JP4843379B2 - Computer system development program - Google Patents

Description

The present invention relates to a Menopu program which to develop a system using a computer, in particular, on the development program of the computer system comprising a plurality of program modules.

  When developing a computer system such as an Internet system or an intranet system, the setting of the performance target value for each program module (hereinafter sometimes simply referred to as “module”) at the design stage is performed as follows, for example. . That is, when the performance of a computer system (hereinafter sometimes simply referred to as “system”) is predicted from the performance set for each module, and the predicted result does not meet the system target, the performance of each module Review target values. Here, depending on the experience and intuition of the designer, which module is reconsidered is selected which is easy to improve or which has come up with an improvement method. Therefore, there are many cases where efficient improvement work that leads to improvement of the performance of the entire system is not performed.

  Moreover, the same thing occurs when the performance of each module is predicted by using the actual performance of each module obtained by measuring each module in a unit test. That is, the system performance is predicted from the actual performance of each module, and when the prediction result does not satisfy the system target, each module is improved. Here, which modules are improved depends on the experience and intuition of the developer, and those that are easy to improve and those that come up with an improvement method are selected. Therefore, there are many cases where efficient improvement work that leads to improvement of the performance of the entire system is not performed.

  As described above, in the conventional technology, when developing a computer system such as an Internet system or an intranet system, if the performance prediction result as a system does not meet the target, which module performance target value should be reviewed at design time, It is up to the developer's experience and intuition to decide which module performance to improve after development. For this reason, there are many cases where efficient system performance improvement cannot be expected.

  In the “Computer System Performance Prediction Program and Design Support System” disclosed in Patent Document 1 below, a scenario composed of a series of requests and their timings, and a request and a program module (part block) activated by the request. A method for predicting the response time and resource utilization rate from the number of scenario occurrences during system operation based on the relationship, processing time of the program module (component block) and resource usage is shown.

However, this disclosure does not indicate which program module (component block) should be improved when the system performance does not satisfy the target value as a result of system performance prediction.
JP 2004-272582 A

The present invention, in an open Hatsupu program of the computer system, by selecting for or determining a program module that greatly affects the system performance, which can contribute to efficient performance improvements should develop systems, computer systems an object of the present invention is to provide a development program.

In order to solve the above-described problems, a computer system development program according to the present invention includes first information relating to the amount of each business carried by the computer system to be developed, each business and each process executed in the computer system. Second information relating to the relationship between the processing, third information relating to the association between each process and each module constituting the computer system, and fourth relating to the processing time for each module involved in each processing. A step of inputting information; a step of predicting performance of the computer system using the first to fourth information to obtain a first prediction result; and a predetermined using the first prediction result of the computer system In the combination of the step of obtaining the first evaluation function value by performing the evaluation using the evaluation function of the first processing function and the one of the modules, Predicting the performance of the computer system for a plurality of different combinations by using the first to fourth information under conditions in which the information is minutely changed, and obtaining second prediction results, respectively, Performing a second evaluation function value by performing an evaluation with the predetermined evaluation function using the second prediction result, the first evaluation function value, the second evaluation function value, and the first Determining the sensitivity to the first evaluation function value in the plurality of combinations of the one of the processes and the one of the modules from the condition in which the information of 4 is slightly changed, and the sensitivity Rearranging the plurality of combinations of the one of the processes and the one of the modules in the order of the sizes of the processes. That.

That is, in the development program of this, to predict system performance, calculated evaluation function value by a predetermined evaluation function using the prediction result, further, the sensitivity of the evaluation function value one one and the modules of the processing Is obtained for a plurality of combinations. Therefore, the user can know the modules that greatly affect the performance of the system. As a result, the target performance of the module is corrected, or the module is actually improved, thereby contributing to efficient system performance improvement.

According to the present invention, in the development program of the computer system, developing a program module which greatly affects the system performance selection assistance with or determined, which can contribute to efficient performance improvements to be developed System program is provided.

As an embodiment of the present invention, using the step of inputting, as fifth information, information relating to the degree of improvement for each module related to each process, the sensitivity and the fifth information, Obtaining a total value for each of the plurality of combinations of the one and the one of the modules; and the plurality of the one of the processes and the one of the modules in order of the magnitude of the total value. in the step of rearranging the combination is performed further, it may be. According to this, the user can know the module that greatly affects the performance of the system in consideration of the ease of improvement.

Further, as an embodiment, information on the degree of improvement for each module related to each process is set as fifth information, and information about the shortenable time for each module related to each process is set as sixth information. Inputting each of the plurality of combinations of the one of the processes and the one of the modules using the sensitivity and the fifth and sixth information; and in the order of magnitude of the total value, the to be executed by the plurality of further the step of rearranging the combination of said one of said each module and the one of the processes may be. According to this, the user can know the module that greatly affects the performance of the system in consideration of the ease of improvement and the shortenable time.

Here, the computer system performance is predicted by reflecting, in the fourth information, the shortenable time corresponding to the combination of each of the corresponding processing and each of the modules from the one having the large total value. Each of obtaining a prediction result, obtaining a third evaluation function value by performing an evaluation with the predetermined evaluation function using the third prediction result of the computer system, and the third evaluation function value and determining whether or not reached a predetermined value is performed further, it may be. According to this, it is possible to automatically present to what extent the shortenable time is actually applied until the target can be achieved as a system to be developed.

Further, as an embodiment, information on the shortenable time for each module related to each process is set as fifth information for each of the plurality of improvement measures, and information on the degree of improvement of each of the plurality of improvement measures is set as sixth information. The step of inputting each as information, the step of obtaining a comprehensive evaluation value for each of the improvement measures using the sensitivity and the fifth and sixth information, and the improvement measures in order of the magnitude of the comprehensive evaluation value to perform the further the steps of sort may be. According to this, by presenting the improvement measures to the user in descending order of the comprehensive evaluation value, the user can know the improvement measures that greatly affect the performance of the system.

Here, the performance of the computer system is predicted by reflecting the shortenable time corresponding to the corresponding improvement measure on the fourth information, and the third prediction result is obtained from the one having the large comprehensive evaluation value. A combination of one of the processes and one of the modules, and a step of performing evaluation using the predetermined evaluation function using the third prediction result of the computer system to obtain a third evaluation function value And predicting the performance of the computer system for a plurality of different combinations by using the first to fourth information according to the condition in which the fourth information is slightly changed in step 4, and obtaining a fourth prediction result, respectively. Performing the evaluation with the predetermined evaluation function using the fourth prediction result of the computer system to obtain a fourth evaluation function value, The evaluation function value, the fourth evaluation function value, and the condition in which the fourth information is slightly changed, in the plurality of combinations of the one of the processes and the one of the modules. A second comprehensive evaluation value for each of the improvement measures using the step of obtaining the sensitivity to the third evaluation function value as the second sensitivity, and the second sensitivity and the fifth and sixth information. and determining a second overall function values to perform the further the steps of determining whether it has reached a predetermined value may be. According to this, each improvement measure is displayed to the user in descending order of the comprehensive evaluation value, and further, up to which improvement measure is applied cumulatively, the goal is achieved.

  Further, as an embodiment, the predetermined evaluation function is a sum total of all the resources of the transition of the resource utilization rate exceeding the upper limit value using the upper limit value of the resource utilization rate used by the computer system as an index. It can be.

  Further, as an embodiment, the predetermined evaluation function is a sum of all the resources having a maximum value of a resource utilization rate exceeding the upper limit value, using the upper limit value of the resource utilization rate used by the computer system as an index. , And can be.

  Further, as an embodiment, the predetermined evaluation function is a sum of all the resources in a time when the resource utilization rate exceeds the upper limit value, using the upper limit value of the resource utilization rate used by the computer system as an index. , And can be.

  Further, as an embodiment, the predetermined evaluation function is a sum of all the resources of the transition of the response time exceeding the upper limit value using the upper limit value of the response time of the resource used by the computer system as an index. can do.

  Further, as an embodiment, the predetermined evaluation function is a sum of all the resources of the maximum value of the response time exceeding the upper limit value, using the upper limit value of the response time of the resource used by the computer system as an index. It can be.

  Further, as an embodiment, the predetermined evaluation function is a sum of all the resources for a time that becomes a response time exceeding the upper limit value, using the upper limit value of the response time of the resource used by the computer system as an index. It can be.

  Furthermore, as an embodiment, the predetermined evaluation function may be a combination of any two or more of the evaluation functions described above.

(First embodiment)
Based on the above, embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing the configuration of a system (tool) for executing a computer system development method according to an embodiment of the present invention. As illustrated in FIG. 1, the system (tool) includes an input unit 101, a sensitivity analysis unit 102, a performance prediction unit 103, and a display unit 104. With these configurations, the system functions as a tool that performs sensitivity calculations and provides the user with support for determining modules that have a greater impact on the performance of the system to be developed.

  The input unit 11 has a function of acquiring work amount information, each related information, and module performance information from the outside and passing them to the sensitivity analysis unit 12. More specifically, for example, an input device such as a keyboard and a mouse and its control program, a drive device for various recording media such as a hard disk and its control program, an interface device for various networks and its control program, etc. can do.

  The business refers to a work classified by work to be developed by the system to be developed. The business amount information is, for example, information indicating the amount of each business in the time direction. Each related information is executed in each business and system to be developed (hereinafter, “system to be developed” is also simply referred to as “system”, but is different from what is described as “system (tool)”). Information relating to the relationship between each process p, information relating to the relationship between each process p and each module m constituting the system (for example, information relating to / without relation), and each resource being the hardware of each module m and the system It is information about the relationship. The module performance information is information on the processing time required for the module m related when the process p is executed or information on the use time of resources used by the module m related when the process p is executed.

  The sensitivity analysis unit 12 evaluates based on a predetermined evaluation function based on the business amount information passed from the input unit 11, each related information, module performance information, and the performance prediction result of the system passed from the performance prediction unit 12. The function value E is obtained. Further, the sensitivity to the evaluation function value E is obtained using the process p and the module m as parameters, and the result is sorted in the order of the combination of m and p having high sensitivity. The performance prediction unit 13 obtains the workload information, each related information, and the module performance information from the sensitivity analysis unit 12, thereby predicting the system performance. The prediction result of the system performance is passed to the sensitivity analysis unit 12. The display unit 14 displays the sensitivity analysis result obtained by the sensitivity analysis unit 12. At that time, the display is performed in the order of the combination of m and p with high sensitivity in the sensitivity analysis result.

  More specifically, the sensitivity analysis unit 12 and the performance prediction unit 13 are configured by, for example, hardware such as a microprocessor or a memory and software such as an OS or an application program that functions on the hardware. Can do. The display unit 14 can be composed of, for example, a display device such as a liquid crystal display device and its control program.

  The operation of the system (tool) shown in FIG. 1 will be described below with reference to the flowcharts shown in FIGS. 2 and 3 are flowcharts showing an operation flow of the system (tool) shown in FIG. First, the business unit information, each related information, and module performance information are input by the input unit 11 (step 201). Each information acquired by the input is transferred to the sensitivity analysis unit 12 and further sent to the performance prediction unit 13 from there. From step 202 to step 218, loop processing (loop 1) is performed.

  The performance prediction unit 13 performs simulation using each piece of information passed to predict system performance (step 203). The prediction result is sent to the sensitivity analysis unit 12. The sensitivity analysis unit 12 obtains an evaluation function value E by a predetermined evaluation function using the sent prediction result (step 204). Then, it is determined whether or not the evaluation function value E has reached a predetermined target (step 205). If the evaluation function value E has been reached, the process exits loop 1 and displays the prediction result of the system performance (step 219).

  When the evaluation function value E does not reach the target (N in Step 205), the process proceeds to sensitivity calculation processing. That is, process p (p is a process number assigned to each process) in loop 2 (step 206 to step 214) is a loop variable, and module m (m is assigned to each module) in loop 3 (step 207 to step 231). The module number) is processed as a loop variable, and the sensitivity to the evaluation function value E is calculated by a combination of p and m.

More specifically, the processing time of the module m related to the processing p is increased by a minute amount Δx (step 208: in the sensitivity analysis unit 12), and simulation is performed again under that condition to predict the system performance (step 209: Performance prediction unit 13). The prediction result is sent to the sensitivity analysis unit 12. The sensitivity analysis unit 12 calculates an evaluation function value from this prediction result in the same manner as in step 204 (step 210). Let this evaluation function value be ^{Em, p} . Further, the sensitivity analysis unit 12 obtains the sensitivity S ^{m, p} as (E ^{m, p} −E) / Δx (step 211). When the sensitivity S ^{m, p} is obtained, the minute amount Δx is returned to the original (step 212), and the loops 2 and 3 are repeated in the same manner to obtain the sensitivity for all combinations of m and p.

When the sensitivities S ^{m, p} are obtained for all combinations of m and p, they are sorted in the order of their magnitudes (step 215: in the sensitivity analyzer 12). Then, display is made on the display unit 14 in the order of the sensitivity S ^{m, p} (step 216). Therefore, the user can reasonably determine which review of the performance of which module m by viewing this display greatly contributes to the performance improvement of the system.

  When the user's judgment is made, the performance information for the module is corrected and input in accordance with the determination (step 217: in the input unit 11). Thereafter, the process returns to step 202 and the processing in loop 1 is repeated. The correction input of the module performance information is a review of the target performance for each module at the design stage, but it may be an input of the module performance after the actual improvement after the unit test of the module. When the loop 1 is exited, the prediction result of the system obtained in the last step 203 of the loop 1 process is displayed (step 219).

As described above, according to the present embodiment, by displaying the combination of the module m and the processing p in descending order of the sensitivity S ^{m, p} , the user knows the module m that greatly affects the performance of the system. it can. As a result, the target performance of the module m can be corrected, or the module m can be actually improved to contribute to efficient system performance improvement.

(Second Embodiment)
Next, a computer system development method according to another embodiment of the present invention will be described with reference to FIG. 4, FIG. 5, and FIG. FIG. 4 is a block diagram showing a configuration of a system (tool) for executing a computer system development method according to another embodiment of the present invention. 5 and 6 are flowcharts showing an operation flow of the system (tool) shown in FIG. 4, 5, and 6, the same reference numerals are given to the same components or steps as those shown in the previous drawings, and the description thereof is omitted.

  This embodiment introduces an improvement degree that expresses the ease of improvement of a module as a numerical value, and presents an evaluation result to the user by combining the sensitivity and the improvement degree.

  As shown in FIG. 4, an improvement ease input unit 15 is added to this system (tool). The improvement ease input unit 15 has a function of acquiring the improvement ease information from the outside and passing it to the sensitivity analysis unit 12. The ease of improvement is a numerical value of the degree of ease in improving the combination of the module m and the processing p, and is determined in consideration of time required for improvement, cost, technical ability, and the like. For example, it is represented by a value of 0 or more and 1 or less (the closer to 1, the easier). More specifically, like the input unit 11, the improved ease input unit 15 includes an input device such as a keyboard and a mouse and its control program, a drive device for various recording media such as a hard disk and its control program, It can be composed of a network interface device and its control program.

  The sensitivity analysis unit 12 calculates the sensitivity in the same manner as in the embodiment shown in FIG. 1, and performs this in the order of the product of the sensitivity and the improvement ease when sorting. Hereinafter, the operation of the system (tool) shown in FIG. 4 will be described more specifically with reference to the flowcharts shown in FIGS.

Step 201 is the same as that in FIG. 2 (first embodiment). Next, the improvement ease information em ^{, p} is input by the improvement degree input unit 15 (step 401). Step 203 to step 204 are the same as those in FIG. Further, Step 205, Step 219, and Steps 206 to 214 are the same as those in FIG.

Then, improved easiness ^{e m} being ^input, the sensitivity was obtained with the ^{p S m,} total value ^{e m} a product of ^p, p · ^{S m,} sorted in order of magnitude of ^p Toshiko (step 415: Sensitivity analyzer 12). Then, the display unit 14 displays the total values em ^{, p} · S ^{m, p} in this order (step 416). Therefore, the user can rationally determine which review of the performance of the module m by looking at this display greatly contributes to the performance improvement of the system, taking into account the ease of improvement. Step 217 after the user's decision is made is the same as in FIG.

As described above, according to the present embodiment, the combination of the module m and the processing p is displayed to the user in descending order of the total values em ^{, p} · S ^{m, p} , so that the user greatly affects the performance of the system. The module m can be known in consideration of the ease of improvement. As a result, the target performance of the module m can be corrected, or the module m can be actually improved to contribute to efficient system performance improvement.

(Third embodiment)
Next, a computer system development method according to still another embodiment of the present invention will be described with reference to FIG. 7, FIG. 8, and FIG. FIG. 7 is a block diagram showing a configuration of a system (tool) for executing a computer system development method according to still another embodiment of the present invention. 8 and 9 are flowcharts showing an operation flow of the system (tool) shown in FIG. 7, 8, and 9, the same reference numerals are given to the same components or steps as those shown in the previous drawings, and the description thereof is omitted.

  In this embodiment, in addition to the above-described embodiment, it is possible to input information on the shortenable time for each process of the module, and the evaluation result combining the sensitivity, the ease of improvement, and the shortenable time can be obtained. It is presented to the user.

  As shown in FIG. 7, a shortenable time input unit 16 is added to this system (tool). The shortenable time input unit 16 has a function of acquiring information on the shortenable time from the outside and passing the information to the sensitivity analysis unit 12. The shortenable time information is the amount that can be shortened in the processing time in the module m and the process p. More specifically, the shortenable time input unit 16 is similar to the input unit 11 and the improved ease input unit 15, and includes input devices such as a keyboard and a mouse, a control program thereof, and drive devices for various recording media such as a hard disk. It can be configured by the control program, various network interface devices and the control program.

  The sensitivity analysis unit 12 calculates the sensitivity in the same manner as the embodiment shown in FIG. 1 and performs this in the order of the product of the sensitivity, the degree of ease of improvement, and the shortenable time when sorting. Hereinafter, the operation of the system (tool) shown in FIG. 7 will be described more specifically with reference to the flowcharts shown in FIGS.

Step 201 is the same as FIG. 2 (first embodiment), and step 401 is the same as FIG. 5 (second embodiment). Next, the shortenable time dx ^{m, p} is input by the shortenable time input unit 16 (step 601). Step 203 to step 204 are the same as those in FIG. Further, Step 205, Step 219, and Steps 206 to 214 are the same as those in FIG.

Then, improved easiness ^{e m} being ^{input, p} and shorten available time ^{dx m, p} and the resulting sensitivity ^{S m,} overall the product of ^p values ^{dx m, p · e m,} p · S m ^{, P} and sort in the order of this size (step 615: sensitivity analysis unit 12). Then, display is performed on the display unit 14 in the order of the total value dx ^{m, p} · em ^{, p} · S ^{m, p} (step 616). Therefore, the user can rationally determine which review of the performance of the module m by looking at this display greatly contributes to the performance improvement of the system, taking into consideration the ease of improvement and the shortenable time. it can. Step 217 after the user's decision is made is the same as in FIG.

As described above, according to the present embodiment, by displaying the combination of the module m and the processing p in the descending order of the total value dx ^{m, p} · em ^{, p} · S ^{m, p} , the user can The module m that greatly affects the performance can be known in consideration of the ease of improvement and the shortenable time. As a result, the target performance of the module m can be corrected, or the module m can be actually improved to contribute to efficient system performance improvement.

(Fourth embodiment)
Next, a computer system development method according to still another embodiment of the present invention will be described with reference to FIG. 10, FIG. 11, and FIG. FIG. 10 is a block diagram showing a configuration of a system (tool) for executing a computer system development method according to still another embodiment of the present invention. 11 and 12 are flowcharts showing an operation flow of the system (tool) shown in FIG. In FIG. 10, FIG. 11, and FIG. 12, the same reference numerals are given to the same components or steps shown in the previous drawings, and the description thereof is omitted.

This embodiment is a comprehensive value like the above embodiments ^{dx m, p · e m,} p · S m, to sort order of magnitude of ^p, furthermore, ^{dx m} based on the sorting ^{result, p} Is applied cumulatively, and a plurality of necessary combinations of m and p such that the evaluation function value becomes zero (= target achievement) is presented to the user.

The sensitivity analysis unit 12A illustrated in FIG. 10 has the same function as the sensitivity analysis unit 12 illustrated in FIG. 7, and further, based on the sorting result of the total values dx ^{m, p} · em ^{, p} · S ^{m, p.} Dx ^{m and p} are applied cumulatively to the module performance information. As a result, it has a function of obtaining a plurality of necessary combinations of m and p such that the evaluation function value is zero (= target achievement). Hereinafter, the operation of the system (tool) shown in FIG. 10 will be described more specifically with reference to the flowcharts shown in FIGS.

  Step 201 is the same as FIG. 2 (first embodiment), step 401 is the same as FIG. 5 (second embodiment), and step 601 is the same as FIG. 8 (third embodiment). Step 203 and step 204 are the same as those in FIG. Further, Step 205, Step 219, and Steps 206 to 214 are the same as those in FIG. Step 615 is the same as FIG.

When the sorting result of the total values dx ^{m, p} · em ^{, p} · S ^{m, p} is obtained in step 615, the sort number k is set to 1 (that is, corresponding to the largest total value) (step 817). The process proceeds to the loop processing (loop 3) of step 818 to step 823. This loop 3 is performed while the evaluation function value E is a positive value (= target not achieved) and k is equal to or less than P (number of processes p) × M (number of modules m).

First, the shortenable time dx ^{m, p} corresponding to the sort number k is reflected in the module performance information (step 819). A simulation is performed under these conditions to predict system performance (step 820). Further, an evaluation function value E _k is calculated from the prediction result (step 821). The method itself of step 820 and step 821 is the same as that of step 203 and step 204 (both in FIG. 2). If the evaluation function value E _k is a condition for continuing the loop 3, the same processing is performed for the next sort number k + 1 (step 822).

After exiting the loop 3, the total values dx ^{m, p} · em ^{, p} · S ^{m, and p} are displayed in this order (step 824: on the display unit 14). At this time, the evaluation function is based on each sort number k. A value E _k is appended (same step). This evaluation function value E _k is typically appended until it reaches zero (= target achievement) (or the evaluation function value E _k does not become zero (= target achievement) until the last sort number). It is added up to the last sort number.) Therefore, the user can see this display and know to what combination of m and p the shortenable time dx ^{m, p} is reflected in the module performance information to achieve the target.

The correction input of the module performance information for dx ^{m, p} is a review of the target performance for each module at the design stage, but the input of the module performance after the actual improvement is performed after the unit test of the module. There may be cases.

As described above, according to the present embodiment, the combination of the module m and the processing p is displayed to the user in the descending order of the total value dx ^{m, p} · em ^{, p} · S ^{m, p} , and which m, p If dx ^{m, p} is reflected in the module performance information up to the combination, it is shown whether the target is achieved. In other words, how far dx ^{m, p} is applied is automatically presented to the extent that the target can be achieved as a system to be developed. As a result, the target performance of the module m can be corrected, or the module m can be actually improved to contribute to efficient system performance improvement.

(Fifth embodiment)
Next, a computer system development method according to still another embodiment of the present invention will be described with reference to FIGS. 13, 14, and 15. FIG. FIG. 13 is a block diagram showing a configuration of a system (tool) for executing a computer system development method according to still another embodiment of the present invention. 14 and 15 are flowcharts showing the operation flow of the system (tool) shown in FIG. 13, 14, and 15, the same reference numerals are given to the same components or steps as those shown in the previous drawings, and the description thereof is omitted.

This embodiment corresponds to the case where a plurality of improvement measures are prepared and the degree of ease and the shortenable time dx ^{m, p} are given for each improvement measure. With the introduction of easiness of each improvement measures, improved ease of e ^m relates p, m as shown in FIG. ^{5, p} is not used.

  The improvement ease input unit 15B has a function of acquiring the ease (for example, a larger numerical value and easier) for each improvement measure from the outside and passing this to the sensitivity analysis unit 12B. More specifically, the improvement ease input unit 15B is similar to the input unit 11 and the improvement ease input unit 15, and includes input devices such as a keyboard and a mouse, a control program thereof, and drive devices for various recording media such as a hard disk. It can be configured by the control program, various network interface devices and the control program.

  The shortenable time input unit 16B has a function of acquiring information on the shortenable time for each improvement measure from the outside and passing the information to the sensitivity analysis unit 12B. The shortenable time information is the amount that can be shortened in the processing time in the module m and the process p as described above. This is acquired for each improvement measure. More specifically, the shortenable time input unit 16B is similar to the input unit 11 and the shortenable time input unit 16, and includes input devices such as a keyboard and a mouse, a control program thereof, and drive devices for various recording media such as a hard disk. It can be configured by the control program, various network interface devices and the control program.

Sensitivity analysis unit 12B has the same function as the sensitivity analysis unit 12 shown in FIG. 1, further, the overall evaluation value _{V k = e k Σ m Σ} p dx k m, p · S m, the ^p is calculated, And it has a function of sorting in the order of this size. Here, e _k is the degree of improvement of the improvement measure k, and dx _k ^{m, p} is the shortenable time information of the improvement measure k. Hereinafter, the operation of the system (tool) shown in FIG. 13 will be described more specifically with reference to the flowcharts shown in FIGS.

Step 201 is the same as FIG. 2 (first embodiment). Then, enter the information of easiness _{e k} for each solving method by improving easiness input unit 15B (step 1002). Further, the shortenable time dx _km ^{, p} for each improvement measure is input by the shortenable time input unit 16B (step 1003). The next step 1004 to step 1021 are loop processing (loop 1). The number of loops 1 increases each time a certain improvement is applied.

The processing of step 203 to step 205 and step 206 to step 214 in the loop 1 is the same as the step with the same sign in each of the embodiments already described. Then, the overall evaluation value _{V k = e k Σ m Σ} p dx k m for each improvement measures unreflected, p · ^{S m,} to calculate the ^p (step 1017). For example, in the first loop 1 process, since all the improvement measures are not reflected, the total evaluation value V _k is calculated by the number _k . Then, each improvement measure k is sorted in the order of the magnitude of the comprehensive evaluation value V _k (step 1018), and the sorting result is displayed (step 1019).

  The user can decide to apply the best improvement based on the sorting result. Therefore, the module performance information is corrected and input in accordance with the applied improvement measure (step 1020). Then loop 1 is repeated.

According to this embodiment, by presenting the improvement measures to the user in descending order of the comprehensive evaluation value V _k , the user can know the improvement measures that greatly affect the performance of the system. Based on this, an improvement measure can be selected. Furthermore, the system performance can be improved efficiently by repeatedly modifying the module performance information and predicting the system performance according to the selected improvement measure.

(Sixth embodiment)
Next, a computer system development method according to still another embodiment of the present invention will be described with reference to FIG. 16, FIG. 17, and FIG. FIG. 16 is a block diagram showing a configuration of a system (tool) for executing a computer system development method according to still another embodiment of the present invention. 17 and 18 are flowcharts showing the operation flow of the system (tool) shown in FIG. 16, 17, and 18, the same reference numerals are given to the same components or steps shown in the previous drawings, and the description thereof is omitted.

This embodiment is the same as the fifth embodiment in that it corresponds to a case where a plurality of improvement measures are prepared and the degree of ease and dx ^{m, p} are given for each improvement measure. Further, in this embodiment, it is possible to present a number of applications required improvement measures to the user to sort the results based on Dzute total evaluation value V _k overall evaluation value V _k is zero (= goal).

The sensitivity analysis unit 12C illustrated in FIG. 16 has the same function as the sensitivity analysis unit 12B illustrated in FIG. 13, and further applies each improvement measure in order based on the sorting result of the comprehensive evaluation value V _k , As a result, it has a function of obtaining an improvement measure necessary for the overall evaluation value V _{k to} be zero (= target achievement). Hereinafter, the operation of the system (tool) shown in FIG. 16 will be described more specifically with reference to the flowcharts shown in FIGS.

Step 201, step 1002, step 1003, and step 203 to step 214 are the same as those in the fifth embodiment. Then, the overall evaluation value _{V k = e k Σ m Σ} p dx k m for each improvement measures k, p · ^{S m,} to calculate the ^p (step 1216), further sorted in order of their total evaluation value _{V k} ( Step 1217). Steps 1216 and 1217 are substantially the same as steps 1017 and 1018 in FIG.

When the improvement measure sort result in the order of the magnitude of the comprehensive evaluation value V _k is obtained in step 1217, the sort number 1 is set to 1 (that is, corresponding to the largest overall evaluation value V _k ) (step 1218), and step 1219. Shifts to the loop processing (loop 3) in step 1224. This loop 3 is performed while the comprehensive evaluation value V _l is a positive value (= target not achieved) and l is equal to or less than the number K of improvement measures.

First, the module performance information is corrected in accordance with dx _l ^{m, p} of improvement measure l corresponding to sort number l (step 1220). A simulation is performed under these conditions to predict system performance (step 1221). Further performs the same process as step 204 to step 214 from the prediction result (excluding step 205), which overall evaluation value after the modification by _{V l = e l Σ m Σ} p dx l m, p · S m, p Is calculated (step 1222). The evaluation function value _{E l} of the improved processes as well for the next sort number l + 1 as long as the conditions to continue the loop 3 (step 1223, 1224).

Once out of the loop 3, the overall evaluation value _{V k = e k Σ m Σ} p dx k m, p · S m, is displayed in the order of ^p is made (step 1225: the display section 14), after the modification this time A comprehensive evaluation value V _k is appended (same step). The improved overall evaluation value V _k is typically appended until it reaches zero (= target achievement) (or the improved overall evaluation value V _k reaches zero (= target) until the last sort number. If not achieved), the last sort number will be added.) Therefore, the user can see this display and know how much improvement measures are applied cumulatively to achieve the target.

As described above, according to this embodiment, and displays total evaluation value _{V k = e k Σ m Σ} p dx k m, p · S m, each improvement measures in descending order of ^p to the user, further, which improves It shows whether the goal is achieved if the measures are applied cumulatively. Thereby, it can contribute to the efficient performance improvement of a system.

Example 1
Next, each of the above embodiments will be described more specifically with reference to examples. The ticket reservation sales system is taken up as a system to be developed. First, Example 1 is an example corresponding to the first embodiment.

  FIG. 19 is a table showing an example of business amount information in this ticket reservation sales system. Here, “business” is 4 business operations of 1) ticket reservation, 2) reservation cancellation, 3) ticket sales, and 4) aggregation. The business volume information is information that represents the volume of each business that occurs every 15 minutes by dividing 24 hours into 15-minute intervals. FIG. 20 is a diagram showing the workload information shown in FIG. 19 in a graph. For these workloads, for example, amounts assumed based on past data are set.

  FIG. 21 is a block diagram showing an example of the configuration of a computer system to be developed. The hardware configuration includes a web server machine 24, an application server machine (hereinafter referred to as “application server machine”) 25, and a database server machine (hereinafter also referred to as “DB server machine”) 26. The web server machine 24 includes a web server A corresponding to access from a member (client 21) via the Internet 23 and a web server B corresponding to access from a window (client 22) via a dedicated line. Has been.

  The application server machine 25 and the database server machine 26 are commonly used by members and the contact point. On the web server machine 24, a web server module A for the web server A and a web server module B for the web server B are mounted. An application server module 251 is installed on the application server machine 25, and a database server module 261 is installed on the database server machine 26. A network between the web server machine 24 and the application server machine 25 is a network A, and its communication module is a communication module network A. A network between the application server machine 25 and the database server machine 26 is a network B, and its communication module is a communication module net B.

  The related information in this embodiment includes a business definition table, a module definition table, and a resource definition table. FIG. 22 is a table showing an example of the business definition information among these pieces of related information. The business definition table is a table representing the relationship between each business and each process p. In this table, the display of “1” indicates that it is related. For example, the ticket reservation business is a business operated by the member via the Internet 23, and a series of processes of “login”, “reservation selection”, “condition designation”, “reservation”, and “logout” are performed. The column of is 1.

  FIG. 23 is a table showing an example of module definition information among the related information. The module definition table is a table showing the relationship between the module m and standard resources. In this embodiment, a machine or a network will be described as one resource. Each element of the table has a value of 0 or 1, meaning that an element of 1 is relevant. For the module m, the field of the server machine in which the module m is mounted is 1, or in the case of a communication module, the field of the network used by the communication module is 1.

  FIG. 24 is a table showing an example of resource definition information among the related information. The resource definition table is a table showing the relationship between resources and standard resources. It has the performance ratio of each standard resource to each resource as an element.

  Module performance information in this embodiment is represented by a process definition table. FIG. 25 is a table showing an example of module performance information, and is a process definition table. This process definition table has, as an element, a processing time (millisecond ms unit) required by the related program module m when the process p is executed when the reference machine or the reference network is used.

  Next, prediction of system performance in the present embodiment will be described. In this embodiment, a resource usage accumulation model is used as an example of the performance prediction model. The resource usage accumulation model is a model that adds the processing time of resources used by the process p by the number of processes for the process p requested within a certain period of time. Since the response time is not calculated in this model, it is assumed here that the upper limit value of the resource utilization rate is set as an index of system performance.

The resource usage rate of the resource usage accumulation model is calculated by the above-described business amount information table (FIG. 19), business definition table (FIG. 22), module definition table (FIG. 23), resource definition table (FIG. 24). If the element part of each table of the table (FIG. 25) is regarded as a matrix, it can be calculated by multiplication of these matrices. That is, the element part of the table of the workload information is the workload matrix [TM], the element part of the workflow definition table is the workflow definition matrix [TD], the element part of the workflow definition table is the process definition matrix [PD], and the module definition table If the element part is a module definition matrix [MD] and the element part of the resource definition table is a resource definition matrix [RD], the resource utilization matrix [R] representing the resource utilization at 15 minute intervals is
[R] = [RD] [MD] [PD] [TD] [TM] / 9000
(The unit of the [R] element is%). / 9000 in the formula means that each element is divided by 9000.

  FIG. 26 is a table showing the resource utilization rate, which is an example of the prediction result of the system performance, obtained as described above. Each element of this table indicates the utilization rate of each resource at each time. FIG. 27 is a graph showing the resource utilization rate shown in FIG. The horizontal axis indicates 24 hours in time, and the vertical axis indicates the resource utilization rate. It can be seen that the utilization rate of the database server machine has reached 101% (largely overloaded) immediately after 10:00.

となる。 Next, the evaluation function in the present embodiment will be described. Assume that the evaluation function of the present embodiment is the sum of the transitions of the resource utilization rate in the part exceeding the resource utilization rate upper limit value set as the index of system performance. FIG. 28 is a schematic diagram for explaining such an evaluation function. Each curve represents the time change of the resource utilization rate for each resource, and the area of the hatched portion above the resource utilization rate upper limit (broken line) is the evaluation function value E. That is,

It becomes.

Here, R _{r, t} represents the average utilization rate of the resource r from time 15 (t−1) to 15 t, and R hat is the resource utilization rate upper limit value set as an index of system performance (this embodiment) (40%). Further, Σ (from t = 1 to T) represents the sum of each time from 0:00 to 24:00, and T is 96 representing 24:00. Further, Σ (for r) represents the total sum of each resource. In the above, when the evaluation function value E is actually calculated in this embodiment, it becomes 64.4693.

となる。 As another evaluation function, it can be defined as the maximum value (total sum for resources) of the amount of resource utilization that exceeds the resource utilization rate upper limit set as an index of system performance. The evaluation function in this case is as follows. That is,

It becomes.

となる。さらに加えて、以上とは異なるさらに別の評価関数として、以上の評価関数を組み合わせたものとすることもできる。 Further, as another evaluation function, it can be defined as the sum of the times exceeding the resource utilization rate upper limit value set as an index of system performance (also the sum of resources). The evaluation function in this case is as follows. That is,

It becomes. In addition, as another evaluation function different from the above, the above evaluation functions may be combined.

Next, the result of performing the processing shown in FIGS. 2 and 3 in this embodiment and sorting in the order of the sensitivity S ^{m, p} will be described. FIG. 29 and FIG. 30 show the results. This is a diagram (display in order of sensitivity) showing an example of a table displayed for supporting selection of a program module that greatly affects the system performance.

  The procedure for obtaining the tables shown in FIGS. 29 and 30 is almost exhausted in the description of FIGS. 2 and 3, but will be described again just in case. Since the processing time required for the related module m when the processing p is executed is written in the processing definition table (processing definition matrix [PD]), the processing time of each element of the processing definition matrix [PD] is increased by a small amount. If the resource utilization at that time is calculated, the system performance is predicted under this condition.

で計算できる。さらに感度Ｓ^ｍ，ｐは、
Ｓ^ｍ，ｐ＝（Ｅ^ｍ，ｐ−Ｅ）／Δｔ
で計算できる。なお本実施例ではΔｔ＝０．１ｍｓである。 That is, a matrix obtained by increasing the element of m rows and p columns of the process definition matrix [PD] by Δt is represented as [PD] ^{m, p} . The resource utilization matrix [R] ^{m, p} representing the resource utilization at this time is
[R] ^{m, p} = [RD] [MD] [PD] ^{m, p} [TD] [TM] / 9000
Can be calculated as Using this system performance prediction result, each element of the resource utilization matrix [R] ^{m, p} is R ^m, _{pr, t} , and the evaluation function value ^{Em, p} is

It can be calculated with Furthermore, the sensitivity S ^{m, p} is
S ^{m, p} = (E ^{m, p} −E) / Δt
It can be calculated with In this embodiment, Δt = 0.1 ms.

The above calculation is performed for all elements of [PD]. 29 and 30 are displayed when the combinations of m and p are displayed in the order of magnitudes of the sensitivities S ^{m and p} thus obtained. In the tables shown in FIGS. 29 and 30, the processing p, the module m, the evaluation value E ^{m, p} , the sensitivity S ^{m, p} , and the processing time (from the processing definition table) are arranged.

  29 and FIG. 30, database server modules (reference numeral 261 in FIG. 21) are arranged at the top. This is because the resource utilization rate exceeds 40% in FIG. 27 is the database server machine, and the module related to the database server machine is the database server module. It can be seen that the sensitivity of all other modules is 0, and even if the other modules are improved, the portion where the resource utilization rate exceeds 40% is not improved. The reason that the sensitivity differs depending on the process p in the same database server module is because the frequency of occurrence of these processes p is different.

  The user can reasonably determine which review of the performance of the module m greatly contributes to the performance improvement as a system with reference to the order of this table. As a result, the target performance of the module m can be corrected, or the module m can be actually improved to contribute to efficient system performance improvement.

(Example 2)
Next, an example corresponding to the second embodiment will be described as a second example. The description of the system to be developed is the same as the above embodiment. FIG. 31 is a diagram illustrating an example of the improvement ease information in a table (improvement degree matrix). FIGS. 32 and 33 are diagrams (displays in the order of the product of sensitivity and ease) showing an example of a table displayed for supporting selection of program modules that greatly influence system performance. Sorting in this embodiment It is a display of the result.

Here, the improvement ease matrix is a matrix in which the ease of improvement is expressed in the range of 0 to 1 in accordance with the combination of each process p and each module m. The procedure for obtaining the tables shown in FIGS. 29 and 30 is almost exhausted in the description of FIGS. 5 and 6. By following this procedure, FIG. 32, as shown in FIG. 33, the sensitivity ^{S m, p} and improved ease of ^{e m,} total value ^{e m} is the product of the ^p, p · ^{S m,} it is sorted in the order of ^p Made.

In this table, the processing p, the module m, the evaluation value E ^{m, p} , the sensitivity S ^{m, p} , the ease em E ^{, p} , the total value em ^{, p} · S ^{m, p} , and the processing time are arranged. Items related to the database server module (reference numeral 261 in FIG. 21) are also arranged at the top of the table. For the same reason as in the first embodiment, only the database server machine has a resource utilization rate exceeding 40% (from FIG. 27), and the module related to the database server machine is the database server module. This is because. The sensitivity of all other modules is 0, and it can be seen that even if these modules are improved, the portion where the resource utilization rate exceeds 40% is not improved. In the second embodiment, the overall value of the database server module for the “condition designation” process having the improvement degree of 1.0 was maximized as a result of the evaluation of the sensitivity and the improvement degree.

  The user can reasonably determine which review of the performance of the module m greatly contributes to the performance improvement as a system with reference to the order of this table. As a result, the target performance of the module m can be corrected, or the module m can be actually improved to contribute to efficient system performance improvement.

(Example 3)
Next, an example corresponding to the third embodiment will be described as Example 3. The description of the system to be developed is the same as in the above embodiments. FIG. 34 is a table illustrating an example of the shortenable time information. FIG. 35 and FIG. 36 are diagrams showing examples of tables displayed for supporting selection of program modules that greatly affect system performance (display of products in order of sensitivity, ease, and shortenable time). It is a display of the sort result in an Example.

The shortenable time information indicates the amount of time that can be shortened according to the combination of each process p and each module m. The procedure for obtaining the tables shown in FIGS. 35 and 36 is almost exhausted in the description of FIGS. By following this procedure, 35, as shown in FIG. 36, the sensitivity ^{S m, p} and improved ease of ^{e m, p} and shorten available time ^{dx m,} total value is the three parties of the product of the ^{p e m , P} · S ^{m, p} · dx ^{m, p} .

Process in this table p, the module m, the evaluation value ^{E m, p,} sensitivity ^{S m, p,} easiness ^{e m, p,} can be shortened time ^{dx m, p,} total value ^{e m, p · S m,} p · dx ^{m, p} and processing time are arranged. Items related to the database server module are arranged at the top of the table. For the same reason as in the first and second embodiments, only the database server machine has a resource utilization rate exceeding 40% (from FIG. 27), and the modules related to the database server machine are the database. This is because it is a server module. The sensitivity of all other modules is 0, and it can be seen that even if these modules are improved, the portion where the resource utilization rate exceeds 40% is not improved. In the third embodiment, as a result of evaluation based on a total value that is a product of the three of sensitivity, ease of improvement, and shortenable time, a database server for “condition designation” processing having an ease of improvement of 1 and a shortenable time of 540 ms. The overall value is the maximum for the module.

  The user can reasonably determine which review of the performance of the module m greatly contributes to the performance improvement as a system with reference to the order of this table. As a result, the target performance of the module m can be corrected, or the module m can be actually improved to contribute to efficient system performance improvement.

Example 4
Next, an example corresponding to the above-described fourth embodiment will be described as Example 4. The description of the system to be developed is the same as in the above embodiments. FIGS. 37 and 38 are diagrams showing examples of tables displayed for the selection support or determination of program modules that greatly affect system performance (displayed in the order of products of sensitivity, ease, and shortenable time, and after improvement) Is also an additional display), and is a display of the sort result in this embodiment.

The procedure for obtaining the tables shown in FIGS. 37 and 38 is almost exhausted in the description of FIGS. 11 and 12. By following this procedure, Figure 37, as shown in FIG. 38, the sensitivity ^{S m, p} and improved ease of ^{e m, p} and shorten available time ^{dx m,} total value is the three parties of the product of the ^{p e m , P} · S ^{m, p} · dx ^{m, p} are sorted in order, and dx ^{m, p} is applied cumulatively based on the sorting result, and the evaluation function value becomes zero (= target achievement). A required combination of m and p is presented to the user.

The result table includes the processing p, the module m, the evaluation value E ^{m, p} , the sensitivity S ^{m, p} , the ease em E ^{, p} , the shortening amount dx ^{m, p} , the total value E ^{m, p} · S ^{m, p} · dx ^{m, p} , processing time, expected time, time after improvement, and overall value after improvement are arranged. Among these, processing p, module m, evaluation value E ^{m, p} , sensitivity S ^{m, p} , ease em ^{, p} , shortening amount dx ^{m, p} , total value em ^{, p} · S ^{m, p} · dx ^m, About ^p and processing time, it is the same as that of the case of said Example 3. FIG.

The estimated time represents an estimated value of the processing time after applying dx ^{m, p} . The time after improvement represents the processing time after applying dx ^{m, p} until the total value after improvement becomes 0 (= target achievement). The total value after improvement indicates a calculation result of the total values em ^{, p} · S ^{m, p} · dx ^{m, p} when dx ^{m, p} is applied cumulatively up to the combination of ^{m, p} . In the fourth embodiment, as shown in FIG. 37, the improved total value becomes 0 (= target achieved) by applying dx ^{m, p} up to the combination of the top eight modules m and processing p. I understand. Therefore, if the target performance of the module m is corrected for the combination of the eight modules m and the processing p, or the module m is actually improved, the target can be achieved as a system to be developed.

As described above, in the fourth embodiment, not only the combination of the module m and the process p that greatly influences the performance of the system is presented to the user, but also the development is further advanced by applying dx ^{m, p} . It automatically presents until the goal is achieved as a system to power.

(Example 5)
Next, an example corresponding to the fifth embodiment will be described as Example 5. The description of the system to be developed is the same as in the above embodiments. FIG. 39 is a table showing an example of the ease of improvement information for each improvement measure. FIG. 40 is a table showing an example of the shortenable time information in the improvement measure 1 in FIG. Although not shown, a shortenable time corresponding to FIG. 40 is given for each improvement measure. FIG. 41 is a table showing an example of a comprehensive evaluation value obtained for each improvement measure.

  The procedure for obtaining the table shown in FIG. 41 is almost exhausted in the description of FIGS. By following this procedure, as shown in FIG. 41, a comprehensive evaluation value is obtained for each improvement measure. The result of FIG. 41 corresponds to the result of the processing of Step 1017 and Step 1018 and the first processing of Loop 1 in FIG. As shown in FIG. 41, sorting is performed according to the order of the comprehensive evaluation values (in FIG. 41, it happens to be the same as the order of improvement measures shown in FIG. 39). Thereafter, loop 1 can be further continued by selecting an improvement measure to be arbitrarily applied and applying the applied information (as shown in FIG. 14).

  As described above, the user can know improvement measures that greatly affect the performance of the system with reference to the order shown in FIG. Based on this, an improvement measure can be selected. The system performance can be improved efficiently by repeatedly modifying the module performance information and predicting the system performance according to the selected improvement measure.

(Example 6)
Next, as Example 6, an example corresponding to the above-described sixth embodiment will be described. The description of the system to be developed is the same as in the above embodiments. FIG. 42 is a table showing an example of the comprehensive evaluation value obtained by applying the improvement measure. FIG. 43 is a table showing an example of module performance information (= processing time information) corresponding to a case where the evaluation value after improvement is zero (= target achievement) in FIG.

  The procedure for obtaining the table shown in FIG. 42 is almost exhausted in the description of FIGS. By following this procedure, the “total evaluation value” in FIG. 42 is obtained in step 1217. Furthermore, the “improved comprehensive evaluation value” in FIG. 42 is obtained by the processing of step 1218 to step 1225. As shown in FIG. 42, in the sixth embodiment, when the improvement measures 1 to 3 are applied cumulatively, the improved overall table value becomes 0 (= target achievement). Then, by applying the shortenable time of improvement measures 1 to 3 to the process definition table (FIG. 25), the “process time after improvement” as shown in FIG. 43 is obtained.

  Thus, in the case of the sixth embodiment, it can be seen that the evaluation value becomes 0 (= target achieved) by the top three improvement measures, and the module performance target is reset or improved according to these three improvement measures. Indicates that it should be. That is, in the sixth embodiment, not only the improvement measures that greatly affect the performance of the system are presented to the user in the order of the magnitude of the effects, but also the development measures can be developed by applying the improvement measures up to this point. It can be automatically presented until the goal is achieved as a system to be operated.

  In the sixth embodiment, when the module processing time is shortened by a plurality of improvement measures, the total value of the shortenable time of each improvement measure is subtracted. However, depending on the improvement measure, for example, even if the improvement measure 2 is executed after the improvement measure 1 is executed, the possible shortening time of the improvement measure 2 may not be realized. Therefore, a method of setting the shortening processing time by a ratio to the processing time instead of setting by time is also conceivable. Also, instead of setting a shortened time for improvement measures, set a processing time after shortening for each improvement measure, and when multiple improvements are implemented, out of the multiple improvements, the processing after shortening It is also possible to set so that the shortest processing time is selected.

(Other examples)
In each of the embodiments described above, the resource utilization rate is predicted using the resource usage accumulation model. However, the performance prediction model of the performance prediction unit 13 (FIG. 1 and the like) includes a queuing network, a dynamic simulator, Sensitivity can be calculated in the same way even when the component block model in “Computer System Performance Prediction Program and Design Support System” disclosed in Japanese Unexamined Patent Publication No. 2004-272582 is used.

  In addition, as an index of system performance, this has been explained as the upper limit value of resource utilization, but sensitivity is calculated even when response time is set or when both resource utilization and response time are used as system performance indices. can do. In the above embodiment, the computer is described as a unit resource. However, the CPU, hard disk, and memory of the computer are used as resources, and the CPU usage, disk access time, memory access time, etc. when the module calculates by processing are used as resources. It can also be set as time.

The block diagram which shows the structure of the system (tool) for performing the development method of the computer system based on one Embodiment of this invention. The flowchart which shows the operation | movement flow of the system (tool) shown in FIG. FIG. 3 is a continuation diagram of FIG. 2 and is a flowchart showing an operation flow of the system (tool) shown in FIG. 1. The block diagram which shows the structure of the system (tool) for performing the development method of the computer system based on another embodiment of this invention. The flowchart which shows the operation | movement flow of the system (tool) shown in FIG. 6 is a continuation diagram of FIG. 5, and is a flowchart showing an operation flow of the system (tool) shown in FIG. 4. The block diagram which shows the structure of the system (tool) for performing the development method of the computer system based on further another embodiment of this invention. The flowchart which shows the operation | movement flow of the system (tool) shown in FIG. FIG. 9 is a continuation diagram of FIG. 8, and is a flowchart showing an operation flow of the system (tool) shown in FIG. 7. The block diagram which shows the structure of the system (tool) for performing the development method of the computer system based on further another embodiment of this invention. 11 is a flowchart showing an operation flow of the system (tool) shown in FIG. 10. FIG. 12 is a continuation diagram of FIG. 11 and is a flowchart showing an operation flow of the system (tool) shown in FIG. 10. The block diagram which shows the structure of the system (tool) for performing the development method of the computer system based on further another embodiment of this invention. 14 is a flowchart showing an operation flow of the system (tool) shown in FIG. 13. FIG. 15 is a continuation diagram of FIG. 14, and is a flowchart showing an operation flow of the system (tool) shown in FIG. 13. The block diagram which shows the structure of the system (tool) for performing the development method of the computer system based on further another embodiment of this invention. The flowchart which shows the operation | movement flow of the system (tool) shown in FIG. FIG. 18 is a continuation diagram of FIG. 17, and is a flowchart showing an operation flow of the system (tool) shown in FIG. 13. The figure which shows an example of work amount information with a table | surface. The figure which shows the work amount information shown in FIG. 19 with a graph. The block diagram which shows an example of a structure of the computer system which should be developed. The figure which shows an example of business definition information among each related information in a table | surface. The figure which shows an example of module definition information by table among each relevant information. The figure which shows an example of resource definition information by a table among each relevant information. The figure which shows an example of module performance information in a table | surface. The figure which shows the resource utilization rate which is an example of the prediction result of a system performance with a table | surface. The figure which shows the resource utilization rate shown in FIG. 26 with a graph. Explanatory drawing which shows an example of an evaluation function typically. The figure which shows an example of the table | surface displayed for the selection assistance of the program module which influences a system performance greatly (display in order of sensitivity). FIG. 30 is a continuation diagram of FIG. 29 showing an example of a table displayed for assistance in selecting program modules that greatly affect system performance (display in order of sensitivity); The figure which shows an example of improvement ease information by a table | surface. The figure which shows an example of the table | surface displayed for the selection assistance of the program module which influences system performance greatly (display of the order of the product of a sensitivity and ease). FIG. 33 is a continuation diagram of FIG. 32, illustrating an example of a table displayed for supporting selection of program modules that greatly affect system performance (display of the order of products of sensitivity and ease). The figure which shows an example of shortenable time information with a table | surface. The figure which shows an example of the table | surface displayed for the selection assistance of the program module which influences system performance greatly (display of the order of the product of a sensitivity, an ease degree, and the shortenable time). FIG. 36 is a continuation diagram of FIG. 35, showing an example of a table displayed to assist selection of program modules that greatly influence system performance (display in order of products of sensitivity, ease, and shortenable time). The figure which shows an example of the table | surface displayed for the selection support or determination of the program module which influences system performance greatly (it displays in order of the product of a sensitivity, an ease, and a shortenable time, and the product after improvement is also displayed additionally). FIG. 38 is a continuation diagram of FIG. 37, showing an example of a table displayed for the selection support or determination of program modules that greatly influence system performance (displayed in the order of the product of sensitivity, ease, and shortenable time; The product after improvement is additionally displayed). The figure which shows an example of the improvement ease information for every improvement measure with a table | surface. The figure which shows an example of the shortenable time information in the improvement measure 1 in FIG. 39 with a table | surface. The figure which shows an example of the comprehensive evaluation value calculated | required for every improvement measure with a table | surface. The figure which shows an example of the comprehensive evaluation value calculated | required by applying an improvement measure. FIG. 43 is a table showing an example of module performance information (= processing time information) corresponding to a case where the improved evaluation value is zero (= target achievement) in FIG.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 11 ... Input part, 12, 12A, 12B, 12C ... Sensitivity analysis part, 13 ... Performance prediction part, 14 ... Display part, 15, 15B ... Improvement ease input part, 16, 16B ... Shortening time input part, 21 ... Client (for member), 22 ... Client (for window), 23 ... Internet, 24 ... Web server machine, 25 ... Application server machine, 26 ... Data server machine, 27 ... Network A, 28 ... Network B, 251 ... Application server Module, 261 ... Database server module.

Claims (6)

1st information regarding the amount of each business which the computer system which should be developed carries out, 2nd information regarding the relation between each said business and each process performed in the said computer system, Each said process and the said computer system are comprised. Inputting third information related to the relationship with each module, and fourth information related to the processing time for each module related to each processing;
Predicting the performance of the computer system using the first to fourth information to obtain a first prediction result;
Performing evaluation with a predetermined evaluation function using the first prediction result of the computer system to obtain a first evaluation function value;
In the combination of one of the processes and one of the modules, the computer system is used for a plurality of different combinations by using the first to fourth information based on a condition in which the fourth information is slightly changed. Predicting performance and obtaining respective second prediction results;
Performing evaluation by the predetermined evaluation function using the second prediction result of the computer system to obtain respective second evaluation function values;
Based on the conditions in which the first evaluation function value, the second evaluation function value, and the fourth information are slightly changed, the plurality of the one of the processes and the one of the modules Determining each sensitivity to the first evaluation function value in combination;
A computer system development program for causing a computer to execute the step of rearranging the plurality of combinations of the one of the processes and the one of the modules in order of the sensitivity. Inputting, as fifth information, information on the degree of ease of improvement for each of the modules related to each of the processes;
Using the sensitivity and the fifth information, respectively, obtaining total values for the plurality of combinations of the one of the processes and the one of the modules;
In the order of magnitude of the total value, the development of the computer system according to claim 1, wherein for execution by the further computer and the step of rearranging said plurality of combinations of said one of said modules and said one of the processing program. Inputting information on the degree of ease of improvement for each module related to each processing as fifth information, and inputting information on the shortenable time for each module related to each processing as sixth information,
Using the sensitivity and the fifth and sixth information, respectively, obtaining total values for the plurality of combinations of the one of the processes and the one of the modules;
In the order of magnitude of the total value, the development of the computer system according to claim 1, wherein for execution by the further computer and the step of rearranging said plurality of combinations of said one of said modules and said one of the processing program. The performance of the computer system is predicted by reflecting the shortenable time corresponding to the combination of each processing and each module corresponding to the combination of the corresponding processing and the module, and the third prediction result Each obtaining step,
Performing evaluation with the predetermined evaluation function using the third prediction result of the computer system to obtain respective third evaluation function values;
The computer system development program according to claim 3 , further causing the computer to execute whether or not the third evaluation function value has reached a predetermined value. Step of inputting information on the shortenable time for each module related to each processing as fifth information for each of a plurality of improvement measures and information on the ease of improvement of each of the plurality of improvement measures as sixth information When,
Using the sensitivity and the fifth and sixth information to determine a comprehensive evaluation value for each of the improvement measures;
Wherein the order of magnitude of the total evaluation value, according to claim 1 wherein the computer system development program to be executed on the improvement measures the sort step and the further computer. From the one having a large comprehensive evaluation value, reflecting the shortenable time according to the corresponding improvement measure in the fourth information to predict the performance of the computer system to obtain a third prediction result, respectively;
Performing evaluation with the predetermined evaluation function using the third prediction result of the computer system to obtain a third evaluation function value;
In the combination of one of the processes and one of the modules, the computer system is used for a plurality of different combinations by using the first to fourth information based on a condition in which the fourth information is slightly changed. Predicting performance and obtaining each of the fourth prediction results;
Performing evaluation with the predetermined evaluation function using the fourth prediction result of the computer system to obtain a fourth evaluation function value;
From the conditions in which the third evaluation function value, the fourth evaluation function value, and the fourth information are slightly changed, the plurality of the one of the processes and the one of the modules Obtaining a sensitivity to the third evaluation function value as a second sensitivity in each combination; and
Using the second sensitivity and the fifth and sixth information to determine a second overall evaluation value for each of the improvement measures;
The computer system development program according to claim 5 , further causing the computer to execute a step of determining whether or not the second total function value has reached a predetermined value.

Images