JP2011258058A - Performance prediction apparatus, computer program and performance prediction method - Google Patents

Abstract

Even when an application program cannot be executed on a certain computer system, the time required to execute the application program on the computer system is accurately predicted.
A log analysis unit 114 (execution log acquisition unit) acquires an application log 107 (execution log) obtained by executing an application 102 in a computer system 101. The log analysis unit 114 (total execution time calculation unit, second execution time calculation unit) analyzes the execution log and calculates the time taken to execute the application 102 or the time taken for a specific process. The new system performance estimation unit 118 (total execution time prediction unit) takes the time required for specific processing out of the time required to execute the application 102 in the computer system 101 for the specific processing in the computer system 102. Replace with time.
[Selection] Figure 1

Description

  The present invention relates to a performance prediction apparatus that predicts the time required to execute an existing application program in a computer system under development.

A benchmark test is a technique for measuring the performance of a computer system. The benchmark test measures the general performance of a computer system by causing the computer system to execute some general processes and measuring the time taken for the processes.
In a computer system in which an application program to be executed is limited, such as an embedded system, performance when the application program is executed is more important than general performance. The simplest method for measuring the performance when the application program is executed is to actually execute the application program on a computer system.

JP-A-8-286940 Japanese Patent Laid-Open No. 10-78892 JP 2000-122898 A JP 2004-264914 A

EEMBC “AutoBench ™ 1.1 software benchmark data book” http: // www. embc. org / techlit / datasheets / autobench_db. pdf

In some cases, such as when the computer system is under development, the application program cannot actually be executed on the computer system. In addition, a general performance ratio measured by a benchmark test or the like between two computer systems does not always apply to a performance ratio when the application program is executed as it is.
The present invention has been made to solve the above-described problems, for example, and even when an application program cannot be executed on a certain computer system, the time taken to execute the application program on the computer system. The purpose is to accurately predict.

The performance prediction apparatus according to the present invention is:
A processing device that processes data, a first execution time acquisition unit, an execution log acquisition unit, a total execution time calculation unit, a second execution time calculation unit, and a total execution time prediction unit;
The first execution time acquisition unit acquires the time required to execute the first specific program for the first specific program that realizes a specific function in the first computer system, using the processing device,
The execution log acquisition unit uses the processing device to obtain an execution log for executing the application program in the second computer system for an application program executable in the first computer system and the second computer system. Acquired,
The total execution time calculation unit uses the processing device to analyze the execution log acquired by the execution log acquisition unit and calculate the time taken to execute the application program,
The second execution time calculation unit analyzes the execution log acquired by the execution log acquisition unit using the processing device, and performs the specific function realized by the first specific program in the first computer system. For the second specific program realized in the second computer system, when the application program calls the second specific program, it calculates the time taken to execute the second specific program,
The first execution time acquisition unit acquires the time calculated by the second execution time calculation unit among the times calculated by the total execution time calculation unit, using the processing device. By replacing with time, the time required to execute the application program in the first computer system is predicted.

  According to the performance prediction device of the present invention, even when the application program cannot be executed on the first computer system at the present time, the time required to execute the application program on the first computer system can be accurately predicted. .

1 is a system configuration diagram of a performance estimation system according to Embodiment 1. FIG. FIG. 3 is an overall flowchart of new system performance estimation processing in the first embodiment. FIG. 6 is a diagram showing an example of an application log 107 according to the first embodiment. FIG. 7 is a flowchart of log analysis step 204 in the first embodiment. FIG. 6 is a detailed flowchart of event switching matching processing step 404 according to the first embodiment. FIG. 6 is a detailed flowchart of exclusive control matching processing step 406 according to the first embodiment. FIG. 4 is a detailed flowchart of Semaphore / Mutex matching processing step 409 according to the first embodiment. FIG. 3 is a detailed flowchart of new system performance estimation processing in the first embodiment. FIG. 10 is a system configuration diagram illustrating an example of an overall configuration of a performance prediction system 800 according to a third embodiment. FIG. 11 is a hardware configuration diagram illustrating an example of a hardware configuration of computer systems 810 and 820 according to the third embodiment. The figure which shows an example of the structure of the program which the computer system 810 in Embodiment 3 performs. FIG. 14 is a diagram illustrating an example of a configuration of a program executed by a computer system 820 in the third embodiment. FIG. 10 is a perspective view illustrating an example of an appearance of a performance prediction apparatus 100 according to a third embodiment. FIG. 10 is a diagram illustrating an example of hardware resources of the performance prediction apparatus 100 according to the third embodiment. FIG. 10 is a block configuration diagram illustrating an example of a functional block configuration of a performance prediction apparatus 100 according to a third embodiment. FIG. 10 is a flowchart showing an example of a flow of performance prediction processing S610 in the third embodiment.

In an embedded computer system, when an operating system (hereinafter referred to as “OS”) or hardware (hereinafter referred to as “HW”) is changed from an existing computer system and migrated to a new computer system, a performance estimate is required before development. There is a need to do.
As a performance estimation method for a computer system, there is a method for changing the combination of a plurality of benchmarks depending on the field of embedded devices and estimating the performance of the target embedded device field.
There is also a method of automatically recognizing the configuration of the evaluation system, extracting an appropriate evaluation target program, executing the evaluation target program, and performing evaluation.
In such a computer system performance estimation method, it is not possible to estimate the system performance in accordance with the characteristics of the application.

The performance prediction device described below measures basic performance such as OS basic processing and file access on a new / existing computer, acquires an application log on the existing computer system, and uses the actual value using unit and the performance ratio. Using the basic performance measured on the new computer for the measured value used part, and using the performance ratio of the new computer and the existing computer measured by the benchmark for the performance ratio used part, the performance of the new computer system Make an estimate.
Since the performance prediction apparatus can estimate the performance of the new system according to the characteristics of the application without moving the application on the new computer system, the performance prediction of the new system can be performed in detail before development.

Embodiment 1 FIG.
The first embodiment will be described with reference to FIGS.

  FIG. 1 is a system configuration diagram of a performance estimation system according to this embodiment.

The existing computer system 101 (second computer system) includes an existing application 102 (application program) that is a performance estimation target, middleware 103 that is a software layer that absorbs differences between operating systems, a log acquisition unit 104, an existing It comprises an operating system 105 on a computer system and hardware 106 on an existing computer system.
The log acquisition unit 104 (execution log acquisition unit) acquires a middleware API, a time stamp immediately before and after the middleware API starts, and an argument of the middleware, and acquires a thread ID indicating in which thread the middleware is executed. Also, the system call called in the middleware, the time stamp immediately before and after the start of the system call, and the argument of the system call are obtained.
The application log 107 (execution log) is an application log that can be acquired by the log acquisition unit 104 when an application that is a performance estimation target is executed on an existing computer system.

A new or existing computer system 108 (first computer system) includes a basic performance acquisition unit 109 (first execution time acquisition unit) that acquires basic performance such as OS basic processing and file access, and a new / existing computer. It comprises an operating system 111 and hardware 112 on new and existing computers.
The configuration of the new / existing operating system 111 and the new / existing hardware 112 in the computer system 108 may be a minimum configuration that can acquire basic performance such as an evaluation board.
The basic performance result 113 is a basic performance result acquired on a new / existing computer.

The log analysis unit 114 (total execution time calculation unit / second execution time calculation unit) analyzes the application log 107 obtained by the log acquisition unit 104 and obtains a log analysis result 115.
The log analysis result 115 includes an actual value use unit 116 that uses the basic performance result 113 measured by the new / existing computer, and a performance ratio use unit 117 that uses the performance ratio of the new computer and the existing computer measured by the benchmark. .

  The new system performance estimation unit 118 (total execution time prediction unit) uses the log analysis result 115 obtained by the log analysis unit 114, and uses the basic performance obtained by the new computer for the actual measurement value use unit 116. The ratio use unit 117 obtains a new system performance estimation result 119 by using the result of adding the performance ratio of the new / existing computer to the result of the existing system.

FIG. 2 is an overall flowchart of the new system performance estimation process in this embodiment.
The flow when using the measured value use unit of the new system as the OS basic performance and using Windows (registered trademark) CE as the OS will be described as an operation example.

In step 201, the basic performance acquisition unit 109 measures the OS basic performance on the existing / new system and executes a benchmark.
In step 202, the new system performance estimation unit 118 calculates the existing / new performance ratio from the benchmark execution result.
In step 203, the log acquisition unit 104 acquires a log of an existing application.
In step 204, the log analysis unit 114 performs log analysis of the existing application obtained in step 203 and outputs a log analysis result 205.
In step 206, the new system performance estimation unit 118 performs performance estimation of the new system based on the log analysis result 205 and outputs a new system performance estimation result 207.

FIG. 3 is a diagram showing an example of the application log 107 in this embodiment.
ID 301 is the ID of the thread that output the log.
The function name 302 is the name of the middleware API and the system call. In the case of the middleware API, the function name 302 can be distinguished from the system call such as Mw_xxx.
An argument 303 is an argument of middleware API or system call.
The time stamp 304 is a time stamp at the time of log output.

  The log output location is in the middleware function. Logs are acquired immediately after the middleware API starts and immediately before the end, and immediately before and after the start of the system call.

FIG. 4 is a flowchart of the log analysis step 204 in this embodiment.
In step 401, the log analysis unit 114 determines whether to analyze the OS basic process. If it is determined that the analysis is not an OS basic process analysis, in step 402, the log analysis unit 114 performs other analysis processes.
If it is determined in step 401 that OS basic processing is to be analyzed, and if there is an event switching command in the log in step 403, the log analysis unit 114 performs event switching matching processing in step 404.
If there is an exclusive control command in the log in step 405, in step 406, the log analysis unit 114 performs exclusive control matching processing.
If there is a Semaphore instruction in the log in Step 407 or if there is a Mutex instruction in the log in Step 408, the log analysis unit 114 performs Semaphore / Mutex matching processing in Step 409.
If none of Step 403, Step 405, Step 407, and Step 408 is applicable, in Step 410, the log analysis unit 114 checks that it is not a location measured by the OS basic process.

FIG. 5 is a detailed flowchart of the event switching matching processing step 404 in this embodiment.
In step 501, if the thread whose execution wait has been released (SetEvent) is immediately after the end of execution wait (WaitForSingleObject), the log analysis unit 114 sets SetEvent: Start and WaitForSingleObject: End. Check as event switching processing.
This is the time from when a thread that has entered the wait state with the WaitForSingleObject instruction is released from the wait state with SetEvent until the thread is switched.

  If a different log appears in step 501, the log analysis unit 114 determines that event switching has not occurred. In step 503, the log analysis unit 114 checks that it is not a location measured by OS basic performance measurement.

FIG. 6 is a detailed flowchart of the exclusive control matching process step 406 in this embodiment.
“No thread priority reversal after exclusive control resource reservation block” cannot extract the OS basic processing unit only by the log, and therefore requires a work to separate from the part other than the OS basic processing.

  If the same resource is attempted to be secured before releasing the resource in step 601, and the resource securing (Enter Critical Section) thread has a lower priority than the resource release waiting thread in step 602, the log analysis unit 114 has secured the resource in step 603. After the thread releases the resource (Leave Critical Section: Start), the thread waiting for the resource release finishes securing the resource (Enter Critical Section: End) as the location measured by the OS basic performance measurement.

If the priority of the resource securing (Enter Critical Section) thread is higher than that of the resource release waiting thread in Step 602, in Step 604, the log analysis unit 114 tries to secure the resource (Enter Critical Section: Start) immediately after the thread is waiting for the resource release. The log is checked as a location measured by OS basic performance measurement.
However, since the portion measured by the OS basic performance measurement cannot be extracted from the portion corresponding to Step 604 only by the log, it is necessary to separate the portion from the portion other than the OS basic processing.

In step 605, the log analysis unit 114 separates the part measured by the OS basic performance measurement from the other part.
In step 606, the log analysis unit 114 calculates the measurement time corresponding to step 604 from the time stamp.
In step 607, the log analysis unit 114 subtracts the corresponding OS basic processing time in step 604 from the time acquired in step 606, and calculates a time other than the OS basic processing from the time acquired in step 606.
In step 608, the log analysis unit 114 checks that the time calculated in step 607 is not the location measured by the OS basic performance measurement, and can divide it into the location measured by the OS basic performance measurement and other locations.

  If it is determined in step 601 that the same resource is not secured before releasing the resource, in step 609, the log analysis unit 114 determines that the measured OS basic processing of exclusive control has not occurred, and the location measured by OS basic performance measurement Check that it is not.

FIG. 7 is a detailed flowchart of Semaphore / Mutex matching processing step 409 in this embodiment.
In “Semaphore” or “Mutex”, the OS basic processing unit cannot be extracted only by the log, and therefore, it is necessary to separate it from parts other than the OS basic processing.

In step 701, when the semaphore is to be released (ReleaseSemaphore), in step 702, the log analysis unit 114 attempts to release the semaphore (ReleaseSemaphore: Start) and then measures the immediately following log by OS basic performance measurement. Check as done.
If the semaphore is not released in step 701, the mutex is released (ReleaseMutex). Therefore, in step 703, the log analysis unit 114 sets the mutex to be released (ReleaseMutex: Start), and then logs the next log. Check as a location measured by OS basic performance measurement.
However, the portions corresponding to step 702 and step 703 cannot be extracted from the portion measured by the OS basic performance measurement only by the log, and therefore, the work separated from the portions other than the OS basic processing is necessary.

In step 704, the log analysis unit 114 separates the part measured by the OS basic performance measurement from the other part.
In step 705, the log analysis unit 114 calculates the measurement time corresponding to step 702 and step 703 from the time stamp.
In step 706, the log analysis unit 114 subtracts the corresponding OS basic processing time in steps 702 and 703 from the time acquired in step 705, and calculates a time other than the OS basic processing from the time acquired in step 705.
In step 707, the log analysis unit 114 checks that the time calculated in step 706 is not the location measured by the OS basic performance measurement, and can divide it into the location measured by the OS basic performance measurement and other locations.

FIG. 8 is a detailed flowchart of the new system performance estimation process in this embodiment.
In the case of the processing unit measured by the basic performance measurement in step 801 based on the log analysis result 205, in step 802, the new system performance estimation unit 118 corresponds to the corresponding basic processing performance value measured on the new system. Replace with
If it is not the processing unit measured in the basic performance measurement in step 801, in step 803, the new system performance estimation unit 118 replaces the measurement time that is not the processing unit measured in the basic performance measurement with the new / existing performance ratio.

  As described above, measure the basic performance on the new existing computer, calculate the performance ratio of the new / existing computer, acquire the log of the existing application, analyze the log, and use the measured value use part and the performance ratio use part Since the performance estimation of the divided new system is performed, it is possible to estimate the performance of the new system according to the characteristics of the existing application without moving the existing application on the new system.

  The performance estimation method described above calculates the application processing performance when an existing application is operated on a computer system with a different OS or HW based on the operation log.

  In this performance estimation method, the processing performance of an application is calculated by combining the result of measuring the OS basic performance with the execution time measured by the operation log.

Embodiment 2. FIG.
In the first embodiment, the OS basic performance measured on the new system is used as the actual measurement value, but instead of the OS basic performance, the file access performance and graphic performance are measured on the new system and the actual measurement value is obtained. It is good.

Embodiment 3 FIG.
A third embodiment will be described with reference to FIGS.
In addition, about the part which is common in Embodiment 1- Embodiment 2, the same code | symbol is attached | subjected and description is abbreviate | omitted.

FIG. 9 is a system configuration diagram showing an example of the overall configuration of the performance prediction system 800 in this embodiment.
The performance prediction system 800 includes a performance prediction apparatus 100 and two computer systems 810 and 820. The performance prediction system 800 predicts the result when the computer system 820 executes the application program 830 based on the result of the execution of the application program 830 by the computer system 810, and evaluates the performance of the computer system 820.

  The computer system 810 (second computer system) is an already operating system and can execute various programs including the application program 830.

On the other hand, the computer system 820 (first computer system) is a system under development and is incomplete. The application program 830 is a program that can be executed by the computer system 820. However, since the computer system 820 is incomplete, the computer system 820 cannot execute the application program 830 at this time. However, the computer system 820 can execute several programs such as a basic program.
For example, the computer system 820 is an embedded system, and is a system that operates by being incorporated in a device to be controlled. However, since the computer system 820 is under development, it is not yet incorporated into the control target device. Since the application program 830 is a program that controls the control target device, it cannot be executed by the computer system 820 that is not incorporated in the control target device.

  In such a situation, the performance prediction apparatus 100 predicts the time required for the computer system 820 to execute the application program 830. The user of the performance prediction system 800 evaluates the performance of the predicted computer system 820 based on the result predicted by the performance prediction device 100, and takes measures such as changing the configuration of the computer system 820 if necessary. To do. Thereby, a computer system 820 having desired performance can be developed.

FIG. 10 is a hardware configuration diagram showing an example of the hardware configuration of the computer systems 810 and 820 in this embodiment.
The computer system 810 includes, for example, a processing device 811, an input device 812, an output device 813, and a storage device 814. The same applies to the computer system 820.

The processing device 811 processes the data stored in the storage device 814 by executing the program stored in the storage device 814, and controls the computer system 810 as a whole.
The storage device 814 stores a program executed by the processing device 811 and data processed by the processing device 811.
The input device 812 inputs information from outside the computer system 810 and converts it into data that can be processed by the processing device 811. The data converted by the input device 812 may be processed directly by the processing device 811 or temporarily stored by the storage device 814.
The output device 813 converts the data processed by the processing device 811 and the data stored in the storage device 814 and outputs the converted data to the outside of the computer system 810.

FIG. 11 is a diagram showing an example of the configuration of a program executed by the computer system 810 in this embodiment.
The computer system 810 executes programs such as an operating system 815, middleware 816, a benchmark program 817, a log generation program 818, and an application program 830. These programs are stored in the storage device 814 and executed by the processing device 811.

The operating system 815 manages hardware resources of the computer system 810 and allocates hardware resources to programs such as an application program 830 that operates on the operating system 815. The operating system 815 provides system calls to programs such as the application program 830. When a program such as the application program 830 calls a system call, various functions of the operating system 815 such as input / output processing are provided.
Examples of functions provided by the operating system 815 to programs such as the application program 830 include a file access function. The file access function is a function for performing operations such as opening, closing, reading, and writing on a file stored in the storage device 814 or the like.
The functions provided by the operating system 815 to programs such as the application program 830 include, for example, an event switching function, an exclusive control function, and a synchronization control function. In the event switching function, there is a thread waiting for an event such as completion of I / O processing, and when the event that the thread is waiting for occurs, the priority of the executing thread and the waiting thread If the priority of the waiting thread is higher than that of the waiting thread, this function switches the executing thread to the waiting state and resumes the waiting thread. In order to prevent inconsistency due to multiple threads updating one data, the exclusive control function, when one thread is updating data, when another thread tries to update the data, This function waits for another thread until the data update is completed. The synchronization control function is a function of waiting for another thread that uses data updated by a certain thread to complete the update of data in order to cause a plurality of threads to cooperate.

  The middleware 816 is a program that extends the functions of the operating system 815 and provides general-purpose functions to programs such as the application program 830. The functions provided by the middleware 816 to programs such as the application program 830 include, for example, an operating system 815 such as a function for realizing more advanced processing such as a database management function, an extended file access function, a graphic function, and a network function. There is a function that complements the function.

  The application program 830 is a program that performs predetermined processing. The application program 830 calls and uses functions provided by the operating system 815 and the middleware 816.

  The benchmark program 817 is a program for measuring the performance of the computer system 810. The benchmark program 817 measures the time taken to execute a predetermined calculation, and measures the processing speed of the processing device 811. In addition, the benchmark program 817 calls a function provided by the operating system 815 and the middleware 816, measures the time taken for processing, and measures the processing speed of the operating system 815 and middleware 816.

  The log generation program 818 is a program that records the occurrence of an event such as a call of a function provided by the operating system 815 or the middleware 816, or an end of processing of the called function. The log generation program 818 generates an execution log in which event occurrence times and contents are recorded.

FIG. 12 is a diagram showing an example of the configuration of a program executed by the computer system 820 in this embodiment.
The computer system 820 executes programs such as an operating system 825, middleware 826, and a benchmark program 827. These programs are stored in the storage device 814 and executed by the processing device 811.
Since the operating system 825 is the same as the operating system 815, detailed description thereof is omitted. Note that the operating system 825 may be the same OS as the operating system 815 or a different OS.
Since the middleware 826 is the same as the middleware 816, detailed description is omitted. The middleware 826 may be the same middleware as the middleware 816, or may be different middleware that provides an equivalent function.
Since the benchmark program 827 is the same as the benchmark program 817, detailed description is omitted. The benchmark program 827 measures the performance of the computer system 820 for the same items as the items measured by the benchmark program 817 so that the measurement results obtained by the benchmark program 817 can be compared.

  The application program 830 can be executed by the computer system 820 as a general rule, although fine correction such as recompilation may be necessary. However, as described above, it is assumed that the computer system 820 cannot execute the application program 830 at the present time.

FIG. 13 is a perspective view showing an example of the appearance of the performance prediction apparatus 100 according to this embodiment.
The performance prediction device 100 includes a system unit 910, a display device 901 having a CRT (Cathode / Ray / Tube) or LCD (liquid crystal) display screen, a keyboard 902 (Key / Board: K / B), a mouse 903, an FDD 904 (Flexible). (Disk / Drive), compact disk device 905 (CDD), printer device 906, scanner device 907, and other hardware resources, which are connected by a cable or a signal line.
The system unit 910 is a computer, and is connected to the facsimile machine 932 and the telephone 931 via a cable, and is connected to the Internet 940 via a local area network 942 (LAN) and a gateway 941.

FIG. 14 is a diagram illustrating an example of hardware resources of the performance prediction apparatus 100 according to this embodiment.
The performance prediction apparatus 100 includes a CPU 911 (also referred to as a central processing unit, a central processing unit, a processing unit, an arithmetic unit, a microprocessor, a microcomputer, and a processor) that executes a program. The CPU 911 is connected to the ROM 913, the RAM 914, the communication device 915, the display device 901, the keyboard 902, the mouse 903, the FDD 904, the CDD 905, the printer device 906, the scanner device 907, and the magnetic disk device 920 via the bus 912, and the hardware. Control the device. Instead of the magnetic disk device 920, a storage device such as an optical disk device or a memory card read / write device may be used.
The RAM 914 is an example of a volatile memory. The storage media of the ROM 913, the FDD 904, the CDD 905, and the magnetic disk device 920 are an example of a nonvolatile memory. These are examples of a storage device or a storage unit. A communication device 915, a keyboard 902, a scanner device 907, an FDD 904, and the like are examples of an input unit and an input device.
Further, the communication device 915, the display device 901, the printer device 906, and the like are examples of an output unit and an output device.

The communication device 915 is connected to a facsimile machine 932, a telephone 931, a LAN 942, and the like. The communication device 915 is not limited to the LAN 942, and may be connected to the Internet 940, a WAN (wide area network) such as ISDN, or the like. When connected to a WAN such as the Internet 940 or ISDN, the gateway 941 is unnecessary.
The magnetic disk device 920 stores an operating system 921 (OS), a window system 922, a program group 923, and a file group 924. The programs in the program group 923 are executed by the CPU 911, the operating system 921, and the window system 922.

The program group 923 stores programs that execute functions described as “˜units” in the description of the embodiments described below. The program is read and executed by the CPU 911.
In the file group 924, information, data, signal values, variable values, and parameters described as “determination results”, “calculation results”, and “processing results” are stored in “files” and “˜databases”. "Is stored as each item. The “˜file” and “˜database” are stored in a recording medium such as a disk or a memory. Information, data, signal values, variable values, and parameters stored in a storage medium such as a disk or memory are read out to the main memory or cache memory by the CPU 911 via a read / write circuit, and extracted, searched, referenced, compared, and calculated. Used for CPU operations such as calculation, processing, output, printing, and display. Information, data, signal values, variable values, and parameters are temporarily stored in the main memory, cache memory, and buffer memory during the CPU operations of extraction, search, reference, comparison, operation, calculation, processing, output, printing, and display. Is remembered.
The arrows in the flowchart mainly indicate input / output of data and signals. The data and signal values are the RAM 914 memory, FDD 904 flexible disk, CDD 905 compact disk, magnetic disk device 920 magnetic disk, other optical disks, mini-disks, etc. It is recorded on a recording medium such as a disk or DVD (Digital Versatile Disk). Data and signals are transmitted online via a bus 912, signal lines, cables, or other transmission media.

  Further, what is described as “˜part” may be “˜circuit”, “˜device”, “˜device”, and “˜step”, “˜procedure”, “˜processing” May be. That is, what is described as “˜unit” may be realized by firmware stored in the ROM 913. Alternatively, it may be implemented only by software, or only by hardware such as elements, devices, substrates, and wirings, by a combination of software and hardware, or by a combination of firmware. Firmware and software are stored as programs in a recording medium such as a magnetic disk, a flexible disk, an optical disk, a compact disk, a mini disk, and a DVD. The program is read by the CPU 911 and executed by the CPU 911. That is, the program causes the computer to function as “to part” described below. Alternatively, the procedure or method of “to part” described below is executed by a computer.

FIG. 15 is a block configuration diagram showing an example of a functional block configuration of the performance prediction apparatus 100 according to this embodiment.
The performance prediction apparatus 100 includes a first performance acquisition unit 121, a second performance acquisition unit 122, an execution log acquisition unit 123, a first execution time acquisition unit 131, an execution time ratio acquisition unit 132, and a total execution time calculation. Unit 141, second execution time calculation unit 142, and total execution time prediction unit 150.

The first performance acquisition unit 121 uses the CPU 911 to acquire performance measurement data representing the performance of the computer system 820. The first performance acquisition unit 121 uses the magnetic disk device 920 to store the acquired performance measurement data.
The performance measurement data acquired by the first performance acquisition unit 121 is data generated by the computer system 820 executing the benchmark program 827. The performance measurement data represents the performance of the computer system 820 measured by the benchmark test. The performance measurement data includes, for example, the following data.

Processing device performance data The processing device performance data represents the performance of the processing device 811 of the computer system 820. The processing device performance data is data representing, for example, the time taken for the computer system 820 to execute a test program for performing a predetermined calculation.

File Access Performance Data The file access performance data represents the performance of access to data stored in the storage device 814 of the computer system 820 managed as a file by the operating system 825 and the middleware 826. The file access performance data is data representing the time taken for the computer system 820 to execute a test program for reading and writing a file, for example.

Event switching performance data The event switching performance data is data representing the performance of the event switching function provided by the operating system 825. The event switching performance data is data representing the time taken for the computer system 820 to execute a test program that frequently causes event switching, for example.

Exclusive control performance data Exclusive control performance data is data representing the performance of the exclusive control function provided by the operating system 825. The exclusive control performance data is, for example, data representing the time taken for the computer system 820 to execute a test program that frequently causes exclusive control.

Synchronous control performance data The synchronous control performance data is data representing the performance of the synchronous control function provided by the operating system 825. The synchronous control performance data is, for example, data representing the time taken for the computer system 820 to execute a test program that frequently causes synchronous control.

Graphic performance data The graphic performance data indicates the performance of the operating system 825 and middleware 826 displaying a predetermined graphic on the display device when there is a display device that displays an image as one of the output devices 813 of the computer system 820. It is data to represent. The graphic performance data is, for example, data representing the time taken for the computer system 820 to execute a test program for drawing a predetermined graphic.

  As described above, the performance measurement data includes data representing the results of measuring various performances of the computer system 820.

The second performance acquisition unit 122 uses the CPU 911 to acquire performance measurement data representing the performance of the computer system 810. The second performance acquisition unit 122 uses the magnetic disk device 920 to store the acquired performance measurement data.
The performance measurement data acquired by the second performance acquisition unit 122 is data generated when the computer system 810 executes the benchmark program 817. The contents of the performance measurement data of the computer system 810 acquired by the second performance acquisition unit 122 are the same as the performance measurement data of the computer system 820 acquired by the first performance acquisition unit 121, and thus detailed description thereof is omitted.

The first execution time acquisition unit 131 uses the CPU 911 to calculate the time required to execute a specific process in the computer system 820 based on the performance measurement data stored by the first performance acquisition unit 121. The first execution time acquisition unit 131 uses the magnetic disk device 920 to store data representing the calculated time. The data stored in the first execution time acquisition unit 131 is referred to as “first execution time data”.
The specific processing includes, for example, event switching processing, exclusive control processing, synchronization control processing, file access processing, graphic processing, and the like. These processes are realized by the computer system 820 executing programs such as the operating system 825 and the middleware 826. The time required for these processes cannot be simply calculated based on the performance of the processing device 811 or the like, and various factors such as implementation in the operating system 825 and hardware performance of the computer system 820 are involved.
Therefore, the first execution time acquisition unit 131 calculates the time required for these processes based on the actual measurement values. For example, the first execution time acquisition unit 131 uses the CPU 911 to calculate the time required for file access processing based on the file access performance data included in the performance measurement data. Similarly, the first execution time acquisition unit 131 uses the CPU 911 to perform event switching based on event switching performance data, exclusive control performance data, synchronous control performance data, graphic performance data, and the like included in the performance measurement data, respectively. Time required for processing, exclusive control processing, synchronous control processing, graphic processing, and the like is calculated.

Using the CPU 911, the execution time ratio acquisition unit 132 is generally used in the computer system 820 based on the performance measurement data stored by the first performance acquisition unit 121 and the performance measurement data stored by the second performance acquisition unit 122. The ratio of the time required to execute the same process to the time required to execute the same process in the computer system 810 is calculated. The execution time ratio acquisition unit 132 uses the magnetic disk device 920 to store data representing the calculated ratio. The data stored in the execution time ratio acquisition unit 132 is referred to as “execution time ratio data”.
The general process is, for example, a simple calculation process. These processes are realized by the computer systems 810 and 820 executing programs. The time required for these processes is mainly determined by the performance of the processing device 811 and the like. It can be considered that there is a generally proportional relationship between the time taken when the computer system 810 executes the same processing and the time taken when the computer system 820 executes.

The execution log acquisition unit 123 uses the CPU 911 to acquire an execution log of the application program 830 in the computer system 810. The execution log acquisition unit 123 uses the magnetic disk device 920 to store the acquired execution log.
The execution log acquired by the execution log acquisition unit 123 is generated by the computer system 810 executing the log generation program 818 while the computer system 810 is executing the application program 830.

  The total execution time calculation unit 141 uses the CPU 911 to calculate the time taken for the computer system 810 to execute the application program 830 based on the execution log stored by the execution log acquisition unit 123. The total execution time calculation unit 141 uses the magnetic disk device 920 to store data representing the calculated time. The data stored in the total execution time calculation unit 141 is referred to as “total execution time data”.

The second execution time calculation unit 142 uses the CPU 911 to call functions of the operating system 825 and middleware 826 while the computer system 810 is executing the application program 830 based on the execution log stored by the execution log acquisition unit 123. Thus, the time taken to execute the specific process is calculated. The second execution time calculation unit 142 uses the magnetic disk device 920 to store data representing the calculated time. The data stored in the second execution time calculation unit 142 is referred to as “second execution time data”.
The specific process in which the second execution time calculation unit 142 calculates the execution time is the same process as the specific process in which the first execution time acquisition unit 131 calculates the ten lake times.
For example, the second execution time calculation unit 142 uses the CPU 911 to search the execution log for a log indicating the start of a specific process, such as a system call or middleware 826 call, and start the specific process. Get the time. The second execution time calculation unit 142 uses the CPU 911 to search the execution log for a log indicating the end of the specific process such as the called system call or the end of the process of the middleware 826, Get the end time of the process. Using the CPU 911, the second execution time calculation unit 142 calculates the elapsed time from the acquired start time to the end time, and obtains the time taken to execute the specific process.
The second execution time calculation unit 142 uses the CPU 911 to calculate each execution time for all specific processes executed by the computer system 810 during the execution of the application program 830. When the same specific process is executed a plurality of times, the second execution time calculation unit 142 calculates each execution time for each execution.

  The total execution time prediction unit 150 uses the CPU 911 to predict the time required for the computer system 820 to execute the application program 830. The total execution time prediction unit 150 uses the CPU 911 to store the first execution time data stored in the first execution time acquisition unit 131, the execution time ratio data stored in the execution time ratio acquisition unit 132, and the total execution time calculation unit. The estimated time is calculated based on the total execution time data stored in 141 and the second execution time data stored in the second execution time calculation unit 142. The total execution time prediction unit 150 uses the magnetic disk device 920 to store data representing the calculated prediction time. The data stored in the total execution time prediction unit 150 is referred to as “total execution time prediction data”. Using the CPU 911, the total execution time prediction unit 150 outputs the stored total execution time prediction data.

For example, the total execution time prediction unit 150 uses the CPU 911 to total the execution times represented by the second execution time data, and totals the execution times of specific processes executed by the computer system 810 during the execution of the application program 830. Calculate the total time. The total execution time prediction unit 150 uses the CPU 911 to calculate a difference obtained by subtracting the total execution time of the calculated specific process execution time from the execution time represented by the total execution time data, and the computer system 810 executes the application program. A total time of execution times of general processes other than the specific process executed during the execution of 830 is calculated. The total execution time prediction unit 150 uses the CPU 911 to calculate a product obtained by multiplying the total time of the calculated execution time of the general processing by the ratio represented by the execution time ratio data, and the computer system 820 makes the application program The total execution time of the execution time of a general process executed during the execution of 830 is predicted.
Using the CPU 911, the total execution time prediction unit 150 replaces the execution time of each specific process represented by the second execution time data with the execution time of each of the same specific processes represented by the first execution time data. The total execution time prediction unit 150 uses the CPU 911 to add up the replaced execution times, and predicts the total execution time of the specific processing executed by the computer system 820 during the execution of the application program 830.
The total execution time prediction unit 150 uses the CPU 911 to add up the total time of the predicted execution time of the general process and the total time of the predicted execution time of the specific process. The time taken to execute the application program 830 is predicted.

FIG. 16 is a flowchart showing an example of the flow of the performance prediction process S610 in this embodiment.
In the performance prediction process S610, the performance prediction apparatus 100 predicts the time required for the computer system 820 to execute the application program 830. The performance prediction process S610 includes a first performance acquisition process S611, a second performance acquisition process S612, a specific process selection process S613, a first execution time calculation process S614, a general first execution time calculation process S615, and a general first. Two execution time calculation step S616, execution time ratio calculation step S617, execution log acquisition step S621, initialization step S622, log selection step S623, process type branching step S624, elapsed time calculation step S625, Two execution time addition step S626, predicted execution time addition step S627, process type determination step S628, total execution time calculation step S631, general execution time calculation step S632, general execution time prediction step S633, and prediction execution time Calculation step S634. The performance prediction apparatus 100 starts processing from the first performance acquisition step S611.

In the first performance acquisition step S611, the first performance acquisition unit 121 uses the CPU 911 to acquire performance measurement data for the computer system 810.
In the second performance acquisition step S612, the second performance acquisition unit 122 uses the CPU 911 to acquire performance measurement data for the computer system 820.

In the specific process selection step S613, the first execution time acquisition unit 131 uses the CPU 911 to select a specific process for which the execution time has not yet been calculated from among the specific processes for which the execution time is to be calculated.
When the execution time has been calculated for all the specific processes and there is no specific process for which the execution time has not yet been calculated, the first execution time acquisition unit 131 uses the CPU 911 to calculate the general first execution time. The process proceeds to step S615.
When there is a specific process for which the execution time has not yet been calculated, the first execution time acquisition unit 131 uses the CPU 911 to select one specific process from among the specific processes for which the execution time has not yet been calculated. Select one. Using the CPU 911, the first execution time acquisition unit 131 advances the process to the first execution time calculation step S614.

In the first execution time calculation step S614, the first execution time acquisition unit 131 uses the CPU 911 to specify the specific process selection step based on the performance measurement data acquired by the first performance acquisition unit 121 in the first performance acquisition step S611. For the specific process selected in S613, the time required for the computer system 820 to execute the specific process is calculated.
Using the CPU 911, the first execution time acquisition unit 131 returns the process to the specific process selection step S613 and selects the next specific process.

In the general first execution time calculation step S615, the execution time ratio acquisition unit 132 uses the CPU 911, based on the performance measurement data acquired by the first performance acquisition unit 121 in the first performance acquisition step S611. The time required for general processing is calculated.
In the general second execution time calculation step S616, the execution time ratio acquisition unit 132 uses the CPU 911 to calculate the computer system 810 based on the performance measurement data acquired by the second performance acquisition unit 122 in the second performance acquisition step S612. The time required for general processing is calculated.
In the execution time ratio calculation step S617, the execution time ratio acquisition unit 132 uses the CPU 911 to divide the time calculated in the general first execution time calculation step S615 by the time calculated in the general second execution time calculation step S616. The quotient is calculated and the execution time ratio is calculated.

  In the execution log acquisition step S621, the execution log acquisition unit 123 uses the CPU 911 to acquire an execution log when the computer system 810 executes the application program 830.

In the initialization step S622, the total execution time calculation unit 141 uses the CPU 911 to select the execution log acquired by the execution log acquisition unit 123 in the execution log acquisition step S621 line by line in order from the beginning, and then executes the application program 830. Look for a log indicating the start of execution. Using the CPU 911, the total execution time calculation unit 141 acquires the execution start time of the application program 830 from the log indicating the start of execution of the application program 830.
Using the RAM 914, the second execution time calculation unit 142 stores 0 as the total time of the execution time of the specific process executed by the computer system 810. The second execution time calculation unit 142 uses the RAM 914 to store data meaning “general processing” as the processing type.
The total execution time prediction unit 150 uses the RAM 914 to store 0 as the total time of the execution time of the specific process executed by the computer system 820.

  In the log selection step S623, the second execution time calculation unit 142 uses the CPU 911 to select the next one-line log from the execution log acquired by the execution log acquisition unit 123 in the execution log acquisition step S621.

In the process type branching step S624, the second execution time calculating unit 142 uses the CPU 911 to branch the process depending on whether the stored process type is “general process” or “specific process”.
When the process type is “general process”, the second execution time calculation unit 142 uses the CPU 911 to advance the process to the process type determination step S628.
When the process type is “specific process”, the second execution time calculation unit 142 uses the CPU 911 to advance the process to the elapsed time calculation step S625.

In the elapsed time calculation step S625, the second execution time calculation unit 142 uses the CPU 911 to acquire the time when the log is recorded from the log acquired in the log selection step S623. The second execution time calculation unit 142 uses the CPU 911 to calculate the elapsed time from the stored process start time to the acquired time.
In the second execution time adding step S626, the second execution time calculating unit 142 uses the CPU 911 to calculate the elapsed time calculated in the elapsed time calculating step S625 to the total execution time of the specific processing of the computer system 810 stored. Add time. Using the RAM 914, the second execution time calculation unit 142 stores the total time with the elapsed time added.
In the predicted execution time adding step S627, the total execution time predicting unit 150 uses the CPU 911, based on the type of processing stored in the second execution time calculating unit 142, in the first execution time calculating step S614. The execution time for the same specific process is acquired from the execution times calculated by the acquisition unit 131. Using the CPU 911, the total execution time prediction unit 150 adds the acquired execution time to the total execution time of the specific processing execution time of the stored computer system 820. Using the RAM 914, the total execution time prediction unit 150 stores the total time with the execution time added.

In the processing type determination step S628, the second execution time calculation unit 142 uses the CPU 911 to determine the type of processing started at the time when the log was recorded based on the log selected in the log selection step S623. .
When the log selected in the log selection step S623 is a log representing the end of execution of the application program 830, the second execution time calculation unit 142 uses the CPU 911 to advance the process to the total execution time calculation step S631.
When the log selected in the log selection step S623 is a log representing the start of a general process or a specific process, the second execution time calculation unit 142 uses the RAM 914 to store the determined process type. Using the CPU 911, the second execution time calculation unit 142 returns the process to the log selection step S623 and selects the next log.

In the total execution time calculation step S631, the total execution time prediction unit 150 uses the CPU 911 to acquire the execution end time of the application program 830 from the log selected by the second execution time calculation unit 142 in the log selection step S623. Using the CPU 911, the total execution time prediction unit 150 calculates the total execution time of the application program 830 based on the execution start time stored in the initialization step S622 and the acquired execution end time.
In the general execution time calculating step S632, the total execution time predicting unit 150 uses the CPU 911 to execute a specific process stored in the second execution time calculating unit 142 from the total execution time calculated in the total execution time calculating step S631. The total time of the general processing of the computer system 810 is calculated by subtracting the total time of the time.
In the general execution time prediction step S633, the total execution time prediction unit 150 uses the CPU 911 to set the execution time ratio to the total execution time of the general processing of the computer system 810 calculated in the general execution time calculation step S632. By multiplying the execution time ratio calculated in the calculation step S617, the total time of the execution time of the general processing of the computer system 820 is calculated.
In the predicted execution time calculation step S634, the total execution time prediction unit 150 uses the CPU 911 to calculate the total execution time of the specific processing execution time of the stored computer system 820 and the computer system calculated in the general execution time prediction step S633. The total time of the execution time of the general process 820 is added to calculate the time required for the computer system 820 to execute the application program 830.

  As described above, the configuration described in each embodiment is an example, and the configuration described in different embodiments may be combined. Moreover, it is good also as a structure which changed, such as replacing the structure of the non-major part with other structures, such as the existing technique.

The performance prediction device (100) described above includes a processing device (CPU 911) for processing data, a first execution time acquisition unit (131; new system performance estimation unit 118), and an execution log acquisition unit (123; log analysis unit). 114), a total execution time calculation unit (141; 114), a second execution time calculation unit (142; 114), and a total execution time prediction unit (150; 118).
The first execution time acquisition unit (131; 118) uses the processing device (911) to implement a first specific program (operating system 825, which realizes a specific function in the first computer system (820; 108). 111; middleware 826), the time taken to execute the first specific program is acquired.
The execution log acquisition unit (123; 114) is an application program that can be executed in the first computer system (820; 108) and the second computer system (810; 101) using the processing device (911). For (830; application 102), an execution log (application log 107) for executing the application program (830; 102) in the second computer system (810; 101) is acquired.
The total execution time calculation unit (141; 114) analyzes the execution log (107) acquired by the execution log acquisition unit (123; 114) using the processing device (911), and executes the application program. Calculate the time taken to execute.
The second execution time calculation unit (142; 114) uses the processing device (911) to analyze the execution log (107) acquired by the execution log acquisition unit (123; 114), and In the second computer system (810; 101), the second specific program (operating system 815) that realizes the specific function realized by the first specific program (825, 111; 826) in the computer system (820; 108) , 105; 816, 103) when the application program (830; 102) calls the second specific program (815, 105; 816, 103), the second specific program (815, 105; 816, 103). ) Is calculated.
The total execution time prediction unit (150; 118) uses the processing device (911) to calculate the second execution time calculation unit (142) out of the times calculated by the total execution time calculation unit (141; 114). 114), the application program (830) is executed in the first computer system (820; 108) by replacing the time calculated by the first execution time acquisition unit (131; 118). Predict how long it will take.

  As a result, even when the application program cannot be executed by the first computer system at this time, the time required to execute the application program by the first computer system can be accurately predicted.

The performance prediction apparatus (100) includes an execution time ratio acquisition unit (132; 118).
The execution time ratio acquisition unit (132; 118) can be executed in the first computer system (820; 108) and the second computer system (810; 101) using the processing device (911). For the general program (benchmark program 817, 827), the first computer system (820; 108) with respect to the time taken to execute the general program (817, 827) in the second computer system (810; 101). ) To obtain the ratio of the time taken to execute the general program (817, 827).
The total execution time prediction unit (150; 118) uses the processing device (911) to calculate the second execution time calculation unit (142) out of the times calculated by the total execution time calculation unit (141; 114). 114), the application program is executed in the first computer system (820; 108) by multiplying the time excluding the time calculated by 114) by the ratio acquired by the execution time ratio acquisition unit (132; 118). Predict how long it will take.

  Thereby, it is possible to more accurately predict the time required to execute the application program in the first computer system.

The first specific program is a system call of the operating system (825; 111) that operates in the first computer system (820; 108).
The second specific program is a system call of the operating system (815; 105) operating in the second computer system (810; 101).

  Since the time required for processing a system call is determined by various factors, it cannot be converted with a simple proportional relationship. Since the time taken to process the system call is replaced with the actually measured time, the time taken to execute the application program in the first computer system can be accurately predicted.

  The specific function realized by the first specific program and the second specific program is at least one of event switching processing, exclusive control processing, synchronization control processing, file access processing, and graphic processing.

  Since the time required for these processes is also determined by various factors, it cannot be converted with a simple proportional relationship. Since the time required for these processes is replaced with the actually measured time, the time required for executing the application program in the first computer system can be accurately predicted.

The general program is a benchmark program (817, 827).
The execution time ratio acquisition unit (132; 118) obtains the result of executing the benchmark program (817, 827) in the first computer system (820; 108) and the second computer system (810; 101). Based on the above, the ratio is calculated using the processing device (911).

  Based on the results measured by the benchmark program, the ratio of the execution speed of the first computer system to the second computer system is calculated, so the time required to execute the application program on the first computer system can be accurately predicted. it can.

  The performance prediction apparatus (100) described above can be realized by a computer program that causes a computer to function as a performance prediction apparatus when executed by the computer.

  100 performance prediction apparatus, 101, 108 computer system, 102 application, 103 middleware, 104 log acquisition unit, 105, 111 operating system, 106, 112 hardware, 107 application log, 109 basic performance acquisition unit, 113 basic performance result, 114 Log analysis unit, 115, 205 Log analysis result, 116 Actual value use unit, 117 Performance ratio use unit, 118 New system performance estimation unit, 119, 207 New system performance estimation result, 121 First performance acquisition unit, 122 Second performance Acquisition unit, 123 execution log acquisition unit, 131 first execution time acquisition unit, 132 execution time ratio acquisition unit, 141 total execution time calculation unit, 142 second execution time calculation unit, 150 total execution time prediction unit, 301 ID, 302 Function name, 303 arguments 304 time stamp, 800 performance prediction system, 810,820 computer system, 811 processing device, 812 input device, 813 output device, 814 storage device, 815,825 operating system, 816,826 middleware, 817,827 benchmark program, 818 Log generation program, 830 application program, 901 display device, 902 keyboard, 903 mouse, 904 FDD, 905 CDD, 906 printer device, 907 scanner device, 910 system unit, 911 CPU, 912 bus, 913 ROM, 914 RAM, 915 communication Device, 920 magnetic disk device, 921 OS, 922 window system, 923 program group, 924 file group, 931 Telephone, 932 facsimile machine, 940 Internet, 941 gateway, 942 LAN.

Claims (7)

A processing device that processes data, a first execution time acquisition unit, an execution log acquisition unit, a total execution time calculation unit, a second execution time calculation unit, and a total execution time prediction unit;
The first execution time acquisition unit acquires the time required to execute the first specific program for the first specific program that realizes a specific function in the first computer system, using the processing device,
The execution log acquisition unit uses the processing device to obtain an execution log for executing the application program in the second computer system for an application program executable in the first computer system and the second computer system. Acquired,
The total execution time calculation unit uses the processing device to analyze the execution log acquired by the execution log acquisition unit and calculate the time taken to execute the application program,
The second execution time calculation unit analyzes the execution log acquired by the execution log acquisition unit using the processing device, and performs the specific function realized by the first specific program in the first computer system. For the second specific program realized in the second computer system, when the application program calls the second specific program, it calculates the time taken to execute the second specific program,
The first execution time acquisition unit acquires the time calculated by the second execution time calculation unit among the times calculated by the total execution time calculation unit, using the processing device. A performance prediction apparatus for predicting a time required to execute the application program in the first computer system by replacing with time. The performance prediction apparatus has an execution time ratio acquisition unit,
The execution time ratio acquisition unit executes the general program in the second computer system for the general program executable in the first computer system and the second computer system using the processing device. The ratio of the time taken to execute the general program in the first computer system to the time taken to
The total execution time prediction unit uses the processing device to calculate the execution time ratio acquisition unit at a time excluding the time calculated by the second execution time calculation unit from the time calculated by the total execution time calculation unit. 2. The performance prediction apparatus according to claim 1, wherein the time required to execute the application program in the first computer system is predicted by multiplying the acquired ratio. The first specific program is a system call of an operating system that operates in the first computer system.
3. The performance prediction apparatus according to claim 1, wherein the second specific program is an operating system call that operates in the second computer system.   The specific function realized by the first specific program and the second specific program is at least one of an event switching process, an exclusive control process, a synchronization control process, a file access process, and a graphic process. The performance prediction apparatus according to any one of claims 1 to 3, wherein The above general program is a benchmark program,
The execution time ratio acquisition unit calculates the ratio using the processing device based on a result of executing the benchmark program in the first computer system and the second computer system. The performance prediction apparatus according to any one of claims 1 to 4.   A computer program, wherein when executed by a computer having a processing device for processing data, the computer functions as the performance prediction device according to any one of claims 1 to 5. A performance prediction device having a processing device that processes data, a first execution time acquisition unit, an execution log acquisition unit, a total execution time calculation unit, a second execution time calculation unit, and a total execution time prediction unit, In a performance prediction method for predicting the performance of an application program,
The first execution time acquisition unit acquires the time taken to execute the first specific program for the first specific program that realizes a specific function in the first computer system, using the processing device,
For the application program executable by the first computer system and the second computer system using the processing device, the execution log acquisition unit executes an execution log obtained by executing the application program in the second computer system. Acquired,
The total execution time calculation unit uses the processing device to analyze the execution log acquired by the execution log acquisition unit and calculate the time taken to execute the application program,
The second execution time calculation unit analyzes the execution log acquired by the execution log acquisition unit using the processing device, and performs a specific function realized by the first specific program in the first computer system. For the second specific program realized in the second computer system, when the application program calls the second specific program, it calculates the time taken to execute the second specific program,
The first execution time acquisition unit acquires the time calculated by the second execution time calculation unit among the times calculated by the total execution time calculation unit using the processing device. A performance prediction method characterized by predicting the time required to execute the application program in the first computer system by replacing with time.

Images