TW201939413A - Computer resource usage trends prediction and analyzing system and method thereof - Google Patents

Abstract

A computer resource usage trends prediction and analyzing system comprising: a data capture device, a data analysis device, a data display device and a trend prediction device. The data capture device captures at least one Key Performance Indicator (KPI) value of the usage of at least one resource of at least one computer. The data analysis device connects to the data capture device and receives, analysis the at least one KPI value. The data display device connects to the data analysis device and displays the at least one KPI value analyzed by the data analysis device as an image. The trend prediction device analyses the at least one KPI value displaying as the image to generate a trend prediction image, wherein the trend prediction device connects to the data display device and the data display device displays the trend prediction image. Moreover, present invention also provides a computer resource usage trends prediction and analyzing method.

Description

Computer resource usage trend prediction analysis system and method

The invention relates to a computer analysis system and method, and particularly to a computer analysis system that analyzes the usage of various resources in a computer, and graphically uses the usage of each resource, and further uses the usage of each resource for trend prediction analysis Computer resource usage trend prediction analysis system and method.

Due to the advancement of science and technology, the scope of application of information engineering is becoming wider and wider. With the increase of the scale of information engineering, the software used in various technology systems is becoming more and more complex. In order to maintain the operation of the system, the more complex the Software and engineering environments require more computers to perform calculations. In this way, the monitoring of the usage of multiple resources such as the Central Processing Unit (CPU), memory, and disks in multiple computers has become a very important thing in information maintenance projects. Furthermore, if it is possible to predict when multiple resources will be overloaded or the disk capacity is lower than the safe range in the monitoring operation of multiple resources of multiple computers, multiple computers will run more smoothly.

On the other hand, among various display chart types, the K-line chart is a very intuitive and easy-to-see chart display method. The K-line chart is generated based on the four values of "opening price, closing price, highest price, and lowest price" of the stock price. In the K-line chart, you can fully observe the true changes of the market. You can not only see the trend of stock prices (or the market), but also understand the daily fluctuations of stock market conditions.

Based on the above reasons, in the monitoring of the usage of multiple resources by multiple computers, if the K-line diagram can be used to display the usage of each resource, the monitoring party can observe each resource in a more intuitive and clear way In the long-term usage situation, do a big data trend prediction analysis to find out early if there is any abnormality in the use of each resource.

In order to achieve the foregoing object, the present invention provides a computer resource usage trend prediction analysis system, including: a data acquisition device that acquires at least one key index value of at least one resource usage in at least one computer device; The data analysis device is connected to the data acquisition device, and receives and analyzes at least one key index value; a data display device is connected to the data analysis device, and the at least one key index value analyzed by the data analysis device is an image Display; and a trend prediction device that analyzes at least one key index value displayed by an image to generate a trend prediction map; wherein the trend prediction device is connected to the data display device, and the data display device displays the trend prediction map.

Preferably, the at least one key indicator value includes a daily start usage value, a daily end usage value, a daily maximum usage value and a daily minimum usage value of at least one resource.

Preferably, the trend prediction device performs analysis by a linear regression method to generate a trend prediction map.

Preferably, the at least one key index value is at least one KPI value, and the image is a K-line diagram.

Preferably, the usage amount of the at least one resource includes at least one CPU usage rate, at least one memory usage rate, and at least one magnetic disk capacity.

Furthermore, the present invention also provides a computer resource usage trend prediction analysis method, which includes the following steps: using a data acquisition device to retrieve at least one key index value of at least one resource usage in at least one computer device; A data analysis device receives and analyzes at least one key index value; a data display device displays at least one key index value analyzed by the data analysis device as an image; and a trend prediction device analyzes the image Display at least one key index value and generate a trend prediction chart; wherein the data display device displays the trend prediction chart.

Preferably, in the step of analyzing the at least one key index value displayed by the image by the trend prediction device and generating a trend prediction map, the trend prediction device further generates a prediction threshold. When the value reaches the predicted threshold, at least one predicted value that reaches the predicted threshold will be displayed on the trend prediction chart in an alert form.

Preferably, the trend prediction device analyzes at least one key index value in the image by a machine learning method to generate a prediction threshold.

Preferably, the at least one key indicator value includes a daily start usage value, a daily end usage value, a daily maximum usage value and a daily minimum usage value of at least one resource.

Preferably, the trend prediction device performs analysis by a linear regression method to generate a trend prediction map.

The following describes the embodiments of the present invention in more detail with reference to the drawings and component symbols, so that those skilled in the art can implement them after studying this specification.

FIG. 1 is a structural diagram for explaining a computer resource usage trend prediction analysis system according to an embodiment of the present invention. Referring to FIG. 1, in an embodiment of the present invention, a computer resource usage trend prediction analysis system 1 includes a data acquisition device 10, a data analysis device 12, a data presentation device 14, and a trend prediction device 16. The data retrieving device 10 retrieves at least one key index value of the usage of at least one resource in at least one computer device 2, wherein the at least one resource is, for example, the CPU 20, the memory 22, and the magnetic disk in the computer device 2. 24 and other hardware devices, and the resource usage amount is, for example, the usage rate of the CPU 20, the usage rate of the memory 22, and the capacity of the magnetic disk 24. The data analysis device 12 is connected to the data acquisition device 10 and receives and analyzes at least one key index value. The data display device 14 is connected to the data analysis device 12 and displays at least one key indicator value analyzed by the data analysis device 12 as an image, wherein the key indicator value is a KPI (Key Performance Indicator, KPI), and The image is a K-line diagram. The trend prediction device 16 analyzes at least one key index value displayed by the image to generate a trend prediction map. The trend prediction device 16 is connected to the data display device 14, and the data display device 14 can also display the trend prediction map generated by the trend prediction device 16.

It should be noted that, in an embodiment of the present invention, the data acquisition device 10 can acquire the key index values of the CPU 20, the memory 22, or the magnetic disk 24 in the computer device 2 at four different periods. For example, the data capture device 10 can capture the daily start usage value (open), daily end usage value (close), and daily maximum of each resource, such as CPU 20, memory 22, or disk 24. Usage value (high) and minimum daily usage value (low). The daily usage value is the first indicator value generated by each resource at 00:00 each day; the daily end usage value is the last indicator value generated by each resource at 23:59 daily; the maximum daily usage amount The value is the maximum indicator value generated by each resource in the 24 hours of a day; the daily minimum usage value is the minimum indicator value generated by each resource in the 24 hours of the day. According to the present invention, the key index values generated by each resource in four time periods can be further received and analyzed by the data analysis device 12 and the key index values captured by the data acquisition device 10, and the data display device 14 can be used. A K-line diagram showing the usage amount of each resource analyzed by the data analysis device 12. In addition, the K-line chart displayed by the data display device 14 may display a day K-line chart in units of days, a week K-line chart in units of weeks, or a month K-line chart in units of months. , In a more intuitive way to monitor the usage of each resource in a specific time period.

Taking the magnetic disk capacity as an example, the K-line diagram displayed by the data display device 14 of the present invention will now be described. FIG. 2 is a schematic diagram for explaining a usage trend of the capacity of the magnetic disk 24 according to an embodiment of the present invention. Please refer to FIG. 1 and FIG. 2, when the data analysis device 12 takes the root capacity (Root) of the daily capacity of the disk 24 from August to October, the data display device 14 can display, for example, August in FIG. 2. The trend screen of the remaining capacity of the magnetic disk 24 to October. The unit of the remaining capacity of the magnetic disk 24 is Gigabytes, GB. It can be seen from Figure 2 that the remaining capacity of the magnetic disk 24 did not increase or decrease significantly during the three months. Only some capacity was released between August 20 and August 27, but After that, the remaining capacity also stabilized, indicating that the reduction of the capacity of the disk 24 had been stably controlled. It should be understood that, in the present invention, a user can further understand the usage status of the magnetic disk 24 during the time period by clicking the value of the K-line chart displayed in the data display device 14 during the time period.

FIG. 3 is a schematic diagram for explaining a trend of a usage amount of a magnetic disk capacity according to another embodiment of the present invention. Please refer to FIG. 1 and FIG. 3. In another embodiment of the present invention, when the data analysis device 12 logs the remaining capacity of the magnetic disk 24 from August to October, the data display device 14 may display, for example, The trend screen of the remaining capacity of the magnetic disk 24 from August to October in FIG. 3, wherein the unit of the remaining capacity of the magnetic disk 24 is Gigabytes, GB. It can be seen from FIG. 3 that the remaining capacity of the magnetic disk 24 is reduced by about 40GB within 3 months, and is continuously reduced in a stable manner, which indicates that the magnetic disk 24 is in a normal use state, and the current capacity of the magnetic disk 24 is about remaining 100GB. It should be understood that, in the present invention, a user can further understand the usage status of the magnetic disk 24 during the time period by clicking the value of the K-line chart displayed in the data display device 14 during the time period.

FIG. 4 is a schematic diagram for explaining a usage trend of a magnetic disk capacity according to another embodiment of the present invention. Please refer to FIG. 1 and FIG. 4. In yet another embodiment of the present invention, when the data analysis device 12 logs the remaining capacity of the magnetic disk 24 from May to October, the data display device 14 may display, for example, The trend screen of the remaining capacity of the magnetic disk 24 from May to October in FIG. 4, wherein the unit of the remaining capacity of the magnetic disk 24 is Gigabytes, GB. It can be seen from FIG. 4 that the remaining capacity of the magnetic disk 24 is about 50 GB released from May to June, which means that the magnetic disk 24 may be maintained and unnecessary data may be deleted from May to June. After that, from June to October, it was reduced by about 30GB each month, and it continued to decrease in a stable manner, which indicates that the magnetic disk 24 is in a normal use state. It should be understood that, in the present invention, a user can further understand the usage status of the magnetic disk 24 during the time period by clicking the value of the K-line chart displayed in the data display device 14 during the time period.

FIG. 5 is a schematic diagram for explaining a usage trend of a magnetic disk capacity according to another embodiment of the present invention. Please refer to FIG. 1 and FIG. 5. In yet another embodiment of the present invention, when the data analysis device 12 analyzes the remaining capacity of the magnetic disk 24 from May to October by a var method, the data can be obtained by using data. The presentation device 14 displays, for example, a trend screen of the remaining capacity of the magnetic disk 24 displayed in a variable form between May and October in FIG. 5. It can be seen from FIG. 5 that the remaining capacity of the magnetic disk 24 is continuously reduced from May to August, which represents that the magnetic disk 24 is in a normal use state. Disk 24 was then maintained in mid-August, while disk 24 was stable from mid-August to October. It should be understood that, in the present invention, a user can further understand the usage status of the magnetic disk 24 during the time period by clicking the value of the K-line chart displayed in the data display device 14 during the time period.

FIG. 6 is a schematic diagram for explaining a trend of a usage amount of a magnetic disk capacity according to another embodiment of the present invention. Please refer to FIG. 1 and FIG. 6. In yet another embodiment of the present invention, the data analysis device 12 analyzes the remaining capacity of the magnetic disk 24 from May to October, and displays, for example, the data display device 14 in FIG. 5. The trend screen of the remaining capacity of the magnetic disk 24 from May to October. The unit of the remaining capacity of the magnetic disk 24 is Gigabytes, GB. It can be seen from FIG. 5 that the remaining capacity of the magnetic disk 24 is reduced by about 100 GB each month from May to October, and continues to decrease in a stable manner, which indicates that the magnetic disk 24 is in a normal use state. It should be understood that, in the present invention, a user can further understand the usage status of the magnetic disk 24 during the time period by clicking the value of the K-line chart displayed in the data display device 14 during the time period.

FIG. 7 is a schematic diagram for explaining a usage trend of a magnetic disk capacity according to another embodiment of the present invention. Please refer to FIG. 1 and FIG. 7. In yet another embodiment of the present invention, when the data analysis device 12 analyzes the remaining capacity of the magnetic disk 24 from May to October using a var method, the data can be obtained by using data. The presentation device 14 displays, for example, a trend screen of the remaining capacity of the magnetic disk 24 displayed in a variable form between May and October in FIG. 7. It can be seen from FIG. 7 that the remaining capacity of the magnetic disk 24 is continuously reduced from May to July, which represents that the magnetic disk 24 is in a normal use state. Then in July, disk 24 was maintained, and from July to October, disk 24 was stable. It should be understood that, in the present invention, a user can further understand the usage status of the magnetic disk 24 during the time period by clicking the value of the K-line chart displayed in the data display device 14 during the time period.

It should be understood that the way shown in the K-line diagram in Figures 2 to 7 has several advantages, and the time period shown in Figures 2 to 7 can be adjusted according to actual needs, such as day, week, and month. Is the unit of time. For example, from the maximum magnitude of the current day / week / month K-line chart, you can understand the stability and smoothness of the analyzed resource; from the maximum decrease / maximum increase of the K-line chart for the current day / week / month Quantity, you can understand the maximum capacity of the resource being analyzed.

Further, after the data analysis device 12 analyzes the remaining capacity of the magnetic disk 24 and displays the K-line trend graphs of FIG. 2 to FIG. 7 by the data display device 14, the trend prediction device 16 can analyze the Thousands / tens of thousands of key indicator values to generate a trend forecast chart (not shown in the figure). In detail, the trend prediction device 16 automatically analyzes thousands of tens of thousands of key index values for big data analysis, for example, picks the middle value of each of thousands of tens of thousands of key index values, and uses linear regression (Linear regression The analysis method finds the rule among thousands / tens of thousands of data points, and further displays a trend prediction chart by the data display device 14, and the trend prediction chart can also be displayed in the form of a K-line chart. In this way, the present invention can use the trend prediction device 16 to predict the future increase or decrease of the remaining capacity of the magnetic disk 24 in order to prevent the detection of hidden concerns in advance. It should be understood that, in an embodiment of the present invention, the trend prediction device 16 predicts the future remaining capacity of the magnetic disk 24, and in other embodiments of the present invention, the trend prediction device 16 may predict the future usage rate of the CPU 20 , The usage status of related resources such as the memory 22's future usage rate.

Furthermore, in an embodiment of the present invention, the trend prediction device 16 may set a prediction threshold. When the trend prediction device 16 reaches the predicted threshold value in the trend prediction graph displayed by the data display device 14, At that time, the predicted key indicators that reach the predicted threshold in the trend prediction graph can use different colors to warn users to further remind users when the usage of various resources in the monitoring is in the future, such as which day / Which week / month will reach the forecast threshold and get ready early.

It is worth mentioning that, in other embodiments of the present invention, the trend prediction device 16 may have capabilities such as machine learning and artificial intelligence, and may be combined with big data analysis tools such as elastic search, Hadoop, and spark to pass historical data Accumulation, a large amount of data (Volume), a variety of data type data (Variety) at the same time for effective display (Velocity). A large amount of data is, for example, data of hundreds of thousands to millions of pens generated every day; a variety of data types are: CPU load rate, memory usage, disk capacity, network traffic and other indicators; real-time The display is, for example, providing different display modes such as instant, daily, weekly, and monthly. For example, the trend prediction device 16 may analyze the thousands / tens of thousands of key index values in FIG. 2 to FIG. 7 by a machine learning method to generate the prediction threshold, and may further dynamically adjust the prediction based on the latest data. Threshold.

FIG. 8 is a schematic diagram for explaining a daily average line trend of a usage amount of a magnetic disk capacity according to an embodiment of the present invention. Referring to FIG. 8, in an embodiment of the present invention, the data analysis device 12 may display at least one average line through the data display device 14. For example, the data analysis device 12 may use the middle key index value in a time zone as a standard value to further display the one-day average line through the data display device 14, for example, use the middle key index value every seven days as The average value of the trend can be displayed by the data display device 14 as a natural week average. If the key index value in the middle of every five days is used as the average value of the trend, the data display device 14 can be used to display a Working week average. Figure 8 shows the 20-day moving average, 30-day moving average, 40-day moving average, and 60-day moving average, so that users can more intuitively understand the multi-day average usage of disk 24.

On the other hand, another advantage of the computer resource usage trend prediction analysis system of the present invention is that it can perform trend prediction analysis on each resource of multiple computer devices at the same time. For example, each computer device has 1-N disks. Taking 100 computer devices as an example, if each computer device has 3-10 disks, there will be 300-1000 disks. The present invention The computer resource usage trend prediction analysis system can make a trend chart of the remaining capacity of the disk as shown in Figure 2 to Figure 7 for each disk. In this way, users can view the map by Figure 2 The trend chart of the remaining capacity of the disks shown in 7 is to quickly understand the usage status of each disk and solve the problems encountered in real time. Furthermore, if the memory is taken as an example, each computer device has a set of memory, and if 100-1000 computer devices are taken as an example, there will be 100-1000 groups of memory to be managed, especially if the computer device also includes a virtual machine. For the memory, the present invention can also perform unified management for the virtual memory at the same time, and display trend graphs such as shown in FIG. 2 to FIG. 7 to quickly understand the use status of each group of memory and solve the problems encountered immediately.

FIG. 9 is a flowchart for explaining a computer resource usage trend prediction analysis method according to an embodiment. Referring to FIG. 9, the computer resource usage trend prediction analysis method of the present invention includes steps S10-S16. Step S10 is: acquiring at least one key indicator value of the usage of at least one resource in at least one computer device by a data acquisition device; step S12 is: receiving and analyzing the at least one key indicator by a data analysis device Step S14 is: displaying the at least one key index value analyzed by the data analysis device by an image display device by using an image; and step S16 is: analyzing the image by a trend prediction device by using the image The at least one key indicator value is displayed, and a trend prediction chart is generated.

The data display device may display the trend forecast chart; the at least one key indicator value includes a daily start usage value, a daily end usage value, a daily maximum usage value, and a daily minimum usage value of the at least one resource. Value; the trend prediction device analyzes by linear regression to generate the trend prediction graph.

In addition, in step S16, the trend prediction device analyzes the at least one key index value displayed by the image and generates the trend prediction map. The trend prediction device further generates a prediction threshold. When at least one of the predicted values reaches the predicted threshold, the at least one predicted value that reaches the predicted threshold will be displayed on the trend prediction graph in a warning form. Furthermore, in other embodiments of the present invention, the trend prediction device may analyze the at least one key index value in the image by a machine learning method to generate the prediction threshold.

In summary, the present invention successfully provides a trend prediction analysis system and method for the usage of multiple resources of multiple computers. The present invention uses the form of a K-line diagram to display the usage status of each resource in a time zone, so that the monitoring party can observe the usage status of each resource in a time zone in a more intuitive and clear way, and The big data trend prediction and analysis is performed through the trend prediction device, so as to find out early whether each resource will be abnormal in use in the future, or when each resource will encounter problems in the future, so as to respond early.

The above are only used to explain the preferred embodiments of the present invention, and are not intended to limit the present invention in any form. Therefore, any modification or change related to the present invention made under the same spirit of the invention Should still be included in the scope of the present invention.

1‧‧‧ computer resource usage trend prediction analysis system

10‧‧‧ Data Acquisition Device

12‧‧‧Data Analysis Device

14‧‧‧ Data Display Device

16‧‧‧Trend prediction device

2‧‧‧Computer device

20‧‧‧CPU

22‧‧‧Memory

24‧‧‧Disk

S10-S16‧‧‧ steps

Those with ordinary knowledge in the art can better understand the various aspects of the present invention and its specific features and advantages after reading the following detailed description with reference to the accompanying drawings, wherein these drawings include: FIG. 1 is an illustration A structural diagram of a computer resource usage trend prediction and analysis system according to an embodiment of the present invention; FIG. 2 is a schematic diagram illustrating a usage trend of a magnetic disk capacity according to an embodiment of the present invention; FIG. 3 is a magnetic field illustrating another embodiment of the present invention. Figure 4 is a schematic diagram showing the trend of the amount of disk capacity used; Figure 4 is a schematic diagram showing the trend of the amount of disk capacity used according to another embodiment of the present invention; 6 is a schematic diagram illustrating a trend of a usage amount of a magnetic disk capacity according to another embodiment of the present invention; FIG. 7 is a schematic diagram illustrating a trend of a usage amount of a magnetic disk capacity according to another embodiment of the present invention; FIG. 9 is a flow chart illustrating a method for predicting and analyzing a computer resource usage trend according to an embodiment of the present invention. Fig.

Claims (10)

A computer resource usage trend prediction analysis system includes: a data acquisition device that acquires at least one key index value of the usage of at least one resource in at least one computer device; a data analysis device that is connected to the data Acquiring a device and receiving and analyzing the at least one key indicator value; a data display device connected to the data analysis device and displaying the at least one key indicator value analyzed by the data analysis device as an image; and The trend prediction device analyzes the at least one key index value displayed by the image to generate a trend prediction map. The trend prediction device is connected to the data display device, and the data display device displays the trend prediction map. The computer resource usage trend prediction and analysis system according to item 1 of the scope of patent application, wherein the at least one key index value includes a daily start usage value, a daily end usage value, and a daily maximum of the at least one resource. Usage value and daily minimum usage value. According to the computer resource usage trend prediction analysis system according to item 1 of the scope of the patent application, the trend prediction device performs analysis by a linear regression method to generate the trend prediction map. According to the computer resource usage trend prediction analysis system according to item 1 of the scope of patent application, wherein the at least one key indicator value is at least one KPI value, and the image is a K-line diagram. The computer resource usage trend prediction analysis system according to item 1 of the scope of patent application, wherein the usage of the at least one resource includes at least one CPU usage rate, at least one memory usage rate, and at least one magnetic disk capacity . A computer resource usage trend prediction analysis method includes the following steps: using a data acquisition device to retrieve at least one key index value of at least one resource usage in at least one computer device; receiving and using a data analysis device Analyze the at least one key index value; display the at least one key index value analyzed by the data analysis device by a data display device; and display the at least one key index value displayed by the image by a trend prediction device A key index value, and generate a trend prediction chart; wherein the data display device displays the trend prediction chart. The computer resource usage trend prediction analysis method according to item 6 of the scope of patent application, wherein in the step of analyzing the at least one key index value displayed in the image by the trend prediction device and generating the trend prediction map The trend prediction device further generates a prediction threshold value. When at least one predicted value in the trend prediction graph reaches the predicted threshold value, the at least one predicted value that reaches the predicted threshold value is displayed on the trend prediction in an alert form. On the map. According to the computer resource usage trend prediction analysis method according to item 7 of the scope of the patent application, wherein the trend prediction device analyzes the at least one key index value in the image by a machine learning method to generate the prediction threshold . The method for predicting and analyzing the trend of computer resource usage according to item 6 of the scope of the patent application, wherein the at least one key indicator value includes a daily start usage value, a daily end usage value, and a daily maximum of the at least one resource. Usage value and daily minimum usage value. According to the computer resource usage trend prediction analysis method according to item 6 of the scope of the patent application, the trend prediction device performs analysis by linear regression to generate the trend prediction map.