JPH10171682A - Generating method for program operation information - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide the generating method for program operation information which enable a user to analyze the operation of a program more efficiently. SOLUTION: A measurement/analysis part 201 obtains event information showing the operation states of respective programs (progA to progB) to be measured from an OS, etc. Then an analytic table 210 wherein the operation times of the respective programs to be measured are totalized at intervals of an analysis unit time is generated and outputted. Further, the analysis unit time is varied according to an object and an analytic drawing representing values of the analytic table 210 with a light-shade image is generated. A similar analytic table regarding the states of access of the respective programs to I/O devices can be generated. Consequently, temporal transition of the operation of each program can be grasped from the analytic table 210 and the operation time is also known.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for generating program operation information indicating an operation state of a program in a computer system.

[0002]

2. Description of the Related Art There is a computer system having a function of measuring the occupancy of a CPU as a means for presenting the operation status of a program to a user. In this computer system, the time during which the program has been running on the CPU during the specified measurement period is measured, and the total operation time during the measurement period and the ratio (CPU occupancy) are provided to the user as operation information. Can be provided. Further, when the CPU selectively executes a plurality of programs one by one by a multitasking function, the occupancy of the CPU can be measured for each program.

[0005] On the other hand, a time chart is used to visually recognize the operation status of a program.
This time chart shows the operation period of each program on the CPU in a time-series manner by a line segment, and is created by the user using information measured by the computer system. According to this time chart, the user can know the period from the start of the operation of the program to the end thereof, the operation order of each program, and the like.

[0004]

According to the conventional method using the occupancy of the CPU, the occupancy of the CPU during the measurement period can be known. It is not possible to obtain information as to whether the operation has been performed on the CPU.

In the method using a time chart, the user can know the operation order and elapsed time of each program, but cannot easily grasp the occupancy of the CPU. Further, when the program operating on the CPU is frequently switched in a very short period, the time chart is likely to be a complicated dotted line chart, and it is difficult to create and understand the operation.

Accordingly, an object of the present invention is to provide a method of generating program operation information that enables a user to more efficiently analyze the operation of a program.

[0007]

In order to solve the above problems, the present invention provides a CPU, a storage device for storing one or more programs to be analyzed, and selectively operating the programs on the CPU. Generating a program operation information indicating an operation state of the program in a computer system comprising: a program that has been operated and a unit that generates monitoring information indicating an operation period, based on the monitoring information. At each analysis unit time interval, a process of obtaining, for each program, time information indicating the total time during which the program has been operating on the CPU within the interval, and a process of obtaining the obtained time information for each program. Creating an analysis table to be displayed in a lined order as the program operation information. To.

According to the method of the present invention, it is possible to create an analysis table that allows the user to easily know the temporal transition and the operation time of the operation period of each program, so that the user can more efficiently analyze the operation of the program. You can do so.

[0009]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described.
This will be described with reference to the drawings.

FIG. 1 is a diagram showing a configuration example of a computer according to the first embodiment of the present invention. Computer 100 shown in FIG.
Comprises a CPU 101, a memory 102 for storing a processing program and processing data of the CPU, a clock 103 for providing a current time, and a plurality of I / O devices.
The I / O device includes a display 105, a keyboard 10
7, a disk device 106, a network communication interface 108, and the like. The processing programs of the CPU include an operating system (OS) 110 having a multitasking function and a program (progA, pro
gB,..., progE) 112 and an operation analysis program 111.

The OS 110 is a processing target of the CPU.
One program to be measured is sequentially allocated. When and which program runs is measured in event format at the OS 110 or middleware level.

The operation analysis program 111 is provided for the computer 10.
The measurement / analysis unit 201 shown in FIG. 2 is realized on the device 0 or a device connected thereto. The measurement / analysis unit 201 includes a measurement data acquisition unit 202, a measurement data storage unit 203, an analysis table creation unit 204, an analysis unit time setting unit 205, and an analysis diagram creation unit 2.
06. The measurement data acquisition unit 202 acquires measurement data indicating the operation history of each program measured at the OS or middleware level. The acquired measurement data is temporarily stored in the measurement data storage unit 203 and passed to the analysis table creation unit 204. The analysis table creation unit 204 aggregates the passed measurement data for each analysis unit time period instructed by the analysis unit time setting unit 205, and generates an analysis table 210.

Note that the measurement data is obtained in real time in accordance with the operation of the CPU and becomes a large amount of data depending on the measurement target. In this embodiment, the obtained measurement data is stored in the measurement data storage unit 203 in FIG. It is kept temporarily. However, by suppressing the acquisition amount of the measurement data, the acquired measurement data can be directly passed to the analysis table creation unit 204 for processing without using the measurement data storage unit.

FIG. 3 shows an example of the measurement data obtained by the measurement data obtaining unit 202.

As shown in FIG. 3, the measurement data is composed of data indicating the time when the event occurred, the name of the program targeted by the event, and the type of the event, and is collectively acquired when the event occurs. In the example of FIG.
It consists of hours, minutes, seconds and m seconds. The event includes “start” indicating the start of the process of the program, “pause” indicating the suspension of the temporary process, “resume” indicating the restart of the interrupted process, and “end” indicating the end of the process. Conventionally, a user has examined the above-described measurement data for each event, created a time chart, and calculated how much time the user has moved.

FIG. 4 shows a procedure for creating an analysis table in the analysis table creation unit 204.

The analysis table creation unit 204 reads the measurement data stored in the measurement data storage unit 203 (S1), and determines a start time and an end time of the analysis based on the time in the measurement data (S2). In the example of FIG. 3, since the occurrence time of the first event is “10: 20: 15: 188”, the start time is earlier, for example, “10: 20: 15: 000”, and the end time is the last event. For example, "10: 20: 16: 00" is later than the occurrence time "10: 20: 15: 840". Next, the analysis table creation unit 2
In 04, an analysis unit time (here, 100 msec) is obtained from the analysis unit time setting unit 205 (S3). And
The operation time of the program operated until the analysis unit time elapses from the start time of the analysis is calculated for each program (S4), and the table managed by the analysis table creation unit 204 is stored in the table.
The operated program and its operation time (excluding "0") are registered in an analysis table format (S5). In the example of FIG. 3, no program is running during the period from time “10: 20: 15: 000” until 100 ms elapses, so no program and operating time are registered. Next, the analysis table creation unit 204 determines whether or not the processes in S4 and S5 have been performed until the end time described above (S6). To end. If the processing has not been performed until the end time, the processing of S4 and S5 is repeated for the subsequent analysis unit time (100 msec) (S6). In the example of FIG. 3, the time “10: 20: 15: 100”
From "10: 20: 15: 200", program A is "10: 20: 15: 1"
Since the operation has been performed for 12 msec from "88", the program A and the operation time "12" are registered in the corresponding areas as shown in FIG. And from the next "10: 20: 15: 200" to "10: 20: 1
At 5: 300, Program A changes from 10: 20: 15: 200 to 1
0: 20: 15: 225 "for 25 ms, and program B
Has been operating for 30 msec from "10: 20: 15: 270", program B is newly registered, and each operation time is registered. This process is repeated every analysis unit time (100 ms) until the end time, whereby the analysis table shown in FIG. 6 is completed.

The advantages of the analysis table shown in FIG. 6 will be specifically described below.

(1) From the analysis table, the user can easily grasp the outline of the operation order of a plurality of programs. For example, from the analysis table of FIG. 6, it can be easily understood that the program to be measured operates in the order of the programs A, B, and C, and the program A operates again. In the conventional method, the operation order of the program could be known from the measurement data shown in FIG. 3, but a large amount of event information had to be followed, which was difficult.

(2) From the analysis table, the user can easily grasp the status of the operation of each program. From the analysis table of FIG. 6, for example, it can be seen that the program B intermittently operates by interrupting and resuming the processing from the start to the end of the operation.

(3) From the analysis table, the user can easily know the outline of the time (response time) from the start of the program to the end thereof. From the analysis table in FIG.
For example, if the program B has three consecutive
It can be seen that the operation is performed over one analysis unit time (about 300 ms). If it is desired to know a more detailed operation state, the event information shown in FIG.

(4) Further, the user can easily calculate the operation time on the CPU for each program from the analysis table. The operation time of a program is important information for faulty measures and verification. In the analysis table of FIG. 6, when focusing on the program A, it can be easily calculated that the processing time is moving for 37 msec in the first operation and 55 msec when the operation is restarted next. Conventionally, in order to obtain such a result, it is necessary to perform a troublesome operation such as calculating event information on a target program one by one using a calculator or the like.

The programs to be analyzed may be divided into groups, and the operation time may be totaled for each group. In a computer, a plurality of programs may be processed as a set. In such a case, it is convenient to be able to check the operation status and operation time for each group. FIG. 7 shows an example of an analysis table in which programs A and B are totaled as one group. In the analysis table of FIG. 7, a column 71 in which the operation times are totaled for each group, a column 72 in which the operation times of all programs are totaled are added, a symbol 73 indicating the type of event that has occurred, A symbol 74 representing activation and return is displayed at the position of the corresponding program and unit time. This example shows a case where the program B is a subroutine of the program A. According to the analysis table of FIG. 7, it is easier to grasp the operation status of the program.

Next, the analysis unit time setting unit 2 shown in FIG.
05 will be described.

The analysis unit time setting unit 205 determines the analysis unit time used when creating the analysis table based on one of the average operation time of the program, the operation time of the program to be analyzed, the operation cycle time, and the set value. To decide. The parameters used to determine the analysis unit time are selected according to the user's instruction.

First, a method of determining the analysis unit time from the average operation time of each program will be described.

In a multitasking computer control system,
The program starts and ends repeatedly in the range of several milliseconds to several hundred milliseconds, and a plurality of programs operate while switching. In such a case, the method determines the analysis unit time based on the average operation time of each program (that is, the average value of the elapsed time from the start of each program to the end).

FIG. 8 shows an example of processing performed by the analysis unit time setting unit 205 according to the present method. Analysis unit time setting unit 2
In step 05, the elapsed time from the start to the end of each program to be measured is obtained based on the acquired measurement data, and the averaged elapsed time is obtained to obtain an average operation time (S11). Then, the most significant digit of the obtained average operation time is set to “1” (S1).
2), all lower digits are set to "0" (S13)
The obtained numerical value is set as the analysis unit time (S14). The analysis unit times obtained in this way are 100 ms, 10 ms, and 1 ms.
It becomes a sharp number such as seconds.

With some exceptions, in many computer control systems, the average operation time is several hundred milliseconds, and thus the analysis unit time is 100 milliseconds. In the case of the measurement data of FIG.
The elapsed time from the start to the end of each program is 652 ms for program A, 140 ms for program B, and
Is 255 ms and their average operation time is 349 ms, so the analysis unit time is 100 ms. When the analysis unit time is 100 ms, the operation time [ms] in each column in the analysis table is
The occupancy of the CPU is indicated as it is. Even when the analysis unit time is 10 ms or 1 second, the occupancy can be easily known by multiplying the operation time [msec] in the analysis table by 10 times or 1/10.

Next, a method for determining the analysis unit time from the operation time of the program to be analyzed will be described.

In the above-described method, when the operation time of a specific program to be analyzed is significantly different from that of another program, the analysis unit time does not match the analysis target program. For example, when the analysis unit time is extremely small for the program to be analyzed, one operation period is represented by numerical values in many columns in the analysis table, and it is difficult to calculate the operation time. For this reason, in this method, instead of the average operation time, the operation time of the program to be analyzed specified by the user (the elapsed time from the start to the end of the operation)
Use to determine the analysis unit time. Other than this point, FIG.
Method is the same. According to the present method, an analysis unit time suitable for a program to be analyzed can be determined, thereby making it possible to create an analysis table that facilitates analysis of the program to be analyzed.

Next, a method of determining the analysis unit time based on the operation cycle time will be described.

For a program such as a monitoring program that operates periodically at regular intervals, it may be desired to grasp the behavior of the program for each operation cycle. In this method, the operation cycle time (elapsed time from the start of the program operation to the next start) of the program that operates periodically is set as the analysis unit time. The operation cycle time is determined using a value specified by the user or based on measurement data. By determining the analysis unit time by this method, it is possible to create an analysis table that makes it easy to grasp the behavior of the program for each cycle.

Next, a method for determining the analysis unit time based on the set value will be described.

The analysis unit time setting unit 205 has a correspondence table in which different analysis unit times are set depending on the nature of the system to be analyzed and the type of analysis. FIG. 9 shows an example of the registered contents of the correspondence table. The correspondence table has “summary analysis” and “detailed analysis” as analysis levels.
In the example of FIG. 9, “traffic monitoring system” and “robot control system” are registered as target systems, and the analysis unit time for each target system is individually registered for each analysis level.

The analysis unit time setting unit 205 determines the analysis unit time registered in the correspondence table corresponding to the target system and analysis level specified by the user as the analysis unit time for creating the analysis table. In the example of FIG. 9, when “detailed analysis” of the “traffic monitoring system” is designated, the analysis unit time is determined to be 100 ms. As described above, according to the present method, it is possible to determine the analysis unit time suitable for the target system and the analysis level.

Next, the analysis diagram creation unit 206 will be described.

The analytic chart creator 206 includes the analytic chart creator 20
In the analysis table created in step 4, an analysis chart is created in which the magnitude of the numerical value in each cell is changed to a shade of color. FIG.
Fig. 1 shows an example of the analysis diagram. Here, the CP of each program
The portion where the occupancy of U is 100% is shown in black, the density is increased as the occupancy becomes smaller, and the portion of 0% is shown in white. In other words, in this analysis diagram, it can be seen that the closer the color is to black, the more the CPU operates using the CPU, and it is possible to visually easily understand how each program operates on the CPU.

The analysis diagram can be created as a graph having a time axis and an operation time axis as shown in FIG. Further, it is also possible to create an analysis diagram as shown in FIG. In the analysis diagram of FIG. 12, the occupancy of the CPU is expressed in the form of contour lines, and is shaded according to the magnitude of the occupancy. From the example of Fig. 12, the user can say that "Program A is operating in a two-second cycle" or "Program A has not completely moved at 9: 20: 57: 780. The amount of movement that has not been completed is taken into the next cycle (9: 20: 59: 78).
0)).

As described above, according to the computer system of the present embodiment, the operation status of each program can be grasped from the viewpoint of a time chart, and at the same time, the value of the operation time of each program can be known. A table can be provided. Further, by changing the analysis unit time according to the measurement state or the system or program to be measured, it is possible to provide an analysis table more suitable for the analysis content. Further, it is possible to provide an analysis diagram that allows the operation status of each program to be easily grasped visually.

Next, a computer system according to a second embodiment of the present invention will be described.

The configuration of the computer system according to this embodiment is basically the same as the configuration shown in FIGS. However, in the computer system of the present embodiment, the analysis table to be created is based on the I / O
The difference is that the usage status of the O device is represented.

The measurement data acquisition section acquires event information of an I / O device (for example, the disk device 106) specified by a user as measurement data used for creating an analysis table. The OS also manages access to I / O devices,
The measurement data acquisition unit acquires the event information from the OS or the middleware. FIG. 13 shows an example of measurement data of event information measured for the I / O device.

The analysis table creation section creates an analysis table in the same procedure as in FIG. However, in this processing, instead of calculating the operation time of the program (S4), the I-value of each program is used.
Counts the number of accesses to the / O device for each analysis unit time,
The result of the aggregation is set on the analysis table. FIG. 14 shows an analysis table created based on the event information in FIG. Here, the reason why the number of accesses is used as the registration target instead of the time required for the access is that a time lag occurs due to buffering of the access data in the I / O device. Of course, an analysis table of the time required for access may be created in the same manner as in FIG.

Note that the measurement may be performed in parallel for each of the plurality of I / O devices, and the totalized results may be arranged and displayed in one analysis table. Further, by combining with the functions of the first embodiment, it is also possible to realize a computer capable of creating both an analysis table indicating an operation state on a CPU and an analysis table indicating an access state to an I / O device. . Further, the contents of the two types of analysis tables may be incorporated into one analysis table. In some cases, the use of two measurement results in this manner makes analysis easier.

[0046]

According to the present invention, it is possible to generate program operation information that enables a user to more efficiently analyze the operation state of a program.

[Brief description of the drawings]

FIG. 1 is a configuration example of a computer system according to a first embodiment of the present invention.

FIG. 2 is a configuration diagram of a measurement / analysis unit realized by an analysis program.

FIG. 3 is an example of event information (measurement data) indicating a result of measuring an operation state of each program on a CPU.

FIG. 4 is an example of a process flow of a process of creating an analysis table.

FIG. 5 is a diagram illustrating an example of an analysis table in a creation process.

FIG. 6 is a diagram showing an example of an analysis table at the time of completion.

FIG. 7 is a diagram showing another example of creating an analysis table.

FIG. 8 is an example of a process flow of a process for determining an analysis unit time.

FIG. 9 is an example of a table in which an analysis unit time is set in accordance with an analysis target system and an analysis level.

FIG. 10 is an example of an analysis diagram in which the values of the analysis table are represented by grayscale images.

FIG. 11 is an example of an analysis diagram in which values in an analysis table are represented in a graph.

FIG. 12 is an example of an analysis diagram in which values in an analysis table are represented by a grayscale image in a contour line format.

FIG. 13 is an example of event information obtained by measuring the use status of an I / O device for each program.

FIG. 14 is an example of an analysis table summarizing the use status of the I / O device.

[Explanation of symbols]

100: Computer system, 201: Measurement / analysis unit, 2
02: Measurement data acquisition unit, 203: Measurement data storage unit, 204: Analysis table creation unit, 205: Analysis unit time setting unit, 206: Analysis diagram creation unit

Claims (10)

[Claims] 1. A CPU, a storage device for storing one or a plurality of programs to be analyzed, and selectively operating the programs on the CPU to generate monitoring information indicating the operated programs and operation periods. Means for generating program operation information indicating an operation state of the program in a computer system comprising: a monitoring unit, based on the monitoring information, at intervals of a predetermined analysis unit time,
A process for obtaining, for each program, time information indicating the total time during which the program has been operating on the CPU within the interval, and an analysis table for displaying the obtained time information in chronological order for each program And generating a program operation information as the program operation information. 2. The method according to claim 1, further comprising the step of determining the analysis unit time based on designated information or the monitoring information. How information is generated. 3. The method for generating program operation information according to claim 2, wherein: the magnitude of the time represented by each piece of time information displayed in the created analysis table and the temporal change of the magnitude are determined. A method for generating an array of color patterns having a plurality of levels of densities in accordance with the magnitude of time represented by time information, or an analysis diagram represented by the shape of a figure. 4. A method for generating program operation information according to claim 2, wherein a table in which time values are registered corresponding to a plurality of types of parameters defining details of the analysis is prepared, and A method for generating program operation information, comprising a step of determining a time value registered in the table corresponding to a parameter as the analysis unit time. 5. The method for generating program operation information according to claim 2, wherein a plurality of programs to be analyzed are allocated to a plurality of groups, and a total of the time information for each analysis unit time is calculated for each group. A method of generating the program operation information, wherein in the process of creating the analysis table, a total of the time information for each analysis unit time is also displayed in time series for each of the groups. . 6. The method for generating program operation information according to claim 2, wherein the monitoring information includes starting, stopping,
Event information indicating which event of restart or end has occurred is also included. In the process of creating the analysis table, a symbol indicating the content of the event information is displayed at a corresponding position in the analysis table. A method for generating program operation information, characterized in that: 7. The method for generating program operation information according to claim 2, wherein the monitoring information includes event information indicating whether the program has started or returned to another program. The processing for creating a table is a method of generating operation information, wherein symbols indicating the activation and the return after the activation are displayed at corresponding positions in the analysis table. 8. A CPU, an I / O device, a storage device for storing one or a plurality of programs to be analyzed, selectively operating the program on the CPU, and executing the program and the program. Means for generating monitoring information indicating the time at which the I / O device was accessed. A method for generating program operation information representing an operation state of the program in a computer system, comprising: For each analysis unit time interval of
Processing for obtaining, for each program, frequency information indicating the total number of times the program has accessed the I / O device within the interval, and analysis for displaying the obtained frequency information in time series for each program; And generating a table as the program operation information. 9. A CPU, a storage device for storing one or a plurality of programs to be analyzed, and selectively operating the programs on the CPU to generate monitoring information indicating the operated programs and operation periods. A computer system comprising: means for generating program operation information by the method for generating program operation information according to claim 1, 2, 3, 4, 5, 6, or 7. 10. A recording medium on which a program for realizing the method for generating program operation information according to claim 1, 2, 3, 4, 5, 6, 7, or 8 is recorded.