JP2019101830A - System for evaluating system characteristics and device for evaluating system characteristics - Google Patents

Abstract

To evaluate the priority of an section of a system to be improved.SOLUTION: A software structure analysis unit 1011 analyzes a relationship between a calling relationship among programs in source code and data, and generates a graph structure with nodes representing programs and directional edges representing the calling relationship among programs. A characteristic analysis unit 1021 calculates a characteristic value of a business system 30 for each node on the basis of an analysis result of the source code and operation information or fault information of the business system 30, in which and the program are executed. An improvement section analysis unit 1023 merges node on the basis of the characteristic value and reconstructs the graph structure on the basis of a result of the node merging.SELECTED DRAWING: Figure 2

Description

  The present invention relates to a system characteristic evaluation system and system characteristic evaluation apparatus applicable to system design, development and maintenance.

  In response to requests from customers etc., it may be required to renovate a large-scale information system such as a production management system. Such an information system has a large divergence between the specification and the source code, and there are many that have been several decades old. For this reason, it only takes enormous man-hours to grasp the availability of the managed source code, and it is often difficult to grasp the entire software structure and design guidelines.

  In addition, since such an information system often has a plurality of functions different in nature, there are many cases in which there are many requests for improvement in function. In such a situation, it is difficult to capture all the improvement requests for the functions, and it is necessary to cope with the high-performance improvement requests while operating the information system.

  In such circumstances, by analyzing dependencies between dependency components of the source code and dependency types, and specifying importance levels for each dependency type, the scope of influence when modifying the source code is extracted, and There is a technology to support the evaluation. (Patent Document 1)

JP 2012-230538 A

  However, the technology disclosed in Patent Document 1 selects a repair target based on the existing structure obtained from the analysis result of the source code, that is, static information. Based on the analysis results of such a structure alone, it is difficult to judge the priority of repair taking into consideration the current business use status and maintainability, or the time-series change of repair results and the like.

For example, even if there is a problem in the structure as a result of source code analysis, if the program has no problem in business use, the priority of repair is likely to be low.
In addition, with regard to improvement of non-functionality such as processing time and availability, the latter process is the same even if the processing time is the same in batch processing performed at night and processing started and executed by the user during the daytime The improvement priority is likely to be high.

  On the other hand, in a system in which a large number of users use at individual timings according to individual purposes, the priority of each function is likely to be independent of the execution time zone. As described above, in order to determine the object to be repaired, it is desirable that the method of use of the system, the track record of development, and the like be recognized as characteristics indicating characteristics of the system.

  The present invention has been made in view of the above circumstances, and an object thereof is to provide a system characteristic evaluation system and a system characteristic evaluation device capable of evaluating the priority of the improvement part of the system.

  In order to achieve the above object, a system characteristic evaluation system according to a first aspect includes a system characteristic evaluation device and a display device. The system characteristic evaluation device analyzes a relation between a call relationship between programs and data in source code, and generates a graph structure with the program as a node and the call relationship of the program as a directed edge; A characteristic analysis unit that calculates a characteristic value of the system for each of the nodes based on the analysis result of the source code and system information of a system on which the program is executed; and merging the nodes based on the characteristic values; And an improvement location analysis unit that reconstructs the graph structure based on the merge result of the nodes. The display device displays the graph structure reconstructed by the improvement point analysis unit.

  According to the present invention, the priority of improvement points in the system can be evaluated.

FIG. 1 is a block diagram showing a schematic configuration of a system characteristic evaluation system according to the embodiment. FIG. 2 is a block diagram showing a schematic configuration of the system characteristic evaluation apparatus of FIG. FIG. 3 is a block diagram showing a specific configuration of the system characteristic evaluation system of FIG. FIG. 4 is a block diagram showing a hardware configuration of the system characteristic evaluation apparatus of FIG. FIG. 5 is a diagram showing an example of the contents of the node information table of FIG. FIG. 6 is a diagram showing an example of the content of the edge information table of FIG. FIG. 7 is a view showing an example of the characteristic data of FIG. FIG. 8 is a flowchart showing the operation of the software structure analysis unit of FIG. FIG. 9 is a flowchart showing the operation of the characteristic analysis unit of FIG. FIG. 10 is a diagram showing an operation flow of the improvement location analysis unit of FIG. 3 at the time of specification of a characteristic item. FIG. 11 is a flowchart showing the operation of the improvement point analysis unit of FIG. FIG. 12 is a view showing an example of a display screen of a graph structure constructed based on the node information table of FIG. 5 and the edge information table of FIG. FIG. 13 is a diagram showing an example in which the characteristic value of the selected node is displayed on the display screen of FIG. FIG. 14 is a diagram showing an example of a display screen of the graph structure after reconstruction displayed on the display device of FIG. FIG. 15 is a diagram showing an example of the contents of the node information table corresponding to the graph structure after reconstruction in FIG. FIG. 16 is a diagram showing an example of the contents of the edge information table corresponding to the graph structure after reconstruction in FIG.

  Embodiments will be described with reference to the drawings. The embodiments described below do not limit the invention according to the claims, and all of the elements described in the embodiments and their combinations are essential to the solution means of the invention. There is no limit.

FIG. 1 is a block diagram showing a schematic configuration of a system characteristic evaluation system according to the embodiment.
In FIG. 1, the system characteristic evaluation system is provided with a system characteristic evaluation unit 10, a display unit 20, a business system 30, a source code management unit 40 and a fault management unit 50. The system characteristic evaluation device 10, the display device 20, the business system 30, the source code management device 40, and the fault management device 50 are connected via the network 70. The network 70 may be a WAN (Wide Area Network) such as the Internet or a digital leased line, or a LAN (Local Area Network) such as Ethernet (registered trademark) or WiFi, or the WAN and the LAN It may be mixed. The network 70 may be wired, wireless, or a combination of wired and wireless.

  The system characteristics evaluation device 10 can evaluate the characteristics of the business system 30 on which the program in the source code is executed, based on the static information and the dynamic information of the source code. Static information about the source code is information that does not change regarding the execution of the source code. As static information on source code, there can be mentioned a graph structure in which a program in the source code is a node and a calling relationship of the program is a directed edge. Dynamic information about the source code is information that can change with respect to the execution of the source code. The dynamic information on the source code can include system information of the business system 30. As system information of the business system 30, operation information and fault information of the business system 30 can be mentioned.

  The display device 20 can display the evaluation result of the characteristic of the business system 30 based on the operation of the user 80. The business system 30 can execute various programs related to the business of the user 90. The source code management device 40 can manage source code executed by the business system 30. The fault management device 50 can manage faults in the business system 30.

  The system characteristics evaluation device 10 can obtain source code from the source code management device 40. The system characterization apparatus 10 can generate static information of source code based on structural analysis of the source code. The system characteristic evaluation device 10 can acquire dynamic information on source code from the business system 30, the source code management device 40 and the fault management device 50.

  At this time, the application log of the business system 30 which is the target of the system characteristic evaluation is acquired from the business system 30, the source code of the business system 30 and the change history of the source code are acquired from the source code management device 40 The fault history can be acquired from the fault management device 50. Then, the system characteristic evaluation device 10 can generate the graph structure of the source code based on the structural analysis of the source code. In addition, the system characteristic evaluation device 10 can manage characteristic values obtained by analyzing various pieces of information for each element of the source code. An element of source code is a program included in the source code. This program may be a shell script. The various information is static information and dynamic information on the source code. The characteristic values can be used to evaluate the characteristics of the business system 30 on which the program is executed. The characteristic values can be evaluated numerically with respect to characteristic items set in view of program implementation and structure, in view of program correction history, and in use of the program.

  Then, the system characteristic evaluation device 10 can symbolize the graph structure of the source code and display it on the display device 20. By visually recognizing the display device 20, the user 80 can determine the priority of the improvement of the source code.

  Here, the system characteristic evaluation apparatus 10 evaluates the characteristic of the business system 30 while referring to not only static information on the source code but also dynamic information on the source code, thereby using the current business of the business system 30. It is possible to determine the priority of the improvement of the source code in which the time-series changes such as the situation, the maintainability, or the repair result of the source code are reflected.

For example, even if there is a problem in the structure as a result of the structural analysis of the source code, the priority of the improvement of the source code can be lowered if the program has no problem in business use in the business system 30.
In addition, with regard to improvement of non-functionality such as processing time and availability of programs, batch processing performed at night and processing started and executed by the user during the day have the same processing time, but the latter processing The priority of the improvement of the program used for

  Furthermore, in the business system 30 in which a large number of users 90 use at individual timings according to each purpose, the priority of improvement of each function can be set so as not to be influenced by the program execution time zone.

  In the example shown in FIG. 1, the system characteristic evaluation unit 10 is connected to the business system 30 and the source code management unit 40 via the network 70. However, the system characteristic evaluation unit 10 is one function of the business system 30. The system characteristic evaluation device 10 may be used as one function of the source code management device 40.

FIG. 2 is a block diagram showing a schematic configuration of the system characteristic evaluation apparatus of FIG.
In FIG. 2, the system characteristic evaluation apparatus 10 is provided with a structure analysis unit 1011 and a characteristic evaluation unit 102. The structure analysis unit 101 is provided with a software structure analysis unit 101 and software structure data 111. The characteristic evaluation unit 102 is provided with a characteristic analysis unit 1021, a characteristic value calculation unit 1022, an improvement point analysis unit 1023, characteristic data 112, and display software structure data 113.

  The structure analysis unit 101 generates static information on the source code based on the structural analysis of the source code. The characteristic evaluation unit 102 evaluates the characteristic of the business system 30 on which the program in the source code is executed, based on the static information and the dynamic information of the source code.

  The software structure analysis unit 1011 analyzes the source code collected from the source code management device 40, specifies a program or data serving as a node, a calling relationship between programs serving as an edge, and a calling relationship between a program and data, Store in 111.

  The characteristic analysis unit 1021 analyzes the characteristic of the business system 30 for each of the nodes stored in the software structure data 111. Specifically, the characteristic items stored in the characteristic data 112 are based on the application log collected from the business system 30, the change history collected from the source code management device 40, the failure history collected from the failure management device 50, etc. The characteristic value is calculated and stored in the software structure data 111.

  The characteristic value calculation unit 1022 calculates, for each node stored in the software structure data 111, the characteristic value of each characteristic item. The characteristic value calculation unit 1022 can be provided for each characteristic item. At this time, an ID specifying the characteristic value calculation unit 1022 can be linked to the characteristic item and set in the characteristic data 112. The characteristic value calculation unit 1022 can be called and operated by the characteristic analysis unit 1021.

  In the example shown in FIG. 2, the characteristic value calculation unit 1022 is provided separately from the characteristic analysis unit 1021, but the characteristic value calculation unit 1022 may be configured as one function of the characteristic analysis unit 1021. .

  The improvement point analysis unit 1023 determines whether the characteristic value of each program and data stored in the software structure data 111 exceeds the threshold stored in the characteristic data 112 with respect to the characteristic item received by the display device 20. can do. Further, the improvement point analysis unit 1023 reconstructs the software structure data 111 based on the determination result, and stores the software structure data 111 in the display software structure data 113.

Hereinafter, the operation of the system characteristic evaluation apparatus 10 will be described with reference to FIGS. 1 and 2.
The software structure analysis unit 1011 acquires the source code of the target business system 30 from the source code management system 40, and analyzes programs (functions) and data, the calling relationship between programs, and the calling relationship between programs and data. Then, the software structure analysis unit 1011 generates a graph structure in which the program and data referenced by the program are nodes, the call relationship of the program and the reference relationship between the program and data are directed edges, and is stored in the software structure data 111.

  Next, the characteristic analysis unit 1021 analyzes the characteristic values of the nodes stored in the software structure data 111. The characteristic analysis unit 1021 calls the characteristic value calculation unit 1022 linked to each characteristic item, analyzes the data to be analyzed for each program, and stores the analysis result in the software structure data 111.

  Furthermore, the system characteristic evaluation device 10 arranges the nodes in order of execution time series based on the software structure data 111 storing the characteristic values. Then, the type, color, thickness, size or display position is set for each node and each edge according to the characteristic value, and each node and each edge can be symbolized based on the setting information. Then, the symbolized information is transmitted to the display device 20 via the network 70. The display device 20 sets the type, color, thickness, size or display position for each node and each edge based on the symbolized information, and displays each node in chronological order of execution.

  Next, when the system characteristic evaluation device 10 receives the characteristic item designated by the user 80 from the display device 20, the system characteristic evaluation device 10 operates the improvement location analysis unit 1023. The improvement point analysis unit 1023 determines, for each node, whether or not the threshold value stored in the characteristic data 112 is exceeded for each characteristic value of the node. Then, the improvement point analysis unit 1023 sets a node exceeding the threshold as an improvement candidate, and sets a node of the software structure data 111 as a node of the display software structure data 113 as it is. Nodes that did not exceed the threshold merge with nodes that did not exceed the other threshold, and the merged node is set as a node of the display software structure data 113. Also, an edge is set for the merged node, and the edge is used as an edge of the display software structure data 113.

  Furthermore, the improvement location analysis unit 1023 recalculates the characteristic value for the merged node. When calculating the characteristic value, the improvement point analysis unit 1023 can call the characteristic analysis unit 1021 to cause the characteristic analysis unit 1021 to recalculate the characteristic value. The characteristic value of the merged node can be calculated based on the characteristic value of the original node.

  Then, the system characteristic evaluation device 10 sets the type, color, thickness, size or display position to each node and each edge according to the recalculated characteristic value, and based on the setting information, each node and each position Edges can be symbolized. Then, the symbolized information is transmitted to the display device 20 via the network 70. The display device 20 sets the type, color, thickness, size or display position for each node and each edge based on the symbolized information, and displays each node in chronological order of execution.

  As a result, the user 80 can easily confirm the node to be improved for the specified characteristic item while confirming the software structure and the execution order and the relationship between the programs based on the arrangement of the graph structure. For this reason, the user 80 can easily specify the improvement target in the source code while considering the business positioning and the like.

  In addition, the system characteristic evaluation device 10 displays nodes with low priority without losing the software structure, execution order, and relationship among programs by merging nodes whose characteristic values do not exceed the threshold. You can reduce the number. For this reason, even in a large-scale information system having hundreds of programs or more, the user 80 can easily specify an improvement target in the source code while considering the business position and the like.

  Furthermore, even if the difference between the specification and the source code becomes large with the change of the source code, and it is difficult to grasp the entire software structure and design guidelines, etc., the user 80 is positioned in the business. While taking into consideration, it is possible to easily identify the improvement target in the source code.

  Further, even when the business system 30 has a plurality of functions of different natures and there are a large number of function improvement requests, it is possible to easily identify the improvement target in the source code while capturing all the improvement requests for the functions.

FIG. 3 is a block diagram showing a specific configuration of the system characteristic evaluation system of FIG.
In FIG. 3, in the software structure data 111, a node information table 1111 and an edge information table 1112 are provided. In the display software structure data 113, a display node information table 1131 and a display edge information table 1132 are provided. The business application system 30 holds an application log 3001. The source code management device 40 holds a source code 4001 and a change history 4002 of the source code 4001. The fault management device 50 holds a fault history 5001 related to the business system 30.

  The software structure analysis unit 1011 acquires the source code 4001 of the target business system 30 from the source code management device 40, and analyzes programs (functions) and data, a calling relationship between programs, and a calling relationship between programs and data. The software structure analysis unit 1011 updates the node information table 1111 with the program and data as nodes based on the analysis result, and the edge information table 1112 with the program call relationship and the program and data call relationship as directed edges. Update.

  Next, the property analysis unit 1021 acquires the software structure data 111 updated by the software structure analysis unit 1011. Also, the characteristic analysis unit 1021 acquires the change history 4001 from the source code management device 40 via the network 70, acquires the application log 3001 from the business system 30, and acquires the failure history 5001 from the failure management device 50. Then, the characteristic analysis unit 1021 operates the characteristic value calculation unit 1022 corresponding to the characteristic item stored in the characteristic table 1121, and calculates the characteristic value of the characteristic item for each node. Then, the characteristic analysis unit 1021 stores the characteristic value calculated for each characteristic item in the node information table 1111 for each node.

  Next, when the system characteristic evaluation device 10 receives the characteristic item designated by the user 80, the improvement point analysis unit 1023 operates. The improvement location analysis unit 1023 refers to the node information table 1111 and the characteristic table 1121 and determines whether each node of the node information table 1111 exceeds the threshold set in the characteristic table 1121. Furthermore, the improvement location analysis unit 1023 reconstructs the graph structure for display based on the determination result, stores the reconstructed program and data of the graph structure as nodes in the display node information table 1131, and The call relation of the restructured program of the graph structure and the call relation of the program and data are stored in the display edge information table 1132 as directed edges.

FIG. 4 is a block diagram showing a hardware configuration of the system characteristic evaluation apparatus of FIG.
In FIG. 4, the system characteristic evaluation apparatus 10 is provided with a main control unit 103, a main storage unit 104, an input unit 105, an output unit 106, and a communication unit 107. The main control unit 103 is provided with a processor 1031 and a RAM 1032.

The main control unit 103 loads a program stored in the main storage unit 104 into the RAM 1032, and the processor 1031 executes the program to control information processing in the system characteristic evaluation apparatus 10.
The main storage unit 104 stores information related to the system characteristic evaluation apparatus 10 in a non-volatile manner. The main storage unit 104 may be a magnetic disk drive or an SSD (Solid State Drive).

  The main storage unit 104 stores a software structure analysis program 1011P, a characteristic analysis program 1021P, a characteristic value calculation program 1022P, an improvement location analysis program 1023P, software structure data 111, characteristic data 112, and display software structure data 113. The characteristic value calculation program 1022P can be provided for each characteristic item.

  The software structure analysis unit 101 can be realized by the processor 1031 executing the software structure analysis program 1011P. The characteristic analysis unit 1021 can be realized by the processor 1031 executing the characteristic analysis program 1021P. The characteristic value calculation unit 1022 can be realized by the processor 1031 executing the characteristic value calculation program 1022P. The improvement location analysis unit 1023 can be realized by the processor 1031 executing the improvement location analysis program 1023P.

The input unit 105 is input by the user to the system characteristic evaluation device 10 when it is necessary to manually input information on the system characteristic evaluation device 10.
The output unit 106 outputs the information regarding the system characteristic evaluation apparatus 10 to the user when the system characteristic evaluation apparatus 10 needs to output information.
The communication unit 107 is connected to the display device 20, the business system 30, the source code management device 40 and the fault management device 50 via the network 70, and the display device 20, the business system 30, the source code management device 40 and the fault management device Communicate with 50

FIG. 5 is a diagram showing an example of the contents of the node information table of FIG.
In FIG. 5, the node information table 1111 includes a node ID 501 allocated for each program or data, a processing name 502 capable of identifying a program name or data name for each node, and a type 503 of program or data. , And stores the characteristic 504 of the business system 30 relating to program execution.

  The characteristic 504 stores at least one set characteristic item. As the characteristic 504, the real number value 5041 of the characteristic 504 and the characteristic value 5042 which performed normalization etc. can be stored, respectively.

  In the example of FIG. 5, a shell script calls a program implemented in COBOL or C language, and an example of storage of a node of a system in which the program calls data is shown. Moreover, the example in which the processing time is set as the characteristic item is shown.

FIG. 6 is a diagram showing an example of the content of the edge information table of FIG.
In FIG. 6, in the edge information table 1112, an edge ID 511, an edge call source node ID 512, and an edge call destination which can be uniquely identified for each call relationship between programs or reference relationships between programs and data shown in FIG. The node ID 513 is stored. The call source node ID and the call destination node ID use the node ID 501 stored in the node information table 1111.

  Note that since the edge is directed, it can be divided and held so that the reference source node and the reference destination node can be specified. For example, when a plurality of programs are called from one program, edge IDs having the same calling source program ID but different calling destination program IDs are created.

  Further, in the node information table 1111, the kind of relationship 514 can also be stored so that it is possible to distinguish between calls between programs or calls between programs and data. In the example of FIG. 6, the calling order of the program is represented by "next", and the reference relationship of data is represented by "in" or "out". Also, when a shell script calls a program, in a hierarchically structured program group, a program in the upper hierarchy is expressed as "call" as a relationship for calling a program in the lower hierarchy. The type of relationship 514 can be changed as appropriate according to the software structure.

FIG. 7 is a view showing an example of the characteristic data of FIG.
In FIG. 7, the characteristic table 1121 stores a characteristic ID 601, a characteristic name 602, a threshold 603, and a characteristic value calculation unit ID 604 for each of the characteristic items managed by the node information table 1111. The characteristic ID 601 is an ID that can uniquely identify the characteristic item. The property name 602 can indicate the name of the property to be displayed on the node information table 1111 and the display device 20. The threshold 603 is a reference of the characteristic of the node that can be an improvement target when the characteristic of the characteristic item designated by the user 80 is represented by a numerical value.

  The threshold 603 can be preset for the real value 5041 or the characteristic value 5042 of the characteristic 504. FIG. 7 shows an example in which the threshold value 603 is set for the real number value 5041 of the characteristic 504. However, the reference used to set the threshold 603 can be changed. For example, a change by the administrator may be made, or a history of the characteristic values 5042 before and after the system repair may be acquired, and the threshold value 603 may be set based on the actual value at that time.

  The characteristic value calculation unit ID 604 is an ID that can uniquely identify a program that analyzes the characteristic ID.

  In the example of FIG. 7, the characteristic name managed as the characteristic item is processing time, number of programs, number of special syntax, importance of data, number of special syntax of mounted language, execution time zone, failure frequency, repair frequency, Although the number of simultaneous repairs and the variation in the number of calling programs have been described, other characteristic items may be set.

FIG. 8 is a flowchart showing the operation of the software structure analysis unit of FIG.
In FIG. 8, the software structure analysis unit 1011 can operate with the detection that the source code has been added or changed in the repository of the source code management device 40 as a trigger.

  At this time, the software structure analysis unit 1011 acquires the source code 4001 of the business system 30 from the source code management device 40, and analyzes the source code 4001 (S701). Specifically, the software structure analysis unit 1011 identifies the identification of the program and data in the source code 4001, the relationship between the program and data, the calling relationship of the program, and the like.

  Next, based on the analysis result of the source code 4001, the software structure analysis unit 1011 assigns a node ID to the identified program and data as a node, and updates the node information table 1111 (S702).

  Next, the call relation and reference relation of the processing and data are analyzed, and the edge information table 1112 is updated based on the analysis result (S703).

  The system characteristics evaluation apparatus 10 can manage the change of the element of the software structure to be displayed on the display device 20 by the process of FIG.

FIG. 9 is a flowchart showing the operation of the characteristic analysis unit of FIG.
In FIG. 9, the characteristic analysis unit 1021 can operate periodically when the software structure data 111 is changed. At this time, the property analysis unit 1021 acquires the node ID 501 from the node information table 1111 (S801).

Next, the characteristic analysis unit 1021 acquires the characteristic value calculation unit ID 604 for each characteristic item from the characteristic table 1121, acquires the software structure from the software structure data 111, and acquires the change history 4001 from the source code management device 40. The application log 3001 is acquired from the business system 30, and the failure history 5001 is acquired from the failure management apparatus 50.
(S802).

  Next, the property analysis unit 1021 operates the property value calculation unit 1022 for each property item to calculate property values 5042 for all the property items (S803).

  Next, the property analysis unit 1021 determines whether the property values 5042 of all the nodes have been calculated (S804). Then, when the characteristic values 5042 of all the nodes are not calculated, the process returns to S803, and the processes of S803 and S804 are repeated until the characteristic values 5042 of all the nodes are calculated. Then, when calculating the characteristic values 5042 of all the nodes, the characteristic analysis unit 1021 stores the characteristic values 5042 in the node information table 1111 together with the real values 5041 of the characteristic 504 (S805).

  The system characteristic evaluation apparatus 10 can manage the change of the characteristic value 5042 of the node by the process of FIG.

  Hereinafter, a specific example of the analysis method will be shown for each of the characteristic names 602 of the characteristic table 1211 described in FIG.

For the processing time of the characteristic IDI1, the application log 3001 is analyzed, and the processing time is calculated from the start time and end time of the process.
As for the number of programs of the characteristic IDI2, the source code 4001 is analyzed to calculate the number of calls from one program to another for a program having a calling relationship.

  For the number of special syntaxes of the characteristic IDI3, the source code 4001 is analyzed to calculate the number of times the syntax and the process which are not common in the implementation language are used for the process implementation method. In particular, in programs that use a legacy language such as the COBOL language as an implementation language, programs with older development dates may be used more frequently, which may make it difficult to repair.

As for the data importance of the characteristic ID I4, the source code 4001 is analyzed, and the number is set based on the number of programs that refer to the data and the number of references between multiple functions. If there is local data that is referenced only by a certain function and data that is referenced by various functions, the latter is set to be of high importance.
With regard to the execution time zone of the characteristic ID I5, the application log 3001 and batch setting information are referred to, and time information at which the program is executed is set.

As for the failure occurrence frequency of the characteristic IDI 6, the failure information 5001 is analyzed, and the number of failure occurrences in a preset period is calculated.
For the modification frequency of the characteristic IDI 7, the change history 4002 is analyzed, and the number of times of modification of the program is calculated.
For the number of simultaneous modifications of the characteristic IDI 8, the change history 4002 is analyzed, and the number of programs simultaneously changed in each modification is calculated.

  The source code 4001 is analyzed, the number of programs called from a certain program is analyzed, and the standard deviation, percentile value, etc. when programs in the same hierarchy are compared are set for the variation in the number of calling programs of the characteristic IDI9.

  In addition, about the characteristic value calculated for every characteristic item, it is good also as a rank of not only calculation value itself but a percentile value or N steps.

FIG. 10 is a diagram showing an operation flow of the improvement location analysis unit of FIG. 3 at the time of specification of a characteristic item.
In FIG. 10, the display device 20 acquires the software structure data 111 and the characteristic data 112 from the system characteristic evaluation device 10 (S901), and displays the software structure 111 arranged according to the characteristic value and the characteristic name 602 (S902) .

  Next, when the user 90 designates the characteristic name 602 (S903), the display device 20 transmits the characteristic ID 601 corresponding to the characteristic item of the characteristic name 602 to the characteristic evaluation unit 102 of the system characteristic evaluation device 10 (S904) .

  The improvement location analysis unit 1023 of the characteristic evaluation unit 102 determines whether the characteristic value of each node stored in the software structure data 1111 exceeds the threshold based on the threshold 603 set in the received characteristic ID 601 ( S 905). By determining whether the characteristic value of each node exceeds the threshold value, it is possible to determine the improvement location of the source code 4001. As shown in FIG. 7, when the threshold 603 is set for the real number 5041 of the characteristic 504, it is determined whether the real number 504 of the characteristic 504 of each node exceeds the threshold 603. It is also good.

  Next, the improvement location analysis unit 1023 merges the nodes that did not become improvement candidates in the analysis of the improvement location, generates a display node and a display edge, and creates the display software structure data 113 (S906). .

  Next, when the display software structure data 113 is created, the characteristic evaluation unit 102 instructs the display device 20 to display the display software structure data 113 (S 907). Then, the display device 20 acquires the display software structure data 113 from the characteristic evaluation unit 102 (S 908), arranges and displays the reconstructed nodes and edges according to the characteristic value (S 909).

FIG. 11 is a flowchart showing the operation of the improvement point analysis unit of FIG.
In FIG. 11, the improvement location analysis unit 1023 operates when the user designates a characteristic item. At this time, the improvement location analysis unit 1023 reads the node information table 1111 and the characteristic table 1121, and acquires a threshold for each characteristic item from the characteristic table 1121 (S1001).

  Next, the improvement location analysis unit 1023 determines, for each node of the node information table 1111, whether the real value 5041 of the characteristic 504 exceeds the threshold 603 for each characteristic item (S1002). In the characteristic table 1121 of FIG. 7, instead of the threshold value 603 for the real number value 5041 of the characteristic 504, a threshold value for the characteristic value 5042 may be set. At this time, it can be determined whether the characteristic value 5042 exceeds the threshold value for each characteristic item.

  Next, the improvement location analysis unit 1023 sorts all nodes into an execution time zone (S1003).

  Next, the improvement location analysis unit 1023 treats the node exceeding the threshold 603 as the improvement candidate as it is. On the other hand, nodes that do not exceed the threshold 603 are considered as non-improved candidates. Then, the improvement location analysis unit 1023 merges the nodes determined as non-improvement candidates, and recalculates the characteristic value of the merged node (S1004). The characteristic value of the merged node can be calculated based on the characteristic value of each node before merging.

  The merge method identifies and merges a plurality of non-improved candidate nodes among the improvement candidate nodes with respect to the nodes sorted in order of execution time in S1003. For example, when the non-improved candidate nodes are continuously arranged in the order of execution time, those non-improved candidate nodes can be merged. On the other hand, when there is an improvement candidate node between non-improved candidate nodes, it is possible not to merge those non-improved candidate nodes. As a result, it is possible to reduce the number of displayed nodes of low priority for improvement without deteriorating the execution order of nodes in the software structure and the relationship between programs, and improve the visibility of the nodes to be improved. it can.

  Also, the characteristic value of the merged node can be set according to the characteristic item. At this time, the characteristic value of the merged node may be the sum of the characteristic values of the pre-merged nodes, or may be the maximum value of the characteristic values of the pre-merged nodes. For example, for nodes that do not have an inclusion relationship, such as processing time and the number of special syntaxes, it is possible to calculate the characteristic value of the merged node by summing the nodes before merging. With regard to the importance of data, the execution time zone, and the like, the maximum value of the pre-merged node can be used as the characteristic value of the merged node. The merge target node may be merged so as not to exceed the characteristic value designated by the user, and may be a plurality of nodes.

  Next, the improvement location analysis unit 1023 sorts the nodes in chronological order, and sets an edge based on the node execution time (S1005).

  Next, the improvement location analysis unit 1023 stores the generated node information and edge information in the display node information table 1131 and the display edge information table 1132 of the display software structure data 113 (S1006).

FIG. 12 is a view showing an example of a display screen of a graph structure constructed based on the node information table of FIG. 5 and the edge information table of FIG.
In FIG. 12, the display screen 1101 can display a software structure including a node 1103 and an edge 1105 and characteristic items 1102 that can be designated by the user 80. In the example of FIG. 12, the processing performance, the operating time zone, the number of programs and the number of special syntaxes are displayed.

  When the software structure is displayed on the display screen 1101, display symbols can be set in the node 1103 and the edge 1105. As this symbol, a figure, a symbol or a character can be used. In the example of FIG. 12, the node 1103 indicating program and data is circled, and the edge 1105 indicating call relationship is indicated by an arrow.

  The node 1103 can be displayed in the execution chronological order 1104 of the first node of the calling relationship. In the example of FIG. 12, the nodes S1 to S5 are executed first, and then the nodes S6 to S14 are executed.

  The node indicating the program or data may change the display method such as color, density, or shape according to the type of node or the size of the characteristic value obtained from the characteristic value calculation unit 1022 data. For example, the node 1103 displays the size of the processing time as the size of the node. By changing the representation of the nodes in this manner, it becomes easy to comprehend a node having a high value among all the nodes.

  In a system in which the shell script has a hierarchical structure in which the program is called, only the relationship between the nodes of the shell script that is the upper hierarchy is displayed, and the calling relationship is expanded according to the specification of the user 80. May be For example, it is assumed that the node S11 is selected on the display screen 1101 when the nodes S1 to S14 of the upper hierarchy are displayed on the display screen 1101. At this time, the hierarchical structure 1107 of the node S11 can be displayed on the display screen 1101.

  In the hierarchical structure 1107 of the node S11, nodes P11 and P12 in the middle hierarchy and nodes D11 to D13 in the lower hierarchy are provided. At this time, the symbols of the nodes S1 to S14, the symbols of the nodes P11 and P12, and the symbols of the nodes D11 to D13 may be made different from each other. In the example of FIG. 12, the nodes P11 and P12 are indicated by squares, and the nodes D11 to D13 are indicated by diamonds.

  Thus, first, the entire software configuration of the business system 30 is confirmed by the nodes S1 to S14 in the upper hierarchy, and then the target is focused on the nodes of interest to the user and expanded into a hierarchical structure, and the detailed program and data relationship It becomes possible to advance the understanding of gender.

FIG. 13 is a diagram showing an example in which the characteristic value of the selected node is displayed on the display screen of FIG.
In FIG. 13, for example, when the node 1103 is pressed on the display screen 1101, a characteristic table 1106 related to the node 1103 is displayed. The characteristic table 1106 displays the characteristic values of the pressed node 1103 for each characteristic item. The characteristic table 1106 can be created from the characteristic 504 of the node information table 1111. Accordingly, the user 80 can confirm the characteristic value of the node 1103 of which the user is interested, among the large number of nodes 1103 displayed on the display screen 1101.

  Further, when the user 80 presses the item name of the characteristic item 1102, the characteristic value of the node 1103 is compared with the threshold value based on the threshold value of the designated characteristic item. At this time, nodes which do not exceed the threshold value are regarded as non-improvement candidates, and nodes which are regarded as non-improvement candidates are merged. Then, the graph structure is reconstructed based on the merged nodes, and the reconstructed graph structure is displayed. This display method is particularly effective when applied to a batch system in which the variation in program execution time is small.

FIG. 14 is a diagram showing an example of a display screen of the graph structure after reconstruction displayed on the display device of FIG.
The display screen 1201 of FIG. 14 shows the case where the processing performance, the operating time zone, and the number of programs are designated as the characteristic items.

  Here, assuming that the processing performance is the processing time, it can be understood from FIG. 5 that the processing time of the node S1 is 7 sec and the processing time of the node S2 is 5 sec. It can be seen from FIG. 7 that the threshold of the processing time is 10 seconds. Therefore, regarding the processing time, it is determined that the processing time of each of the nodes S1 and S2 is less than the threshold. Therefore, the nodes S1 and S2 are regarded as non-improved candidate nodes. Further, the nodes S1 and S2 are continuous in the execution chronological order 1104. As a result, the nodes S1 and S2 are merged, and a node S1 + 2 is newly generated. At this time, the real value of the characteristic of the node S1 + 2 can be the sum of the real value of the characteristics of the original nodes S1 and S2. Also, a characteristic value is given to a real value of the characteristic of node S1 + 2. Then, the characteristic table 1206 of the node S1 + 2 can be displayed by pressing the node S1 + 2.

  On the other hand, it can be understood from FIG. 5 that the processing time of the node S3 is 250 sec. It can be seen from FIG. 7 that the threshold of the processing time is 10 seconds. Therefore, regarding the processing time, the processing time of each node S3 is determined to be equal to or greater than the threshold. Therefore, the node S3 is regarded as an improvement candidate node and treated as it is.

  FIG. 15 is a diagram showing an example of the contents of the node information table corresponding to the graph structure after reconstruction in FIG. The node information table 1131 shows the case where the node S1 + 2 is generated by merging the nodes S1 and S2.

  In FIG. 15, the node information table 1131 includes a node ID 1301 assigned to each program or data, a processing name 1302 capable of identifying a program name or data name for each node, and a type 1303 of program or data. , And store characteristics 1304 of the business system 30 relating to program execution.

  The characteristic 1304 stores at least one or more set characteristic items. As the characteristic 1304, a real number 13041 of the characteristic 1304 and a characteristic value 13042 subjected to normalization or the like can be stored.

  Here, as a result of the improvement location analysis unit 1023 operating, the node information table 1131 stores information on merged nodes among the nodes in FIG. 5 and information on nodes not merged. Also, for the merged node, the characteristic value stores the value calculated again after the merging. FIG. 15 shows an example in which the nodes S1 and S2 of the shell script are stored as the merged node S1 + 2. Also for the processing time of the node S1 + 2, a value obtained by adding up the processing times of the nodes S1 and S2 of the original shell script is stored.

FIG. 16 is a diagram showing an example of the contents of the edge information table corresponding to the graph structure after reconstruction in FIG.
In FIG. 16, by merging the nodes S1 and S2, the edges attached to the nodes S1 and S2 also change. Therefore, the edge information table 1112 of FIG. 6 is updated to the edge information table 1132 of FIG. 16 in accordance with the merge result of the nodes S1 and S2.

  The edge information table 1132 stores an edge ID 1311, an edge call source node ID 1312 and an edge call destination node ID 1313 which can be uniquely identified for each calling relationship between programs or a reference relationship between a program and data shown in FIG. Do. The call source node ID and the call destination node ID use the node ID 1301 stored in the node information table 1131. The node information table 1132 can also store the type of relationship 1314 so that it is possible to distinguish between calls between programs or calls between programs and data.

  By performing the above processing, it is possible to grasp not only the software structure but also the candidate for improvement based on the development status and the operation status, and it is possible to reduce the number of examination man-hours for the innovation target at the time of system renewal.

The present invention is not limited to the above embodiment as it is, and at the implementation stage, the constituent elements can be modified and embodied without departing from the scope of the invention. In addition, various inventions can be formed by appropriate combinations of a plurality of components disclosed in the above-described embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, components in different embodiments may be combined as appropriate.
Although production control has been described as an example in this embodiment, the present invention can be applied to other fields.

DESCRIPTION OF SYMBOLS 10 system characteristic evaluation apparatus, 20 display apparatus, 30 business system, 40 source code management apparatus, 50 failure management apparatus, 101 software structure analysis unit, 102 characteristic evaluation unit, 1011 software structure analysis unit, 1021 characteristic analysis unit, 1022 characteristic value Calculation unit, 1023 improvement location analysis unit, 111 software structure data, 112 characteristic data, 113 display software structure data

Claims (15)

A system characteristic evaluation device,
And a display device,
The system characteristic evaluation device is
A software structure analysis unit that analyzes a relation between a call relationship between programs and data in source code, and generates a graph structure with the program as a node and the call relationship of the program as a directed edge;
A characteristic analysis unit that calculates, for each of the nodes, a characteristic value of the system based on an analysis result of the source code and system information of a system on which the program is executed;
An improvement location analysis unit which merges the nodes based on the characteristic values and reconstructs the graph structure based on the merge result of the nodes;
The display system symbolizes and displays nodes and edges of the graph structure reconstructed by the improvement point analysis unit.   The system characteristic evaluation system according to claim 1, wherein the system information is operation information or failure information of the system. The characteristic analysis unit assigns a characteristic value to each of the operation information, the failure information, the structure information of the program, and the change history of the program.
The system characteristic evaluation system according to claim 2, wherein the improvement point analysis unit merges the nodes based on the characteristic value.   The operation information is a processing time or an execution time zone of the program, the failure information is a failure occurrence frequency of the system, the structure information is the number of programs, the number of special syntaxes, or the data importance, the change history The system characteristics evaluation system according to claim 3, wherein is a program modification frequency or a simultaneous modification number. The characteristic analysis unit calculates the characteristic value for each characteristic item,
The improvement point analysis unit
It is determined for each node whether or not the characteristic value satisfies the condition given for each of the characteristic items,
Merge selected nodes based on the determination result,
The system characteristics evaluation system according to claim 1, wherein an edge of the node is set based on execution time series information of the node. The improvement point analysis unit arranges the nodes in an execution time-series order of the nodes in a graph structure including edges and nodes of the source code.
The system characteristics evaluation system according to claim 1, wherein the display device displays the symbols of the nodes in the execution time series. The improvement point analysis unit
Setting symbols for causing the display device to display the node and the edge on the node and the edge, respectively;
Change the type, color, thickness, size or display position of the symbol according to the characteristic value,
The system characteristics evaluation system according to claim 6, wherein the display device displays a symbol whose type, color, thickness, size or display position has been changed. The improvement point analysis unit changes nodes and edges of the graph structure based on the specification of the characteristic item,
5. The system characteristics evaluation system according to claim 4, wherein the display device displays a symbol reflecting the change of the node and the edge. A source code management device that holds the source code and a change history of the source code;
A business system that holds an application log including operation information of the system;
The system characteristic evaluation system according to claim 2, further comprising: a fault management device that holds fault history reports of the system. A structural analysis unit that generates static information on the source code based on structural analysis of the source code;
A characteristic evaluation unit for acquiring dynamic information about the source code and evaluating characteristics of a system in which a program in the source code is executed based on static information and dynamic information of the source code apparatus.   11. The system characteristic evaluation device according to claim 10, wherein the dynamic information of the source code is operation information or failure information of the system, or a change history of the source code.   12. The system according to claim 11, wherein the operation information is a processing time or an execution time zone of the program, the failure information is a failure occurrence frequency of the system, and the change history is a repair frequency or a simultaneous repair number of the program. System characterization system.   11. The system characteristic evaluation device according to claim 10, wherein the static information of the source code has a graph structure in which a program in the source code is a node and a calling relation of the program is a directed edge. The characteristic evaluation unit
A characteristic analysis unit that calculates, for each of the nodes, a characteristic value of a system in which a program included in the source code is executed, based on static information and dynamic information of the source code;
14. The system characteristic evaluation device according to claim 13, further comprising: an improved location analysis unit that merges the nodes based on the characteristic values and reconstructs the graph structure based on a merge result of the nodes. A threshold is set for each characteristic item of the characteristic of the source code,
The system characteristic evaluation device according to claim 14, wherein the improvement point analysis unit compares the characteristic value with the threshold value for each of the characteristic items, and selects a merge target node based on the comparison result.

Images