain the maximum value and the minimum value of the performance data from the target analysis area of the performance curve, and calculate the difference between the maximum value and the minimum value of the performance data; the target analysis area is: the sub-area obtained by dividing the performance curve, and there are at least two sub-areas.

[0110] Step S313: When the difference between the maximum value of the performance data and the minimum value of the performance data is not greater than the difference threshold of the performance maximum value, it is judged that the fourth judgment condition is met, and the result of the fourth fluctuation analysis is determined to be: there is no abnormal fluctuation; when the difference between the maximum value of the performance data and the minimum value of the performance data is greater than the difference threshold of the performance maximum value, it is judged that the fourth judgment condition is not met, and the result of the fourth fluctuation analysis is determined to be: there is abnormal fluctuation.

[0111] It is understandable that the difference threshold of the maximum performance value is pre-configured, and the embodiment of the present application does not impose any specific restriction on the value of the difference threshold of the maximum performance value.

[0112] Specifically, each sub-region obtained by dividing the performance curve is a target analysis region. For example, if the performance curve is divided into three sub-regions: the SLC region, the mixed SLC and TLC region, and the TLC region, all three sub-regions are target analysis regions. It is understood that steps S312 to S313 are performed in the SLC region, the mixed region, and the TLC region, respectively, completing the fourth fluctuation analysis based on the fourth judgment condition, and obtaining the fourth fluctuation analysis results for each sub-region.

[0113] Through steps S311 to S313 , the fourth fluctuation analysis process based on the fourth judgment condition is automatically completed, which is beneficial to improving the analysis efficiency of the performance curve.

[0114] According to some embodiments of the present application, step S310 is further described, wherein a fifth fluctuation analysis is performed on the performance curve based on a fifth judgment condition to obtain a fifth fluctuation analysis result, including but not limited to steps S314 to S316.

[0115] Step S314: obtaining a second preset difference threshold of the target analysis area from the preset performance fluctuation table; wherein the target analysis area is: a sub-area obtained by dividing the performance curve, and there are at least two sub-areas.

[0116] Step S315: performing a difference analysis process on the target analysis region in the performance curve. In the target analysis region, the absolute value of the difference between two adjacent data is compared with a second preset difference threshold.

[0117] Step S316: When the absolute value of the difference between two adjacent data is not greater than the second preset difference threshold, it is judged that the fifth judgment condition is met, and the result of the fifth fluctuation analysis is determined to be: there is no abnormal fluctuation; when the absolute value of the difference between two adjacent data is greater than the second preset difference threshold, it is judged that the fifth judgment condition is not met, and the result of the fifth fluctuation analysis is determined to be: there is abnormal fluctuation.

[0118] It is understandable that the second preset difference threshold is pre-configured, and the value of the difference threshold for the performance maximum in the embodiment of the present application is not specifically limited. The second preset difference threshold is different for different target analysis areas.

[0119] Specifically, each sub-region obtained by dividing the performance curve is a target analysis region. For example, if the performance curve is divided into three sub-regions: the SLC region, the mixed SLC and TLC region, and the TLC region, all three sub-regions are target analysis regions. It can be understood that steps S314 to S316 are performed in the SLC region, the mixed region, and the TLC region, respectively, completing the fifth fluctuation analysis based on the fifth judgment condition, and obtaining the fifth fluctuation analysis results for each sub-region.

[0120] Through steps S311 to S313 , the fourth fluctuation analysis process based on the fourth judgment condition is automatically completed, which is beneficial to improving the analysis efficiency of the performance curve.

[0121] Through steps S314 to S316 , the fifth fluctuation analysis process based on the fifth judgment condition is automatically completed, which is beneficial to improving the analysis efficiency of the performance curve.

[0122] According to some embodiments of the present application, Figure 5 , further illustrating step S140, when the type of performance test is: random read and write data; the target trade-off algorithm is the third trade-off algorithm; step S140: determine the target trade-off algorithm according to the type of performance test, and perform performance fluctuation analysis on the performance curve according to the target trade-off algorithm and the preset performance fluctuation table to obtain the performance fluctuation analysis situation, including but not limited to steps S410 to S460.

[0123] Step S410: Obtain a third preset difference threshold and an error value from a preset performance fluctuation table.

[0124] Step S420: determining a target difference range according to the third preset difference threshold and the error value.

[0125] Step S430: Acquire a plurality of continuous data in any local area of the performance curve.

[0126] Step S440: For a plurality of continuous data, calculate the difference between every two adjacent data to obtain a plurality of difference values.

[0127] Step S450: When any difference value is not within the target difference value range, it is determined that the performance fluctuation analysis situation is: abnormal fluctuation data exists.

[0128] Step S460: When all the differences are within the target difference range, it is determined that the performance fluctuation analysis result is: there is no abnormal fluctuation data.

[0129] It can be understood that the third preset difference threshold and error value are both pre-configured, and this application does not impose any specific restrictions on the third preset difference threshold and error value, and the target difference range determined by the third preset difference threshold and error value.

[0130] Through steps S410 to S460, the performance curve corresponding to the performance data obtained based on the performance test of random read and write data is automatically analyzed for performance fluctuations based on the third trade-off algorithm. This allows the performance curve to be analyzed more quickly to obtain performance fluctuation analysis results, thereby improving the efficiency of analyzing the performance data of the storage device.

[0131] An example is given to further illustrate the specific process of performance fluctuation analysis and processing based on the third trade-off algorithm provided in the embodiment of the present application.

[0132] Example 5:

[0133] When the third preset difference threshold is 50, the error value is 10, and the target difference range is 40 to 60, and four consecutive data points are obtained: 50, 100, 60, and 90, and the difference between each two adjacent data points is within the target difference range, then there is no fluctuation anomaly at these four points. Continue to obtain four consecutive data points: 50, 200, 100, and 150. If the difference between two adjacent data points is not within the target difference range, then it is determined that there is a fluctuation anomaly and these four points are jittering.

[0134] After automatically drawing the performance curve according to the target data table, the method further includes: receiving a curve zoom instruction; and zooming in on the performance curve in response to the curve zoom instruction, so that technicians can more intuitively view the performance fluctuation analysis of the storage device.

[0135] The above is a specific description of the preferred implementation of the present application, but the present application is not limited to the above implementation mode. Technical personnel familiar with the field can also make various equivalent modifications or substitutions without violating the spirit of the present application. These equivalent modifications or substitutions are all included in the scope defined by the present application.

Claims

1. A method for analyzing performance data of a storage device, characterized in that: Applied to a performance data analysis tool, the method includes: Acquire performance data and a preset performance fluctuation table of a storage device; wherein different performance data are obtained by performing different types of performance tests on the storage device; When the performance data is inconsistent with the preset data format, the target performance data obtained after keyword screening of the performance data is stored in the preset format data table to obtain a target data table; Automatically draw a performance curve based on the target data table; Determining a target trade-off algorithm according to the type of the performance test, performing a performance fluctuation analysis on the performance curve according to the target trade-off algorithm and the preset performance fluctuation table to obtain a performance fluctuation analysis result; wherein the performance fluctuation analysis result includes: data with abnormal fluctuations and data without abnormal fluctuations; When the performance fluctuation analysis result shows that abnormal fluctuation data exists, the abnormal fluctuation data is marked on the performance curve; when the performance fluctuation analysis result shows that abnormal fluctuation data does not exist, it is indicated that the performance curve is normal.

2. The method for analyzing performance data of a storage device according to claim 1, wherein: When the type of the performance test is: full-capacity sequential data writing; the target trade-off algorithm is the first trade-off algorithm; The step of determining a target trade-off algorithm according to the type of the performance test, and performing a performance fluctuation analysis on the performance curve according to the target trade-off algorithm and the preset performance fluctuation table to obtain a performance fluctuation analysis result includes: According to the first trade-off algorithm and the preset performance fluctuation table, performing a first fluctuation analysis process on the performance curve based on a first judgment condition to obtain a first fluctuation analysis result, performing a second fluctuation analysis process on the performance curve based on a second judgment condition to obtain a second fluctuation analysis result, and performing a third fluctuation analysis process on the performance curve based on a third judgment condition to obtain a third fluctuation analysis result; When one of the first fluctuation analysis result, the second fluctuation analysis result, and the third fluctuation analysis result indicates that abnormal fluctuation exists, determining the fluctuation analysis condition as: data with abnormal fluctuation exists; When the first fluctuation analysis result, the second fluctuation analysis result, and the third fluctuation analysis result all indicate that there is no abnormal fluctuation, the performance fluctuation analysis condition is determined to be: there is no abnormal fluctuation data.

3. The method for analyzing performance data of a storage device according to claim 2, wherein: The performing a first fluctuation analysis process on the performance curve based on the first judgment condition to obtain a first fluctuation analysis result includes: Obtaining a performance threshold from the preset performance fluctuation table; Comparing the maximum value of performance data and the minimum value of performance data in the target analysis area of the performance curve with the performance threshold respectively; wherein the target analysis area is: a sub-area obtained by dividing the performance curve, and the sub-area has at least two sub-areas; When the maximum value of the performance data is greater than the performance threshold or the minimum value of the performance data is greater than the performance threshold, it is determined that the first judgment condition is not met, and the first fluctuation analysis result is determined to be: abnormal fluctuation exists; When the maximum value of the performance data is not greater than the performance threshold and the minimum value of the performance data is not greater than the performance threshold, it is determined that the first judgment condition is met, and the first fluctuation analysis result is determined to be: there is no abnormal fluctuation.

4. The method for analyzing performance data of a storage device according to claim 2, wherein: The performing a second fluctuation analysis process on the performance curve based on the second judgment condition to obtain a second fluctuation analysis result includes: Obtaining a first preset difference threshold value of a target analysis region from the preset performance fluctuation table; wherein the target analysis region is: a subregion obtained by dividing the performance curve, and the subregions are at least two; performing a difference analysis process on the target analysis area in the performance curve, and comparing the absolute value of the difference between two adjacent data in the target analysis area with the first preset difference threshold; When the absolute value of the difference between two adjacent data is not greater than the first preset difference threshold, it is determined that the second judgment condition is met, and the second fluctuation analysis result is determined to be: there is no abnormal fluctuation; When the absolute value of the difference between two adjacent data is greater than the first preset difference threshold, it is determined that the second judgment condition is not met, and the second fluctuation analysis result is determined to be: abnormal fluctuation exists.

5. The method for analyzing performance data of a storage device according to claim 2, wherein: The performing a third fluctuation analysis process on the performance curve based on the third judgment condition to obtain a third fluctuation analysis result includes: Obtaining a preset mean threshold of a target analysis area from the preset performance fluctuation table; the target analysis area is: a sub-area obtained by dividing the performance curve, and the sub-area has at least two sub-areas; Performing mean analysis processing on the target analysis area in the performance curve, and comparing the mean between two adjacent data in the target analysis area with the preset mean threshold of the target analysis area; When the mean value between two adjacent data is not less than the preset mean threshold value, it is determined that the third judgment condition is met, and the third fluctuation analysis result of the target analysis area is determined to be: there is no abnormal fluctuation; When the mean value between two adjacent data is less than the preset mean value threshold, it is determined that the third judgment condition is not met, and the third fluctuation analysis result of the target analysis area is determined to be: abnormal fluctuation exists.

6. The method for analyzing performance data of a storage device according to claim 1, wherein: When the type of the performance test is: full-capacity sequential data reading; the target trade-off algorithm is the second trade-off algorithm; The step of determining a target trade-off algorithm according to the type of the performance test, and performing a performance fluctuation analysis on the performance curve according to the target trade-off algorithm and the preset performance fluctuation table to obtain a performance fluctuation analysis result includes: performing, according to the second trade-off algorithm and the preset performance fluctuation table, a fourth fluctuation analysis process based on a fourth judgment condition on the performance curve to obtain a fourth fluctuation analysis result, and performing a fifth fluctuation analysis process based on a fifth judgment condition on the performance curve to obtain a fifth fluctuation analysis result; When one of the fourth fluctuation analysis result and the fifth fluctuation analysis result indicates that abnormal fluctuation exists, the fluctuation analysis condition is determined to be: data with abnormal fluctuation exists; When both the fourth fluctuation analysis result and the fifth fluctuation analysis result indicate that there is no abnormal fluctuation, the performance fluctuation analysis condition is determined to be: there is no abnormal fluctuation data.

7. The method for analyzing performance data of a storage device according to claim 6, wherein: Performing a fourth fluctuation analysis process on the performance curve based on a fourth judgment condition to obtain a fourth fluctuation analysis result includes: Obtaining a performance maximum value difference threshold from the preset performance fluctuation table; Obtaining a maximum value and a minimum value of performance data from a target analysis region of the performance curve, and calculating a difference between the maximum value and the minimum value of the performance data; the target analysis region being a subregion obtained by dividing the performance curve, the subregions being at least two; When the difference between the maximum value of the performance data and the minimum value of the performance data is not greater than the difference threshold of the performance values, it is determined that the fourth judgment condition is met, and the fourth fluctuation analysis result is determined to be: there is no abnormal fluctuation; When the difference between the maximum value of the performance data and the minimum value of the performance data is greater than the difference threshold of the performance maximum value, it is determined that the fourth judgment condition is not met, and the fourth fluctuation analysis result is determined to be: abnormal fluctuation exists.

8. The method for analyzing performance data of a storage device according to claim 6, wherein: The performing a fifth fluctuation analysis process on the performance curve based on the fifth judgment condition to obtain a fifth fluctuation analysis result includes: Obtaining a second preset difference threshold value of a target analysis region from the preset performance fluctuation table; wherein the target analysis region is: a subregion obtained by dividing the performance curve, and the subregions are at least two; performing a difference analysis process on the target analysis area in the performance curve, and comparing the absolute value of the difference between two adjacent data in the target analysis area with the second preset difference threshold; When the absolute value of the difference between two adjacent data is not greater than the second preset difference threshold, it is determined that the fifth judgment condition is met, and the fifth fluctuation analysis result is determined to be: there is no abnormal fluctuation; When the absolute value of the difference between two adjacent data is greater than the second preset difference threshold, it is determined that the fifth judgment condition is not met, and the fifth fluctuation analysis result is determined to be: abnormal fluctuation exists.

9. The method for analyzing performance data of a storage device according to claim 1, wherein: When the type of the performance test is: random read and write data; the target trade-off algorithm is the third trade-off algorithm; The step of determining a target trade-off algorithm according to the type of the performance test, and performing a performance fluctuation analysis on the performance curve according to the target trade-off algorithm and the preset performance fluctuation table to obtain a performance fluctuation analysis result includes: Obtaining a third preset difference threshold and an error value from the preset performance fluctuation table; Determining a target difference range according to the third preset difference threshold and the error value; Acquiring a plurality of continuous data in any local area of the performance curve; For multiple consecutive data, calculate the difference between every two adjacent data to obtain multiple difference values; When any difference is not within the target difference range, the performance fluctuation analysis condition is determined to be: abnormal fluctuation data exists; When all the differences are within the target difference range, it is determined that the performance fluctuation analysis situation is: there is no abnormal fluctuation data.

10. A performance data analysis tool for a storage device, characterized in that: Used to execute the performance data analysis method for a storage device as described in any one of claims 1 to 9.

Images