JP2008210135A - Log acquisition method, log acquisition system and log acquisition program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To perform log analysis of a method performed with parallel processing. <P>SOLUTION: In the method, a log is acquired by use of aspect oriented technology. A join point, identifying time, and log acquisition target information are stored as the log in time of a start of a transaction, and the join point, the identifying time, the log acquisition target information and end time are stored as the log in time of an end of the transaction. Thereby, even when the parallel processing is performed, the join point, the identifying time, and the log acquisition target information common to in time of the start and in time of the end are referred, a processing time spent for the transaction can be accurately and easily decided from a time interval between the end time and the identifying time. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a log acquisition method, a log acquisition system, and a log acquisition program. More specifically, the present invention relates to a log acquisition method, a log acquisition system, and a log acquisition program suitable for logging in a system that is constructed using object-oriented technology and requires real-time performance and reliability.

  System development using object-oriented technology is becoming the mainstream of system development.

  For example, when designing a monitoring control system that centrally monitors and controls equipment and resources (hereinafter referred to as monitoring objects) to be monitored and controlled, the monitoring objects are modeled as object classes, Methods that use objects to manage state / measurements are becoming popular. Such a set of object classes is called an “information model”. At the time of construction of the monitoring control system, the “information model” is realized in the form of a software library and used for software (application) that realizes the monitoring control function.

  In a supervisory control system, real-time performance (end processing within the required time) and reliability are very important. For example, the entire system must be executed within 1 second from the start to the end of a certain control process. There are constraints that must be met.

  In a conventional supervisory control system, a software library or the like is constructed by a single vendor according to the hardware to be used in order to realize real-time performance. For this reason, software libraries cannot be used on different hardware and operating systems and are not versatile. However, in recent years, there is a need for common use of software libraries that do not depend on a single vendor. Therefore, it is desired that a software library that implements an “information model” is used in different operating systems, and applications are used by various developers.

  In addition, analysis of logs, which are operation records, is important for system analysis. Especially in systems that require high real-time performance and reliability, such as supervisory control systems, when a problem occurs, the problem location can be identified. Must be performed accurately and without burdening the entire system.

  For example, Patent Document 1 discloses that in log management of a network, a log level and an output destination change code are inserted into each application without stopping the application of each server, and a command is sent to each application. A log management method has been proposed in which changes and log output destination changes are dynamically executed. Patent Document 2 proposes a software operation monitoring apparatus in which a log output mechanism is mounted on a system kernel.

  Furthermore, features that cross between classes that cannot be separated well in recent years by object-oriented are defined as “cross-cutting concerns”, and this “cross-cutting concerns” is separated and described as an aspect to improve program maintainability. Aspect-oriented technology has attracted attention (Non-Patent Document 1). As a programming language for realizing the aspect-oriented technology, for example, AspectJ exists. AspectJ is a programming language in which the Java (registered trademark) language is provided with extensibility for aspect-oriented technology.

JP2004-152024A JP 2006-53788 A "Aspect Oriented Software Development and Tools", Shiba Chiba, Information Processing, Vol. 45, No. 1, pp. 28-33 (January 2004)

  However, if it is necessary to explicitly embed code for log output in units of applications as described in Patent Document 1, every time the system specifications are changed, the source of all programs that acquire logs There is a problem that the code needs to be reviewed and the amount of correction of the source code becomes enormous, so that it cannot be easily handled. In addition, in the process of correction, if there is a coding mistake by the programmer, for example, if code is added to a place where processing should not be added, or if it is forgotten to add code where it should be added, incorrect code will be described In such a case, it may be assumed that a bug occurs in the original program by changing the code in a part not related to logging.

  In the technique described in Patent Document 2, since the log output mechanism is implemented in the system kernel, the appropriate source code for log acquisition is required every time the operating system that implements the entire system is modified or a function is added. There is a problem that the maintainability is low. Further, since it is assumed that the operating system is corrected and the function is added frequently, the frequency of the correction is high, and the maintenance work becomes complicated. Furthermore, since it is necessary to implement for each operating system, there is a problem that the software library in which the information model is implemented cannot be shared. There is also a problem that it cannot be applied to a system using an operating system whose source code is not publicly available.

  For these problems, it is conceivable to improve the maintainability by defining log output processing as an aspect by aspect-oriented technology as in the invention described in Non-Patent Document 1. When the log output process is performed according to the aspect using the technique described in No. 1, there are the following problems.

  When the scale of the system is large, especially in a system that requires high real-time performance, such as a supervisory control system, a plurality of processes are executed simultaneously (parallel processing, multithreading) on a computer. That is, the same method may be called in parallel. However, even if the technique described in Non-Patent Document 1 is used, when such parallel processing is performed, it is not possible to analyze which log each corresponds to, and an appropriate system analysis can be performed. Can not.

  Also, the processing time may vary depending on the order in which the methods are called (call stack). However, even if the technique described in Non-Patent Document 1 is used, a time lag due to a difference in threads calling the same method cannot be acquired as a log.

  Furthermore, acquiring a log for each execution of each method by aspect orientation is a burden on the entire system, and in particular, in a monitoring control system or the like, there is a problem that it becomes a hindrance to real-time characteristics.

  Therefore, the present invention is capable of measuring a time required for a transaction that is processed in parallel, and acquiring a log that can arbitrarily specify a method for recording a log and a sampling period. It is an object to provide a method, a log acquisition system, and a log acquisition program.

  In order to achieve this object, the log acquisition method according to claim 1 is an aspect-oriented method of acquiring a log, and acquires the time at the time as an identification time at the start of a transaction to be acquired. Then, at least the process of storing the identification time in association with the join point at the time and the process of storing the log acquisition target information of the transaction is performed, and the join point, the identification time and the log are used as a log at the start of the transaction. Stores acquisition target information, and at the end of the transaction, acquires the identification time using the join point at that time as a key and stores them, processing to store the log acquisition target information of the transaction, and time at that time At least the process of acquiring the end time and storing the end time You, as a log at the end of the transaction, join point, identification time, so that to save the log acquisition target information and end time.

  The log acquisition system according to claim 7 is a system for acquiring a log by using aspect-oriented technology, and at the start of a transaction to be acquired by the log, the time at the time is read from the timer and identified. As a log at the start of a transaction, at least a process for storing the identification time in association with the join point at the time and storing the log acquisition target information of the transaction in the storage apparatus is executed. Processing for storing the join point, identification time and log acquisition target information in the log recording unit, and at the end of the transaction, reading the identification time using the join point at that time as a key and storing them in the storage device, Processing for storing transaction log acquisition target information in a storage device, and The time at the point is read from the timer as the end time, and at least the process of storing the end time in the storage device is executed, and as a log at the end of the transaction, the join point, identification time, log acquisition target information, and end time Is provided with a log acquisition unit that executes a process of storing the information in the log recording unit.

  The log acquisition program according to claim 8 is a program for acquiring a log by using aspect-oriented technology, and at the start of a transaction to be acquired by a log, the time at the time is read from the timer and identified. Process for storing the identification time in the storage device in association with the join point at the time, processing for storing the log acquisition target information of the transaction in the storage device, and log at the start of the transaction as a join point, identification Processing for storing the time and log acquisition target information in the log recording unit, processing for reading the identification time using the join point at the time as a key and storing them in the storage device at the end of the transaction, transaction log acquisition Processing to store the target information in the storage device, at the time Is read from the timer as the end time, the process of storing the end time in the storage device and the log at the end of the transaction, the join point, identification time, log acquisition target information, and end time are stored in the log recording unit The processing is executed by a computer.

  Therefore, log acquisition using aspect-oriented technology, at the start of the transaction, the join point, identification time and log acquisition target information are saved as a log, and at the end of the transaction, the join point, identification time, and log acquisition By storing the target information and end time as a log, the system operator can set the common join point, identification time, and log acquisition target information at the start and end even when parallel processing is performed. For reference, it is possible to accurately and easily determine the processing time required for a certain transaction from the time interval between the identification time and the end time. The join point refers to a place in the code where the advice (processing contents, in this case, log acquisition) is targeted in the aspect-oriented technology. Further, in this specification, a transaction refers to the entire processing that is the target of log acquisition, and includes a method and a later-described constructor call, a data display function, and the like.

  The invention according to claim 2 is the log acquisition method according to claim 1, wherein at the start of the transaction, the processing for storing the identification time in association with the join point at the time and the transaction log acquisition target information In addition, after the process of storing is performed, a process of acquiring and storing the time at the time as the start time is performed and stored as a log at the start of the transaction. Therefore, the start time is acquired separately from the identification time immediately before the start of the transaction, and stored as a log at the start.

  Further, the invention according to claim 3 is the log acquisition method according to claim 1 or 2, further storing a call stack indicating in what order the methods are called at the start of the transaction, It is saved as a log at the start of a transaction.

  Therefore, by acquiring the call stack, it is possible to save as a log how many seconds it took at each call position even if the processing time of the method differs depending on the order in which the same method is called.

  According to a fourth aspect of the present invention, in the log acquisition method according to any one of the first to third aspects, an object class and a method that are targets of log acquisition can be arbitrarily set as parameters in advance. Therefore, since the aspect-oriented technology is used, it is possible to change the class and method for which the log is acquired even during operation of the system.

  According to a fifth aspect of the present invention, in the log acquisition method according to any one of the first to fourth aspects, at the start of a transaction, a counter value is incremented, and a preset sampling period setting value is set. The log is acquired only when the counter value matches.

  Therefore, the log acquisition frequency can be arbitrarily set by the set value of the sampling period input in advance as a parameter.

  According to a sixth aspect of the present invention, in the log acquisition method according to the sixth aspect, the log acquisition method according to any one of the first to fifth aspects is applied to a software library used for a monitoring control application that implements a monitoring control function in the monitoring control system. I am trying to apply it.

  Therefore, the code for log acquisition is described as source code different from the “information model” that implements the software library by aspect-oriented technology, and the source code for log acquisition is also used when modifying the “information model” There is no need to change.

  According to the log acquisition method according to claim 1, the log acquisition system according to claim 7, and the log acquisition program according to claim 8, even if a transaction is processed in parallel, a certain transaction is required. The time can be individually identified. For example, in parallel processing, when method 1 is activated in thread A and method 1 is terminated in thread B immediately after that, it is different from the actual situation if it is not properly grasped which transaction is being measured. Although a short time is measured, since the time required for processing each method can be individually identified, such a problem can be avoided. Also, by using the aspect, it is only necessary to define the original processing for each object class, and software maintainability can be improved.

  In addition, according to the log acquisition method according to claim 2, by acquiring the start time again immediately after the acquisition of the identification time and immediately before the start of the transaction, the processing time can be further increased from the start time and the end time. It becomes possible to measure accurately.

  According to the log acquisition method of claim 3, even if the processing time differs depending on which thread the same method is called, by referring to the call stack, the number of seconds at each call position It can be judged whether it took. For example, the processing time of a method can be recorded as a log for each thread that calls the target method.

  Further, according to the log acquisition method of the fourth aspect, it is possible to arbitrarily change the class and the method for which the log is acquired even during the actual operation of the system. Furthermore, the ability to select the log acquisition target arbitrarily allows the operator to select the process that needs to acquire the log as appropriate, reducing the processing load for log acquisition in the entire system and minimizing overhead time. Can be achieved.

  Further, according to the log acquisition method of the fifth aspect, it is possible to arbitrarily set the frequency of log acquisition. As a result, it is possible to acquire a log according to an operator's request. In addition, it is possible to reduce the processing load for log acquisition in the entire system and to minimize the overhead time.

  According to the log acquisition method described in claim 6, the log acquisition in the monitoring control library used by various system developers affects the application that executes the monitoring control processing and the monitoring control library. Therefore, maintainability of the entire monitoring control system can be improved.

  Hereinafter, the configuration of the present invention will be described in detail based on embodiments shown in the drawings.

  The log acquisition method according to the present invention is an aspect-oriented method of acquiring a log. At the start of a transaction to be acquired, the time at the time is acquired as an identification time, and the identification time is determined at the time. At least the process of storing the information in relation to the join point and the process of storing the log acquisition target information of the transaction, and storing the join point, the identification time and the log acquisition target information as a log at the start of the transaction, At the end, the process of acquiring identification times using the join point at the time as a key and storing them, the process of storing the transaction log acquisition target information, and the time at the time as the end time At least when processing to memorize the time, at the end of the transaction As a log, join points, identification time, so that to store the log acquisition target information and end time. Alternatively, the transaction processing time may be measured more accurately by measuring the start time separately from the identification time immediately before the start of the transaction for which the log is to be acquired. Furthermore, as information acquired as a log, necessary information can be appropriately acquired in addition to the above-described items according to the necessity of business.

  The system to which the log acquisition method, the log acquisition system, and the log acquisition program of the present invention are applied is not particularly limited as long as it is a system designed by object-oriented technology. In this embodiment, the monitoring and control system of the water reclamation center (sewage treatment plant) (reference documents: Nakamura, Yoshida, Takahashi, Otani, Shimada “Applicability of general-purpose technology to the water reclamation center monitoring and control system through the construction of an experimental system” Evaluation ”, 2005 Institute of Electrical Engineers of Japan Industrial Application Division, pp.II-265 to II268, August 2005)), the case of applying the log acquisition method, log acquisition system and log acquisition program of the present invention will be described. .

  The configuration of the water reclamation center monitoring control system 1 in the present embodiment is, for example, a hardware configuration as shown in FIG.

  The monitoring control system 1 according to this embodiment includes a central monitoring control facility 2, a process controller 3, and a sequencer 4. Hereinafter, the process controller 3 and the sequencer 4 are collectively referred to as a local control facility 6.

  The central monitoring control facility 2 includes an output device 20 such as a display as shown in FIG. 2, an input device 21 such as a keyboard and a mouse, a central processing unit (CPU) 22 that performs arithmetic processing, and data being calculated. A workstation having a main storage device (RAM) 23 in which parameters and the like are stored, a hard disk as an auxiliary storage device 24 in which each database and calculation results are recorded, a communication interface 25 for communicating with the outside, etc. A dedicated computer is used. Hereinafter, the main storage device 23 and the auxiliary storage device 24 are collectively referred to simply as a storage device. The hardware resources are electrically connected through a bus 26, for example.

  Further, as shown in FIG. 3, the central supervisory control device 2 includes a supervisory control library 17 and supervisory control applications (supervisory control APP 1 to 3) 16. The programs for realizing the monitoring control library 17 and the monitoring control application 16 are recorded in the auxiliary storage device 24, and these are read and executed by the CPU 22 so that the workstation and the dedicated computer can be centrally monitored and controlled. Functions as 2. Data, parameters, and the like necessary for the execution are loaded into the RAM 23.

  The monitoring control library 17 is used in various monitoring control applications 16. The monitoring control library 17 stores, for example, basic data 15 (not shown) necessary for processing of the monitoring control application 16, and monitoring data 18 and control with the monitoring target device 5 via the local control facility 6. Responsible for exchanging data 19. Note that the basic data 15 is, for example, monitoring data 18 obtained from the site, setting values input by the operator 9, and the like.

  The monitoring control application 16 provides a user interface for the operator 9 such as displaying a monitoring result on the output device 20 and processes the monitoring data 18 and the control data 19. In this embodiment, the monitoring control application 16 and the monitoring control library 17 are mounted on the same hardware, but may be realized in a distributed computer environment.

  In the present embodiment, the central monitoring control device 2 functions as the log acquisition system 10. The log acquisition system 10 exchanges the monitoring data 18 and the control data 19 in the monitoring control library 17, saves and acquires the basic data 15, outputs the basic data 15 to a file, and performs monitoring via the monitoring control application 16. The log 8 is acquired and output for transactions related to the display of the data 18 and the like. As described above, the log acquisition system 10 can acquire data useful for maintenance of the library itself, separately from the monitoring control function normally executed by the central monitoring control device 2.

  An example of the functional configuration of the log acquisition system 10 of the present embodiment is shown in FIG. The log acquisition system 10 in this embodiment includes a log acquisition unit 11 that executes log acquisition processing, a program storage unit 12 that stores a program for log acquisition, and a log storage unit 13 that stores the created log 8 as a log file. The display control unit 14 displays the log on the output device 20.

  The process controller 3 is mainly a hardware device developed exclusively for at least an arithmetic unit (CPU), a storage device, a communication interface, and the like.

  The process controller 3 performs calculations such as integral calculation and differential calculation necessary for feedback control based on the data sent from the sequencer 4, and sends a command based on the calculation result to the sequencer 4 and the central monitoring control equipment 2. A function for setting control calculation parameters based on the received command values and sending an operation command for the device to the sequencer 4 is provided. In addition, the calculation required when controlling each monitoring object apparatus 5a-5d is not performed by the central monitoring control equipment 2, but is often performed by the process controller 3.

  For example, a programmable logic controller (PLC) is used for the sequencer 4, and for example, a pump 5a, a valve 5b, a conveyor 5c, a blower 5d, and the like are connected to each of the sequencers 4a to 4d.

  The sequencer 4 receives a signal indicating the analog value or the digital value from each of the monitoring target devices 5a to 5d, a command from the process controller 3, and receives the set logic and the command or device from the process controller 3. The sequence control is performed based on the above signal, and the function of transmitting the signal to each of the monitoring target devices 5a to 5d and the function of transmitting the monitoring data to the process controller are provided. For example, a remote input / output device is used for transmission / reception of signals to / from the monitoring target devices 5a to 5d.

  The monitoring data 18 sent from the local control facility 6 (for example, PLC) to the central monitoring control facility 2 is classified into an analog value (measured value, continuous value) and a digital value (on / off information, 0-1 value). Each record of the monitoring data 18 is given a unique ID (device ID) to each device, and is automatically transmitted to each device at a predetermined time interval. At this time, the correspondence relationship between each device ID and the device may be stored in advance in the auxiliary storage device 24 as an initial setting file.

  The analog value of the monitoring data 18 includes, for example, a water level (difference from the reference water level) (m), pump output (%), valve opening / closing degree (%), current value (A), and the like.

  On the other hand, the digital value of the monitoring data 18 includes, for example, the valve closing status (ON = fully closed, OFF = open), the status of the switchgear in the power receiving / transforming equipment (ON = turned on, OFF = opened), and the pump operation. There are situations (ON = running, OFF = stopped), various alarms (presence / absence of occurrence of ground fault, etc.).

  Similarly, the control data 19 transmitted from the central supervisory control facility 2 to the local control facility 6 can be roughly divided into analog values and digital values.

  The analog value of the control data 19 includes, for example, a pump output set value (%), a valve opening set value (%), a medicine injection rate (%), and the like, and the digital value includes, for example, a pump, a blower, etc. There are operation commands (ON = operation, OFF = stop), valve opening / closing commands (ON = close, OFF = open), etc. for each device.

  The hardware configurations of the monitoring control system 1 and the log acquisition system 10 according to the present embodiment have been described above. However, the hardware configurations are merely examples, and the present invention is not limited thereto. The communication protocol for exchanging data such as control data and monitoring data between the monitoring control library 17 and the local control facility 6 is not particularly limited. In this embodiment, TCP / UDP / IP / Ethernet (registered trademark) is used as a communication protocol below the transport layer and uniquely defined above the session layer, but the communication protocol is particularly limited. Instead, a new or known combination may be used.

  FIG. 5 is a diagram schematically showing how an execution code used in the log acquisition system 10 of the present embodiment is generated. The log acquisition system 10 separately defines object classes 30 and 31 constituting the main body of the monitoring control library 17 and an aspect 32 that defines a log output function, and generates an execution code 37 through an aspect compiler 33.

  The object class A30 acquires the measurement value of the monitoring target device (pump 5a, valve 5b, etc.), and transmits the measurement value when the measurement value exceeds a preset threshold value. For example, when the measured value of the output of a certain pump exceeds a threshold value (120%), the measured value is transmitted as an abnormality. The object class B31 acquires the ON / OFF state of the monitoring target device and transmits it. Note that the object classes 30 and 31 are examples of object classes that may perform log acquisition in the monitoring control library 17 and are not particularly limited.

  For these object classes, an aspect-oriented technique is used to describe a log acquisition program, which will be described later as aspect 32, and these are compiled as aspect 32 by compiling them with aspect compiler 33 (for example, ajc in aspectJ). The log acquisition program is incorporated in the processing of both the object class A30 and the object class B31. These are called intermediate code A34 and intermediate code B35. Further, the execution code 37 is generated by compiling the intermediate code A34 and the intermediate code B35 by the language compiler 36.

  In the present embodiment, the aspect is incorporated into all object classes that are likely to acquire logs by the above-described method. As a result, it is possible to automatically incorporate the log acquisition function according to one aspect for all object classes that may perform log acquisition.

  Below, the process which the log acquisition part 11 of this embodiment performs is demonstrated. Each process described below is executed by reading a log acquisition program stored in the program storage unit 12 and executing it.

  FIG. 7 is a flowchart illustrating an example of a log acquisition process at the start of a transaction.

  First, it is determined whether or not the log acquisition process start condition shown in FIG. 6 is satisfied (S101). If it is satisfied (S101: Yes), whether or not the sampling cycle shown in the flowchart of FIG. (S102: Yes), the current time is acquired as the identification time (S103). The method for acquiring the identification time is not particularly limited, but in the present embodiment, it is acquired from a timer (not shown) of the central monitoring control facility 2.

  On the other hand, when the log acquisition process start condition is not satisfied (S101: No) or when the sampling cycle is not reached (S102: No), the log acquisition process ends and no log is acquired.

  FIG. 6 shows an example of the log acquisition process start condition. In the present embodiment, for example, all of the “transaction log acquisition mode”, “transaction log acquisition target class”, and “transaction log acquisition target method” are satisfied, and “method The start condition is that any one of “execution”, “calling a constructor”, or “calling a display function” is satisfied. Note that the log acquisition process start condition shown in FIG. 6 is an example, and log acquisition may be performed according to the necessity of the monitoring control work.

  In the log acquisition system 10 of the present embodiment, the mode indicating whether or not a transaction log is acquired is a parameter that can be appropriately selected by the operator 9. Similarly, for the object class (acquisition target class) and method (acquisition target method) to be acquired by the log, similarly, the parameter that can be appropriately selected by the operator 9 is used, and there is a possibility of performing log acquisition incorporating an aspect. Any object class can be arbitrarily selected.

  In this embodiment, these modes, acquisition target classes, and acquisition target methods can be changed even during library operation. For example, a mode indicating whether or not a transaction log is acquired is stored as a boolean variable, and the value is referred to every time an aspect is executed. By allowing the operator 9 to arbitrarily set this value from the operation screen on the output device 20, it can be changed even during library operation.

  Also, whether or not it is the acquisition target class or acquisition target method, for example, by checking whether the class name or method name obtained from thisJoinPoint is registered in the set that manages the log acquisition target class, if true Log acquisition is executed, and if it is false, the log acquisition process is terminated. Since the operator 9 can arbitrarily set the registration and deletion of the class name or method name in this set from the operation screen on the output device 20, it can be changed during the library operation.

  In addition, the method in “execution of method” refers to a method possessed by an object designed for the monitoring control library 17. For example, when implemented using Java (registered trademark), Java (registered) Trademark) The methods provided by the original object are not targeted.

  In addition, the constructor in “calling a constructor” is a process that is called when a new object is generated and performs initialization of contents, and is designed for the monitoring control library 17 in this embodiment. This is one of the start conditions for measuring the time required for generating an object.

  In addition, regarding “calling of display function”, in this embodiment, how the display function is implemented is out of the scope of the monitoring control library 17, and for example, for display in Java (registered trademark). When calling a method that redraws an object, it is set to return a true value.

  Next, a pair having the current join point as a key and the time for identification as a value is stored (S104). A join point refers to a position where an aspect is called in aspect-oriented technology, and has a value uniquely determined for each position. In this embodiment, a pair of join point hash code and identification time is stored.

  Next, log acquisition target information is acquired (S105). The log acquisition target information refers to, for example, a class name, a method name, and a monitoring target name that are targets for log acquisition. For example, when a sedimentation basin (“sedimentation basin No. 1”), which is a monitoring control object, is represented by a Tank object and its water level is recorded by the setMeasurement () method, the log acquisition target As information, it is acquired that it is a Tank object, a method called setMeasurement (), and that the monitoring control object is named “Sedimentation basin No. 1”.

  Next, the call stack is acquired (S106). The call stack is the order in which the methods are read, and stores the method that called the processing target method as the call stack. Specifically, for example, in the case of Java (registered trademark), it is possible to acquire a call stack at the time of generation by generating a Throwable class. Furthermore, by using the Throwable.getStackTrace () method, it is possible to decompose it into elements such as combinations of class names and method names constituting the stack. In the present embodiment, for example, by acquiring a character string from this element, the call stack to be finally saved in the log is acquired.

  Next, the current time is acquired as the start time (S107). Although the acquisition method of a start time is not specifically limited, It acquires from the timer (not shown) of the central monitoring control equipment 2 similarly to the time for identification. In this embodiment, the start time is stored separately from the identification time immediately before the start of the transaction that is the target of log acquisition, thereby enabling more accurate transaction processing time measurement. It is also possible to simplify the processing by treating the identification time as the start time without acquiring the start time.

  Next, the log acquisition target information, identification data, call stack, and start time are stored (S108), and the log acquisition process at the start of the transaction ends. The identification data refers to a combination of the method name acquired in S105 and a pair having the identification time as a value using the current join point acquired in S104 as a key. By storing this identification data, it is possible to determine which access is made when analyzing the log even if parallel processing is performed and accesses to the same method occur almost simultaneously. Become. The log acquisition target information, identification data, call stack, and start time acquisition order shown in FIG. 7 is an example and is not limited to this. Moreover, the information acquired as a log at the time of the above-mentioned start is an example, and what is necessary is just to acquire required information suitably according to the necessity of work.

  FIG. 8 shows an example of the sampling determination process (S102) for determining whether it is the set sampling timing in the log acquisition process.

  In the present embodiment, it is possible to set a value (sampling period) indicating how many times the log is stored once (for example, once in 100) for the same method call. The sampling period is a parameter that can be arbitrarily set by the operator 9, and the setting value of the sampling period read out on the memory is compared with a counter indicating the number of processing times. This makes it possible to perform one log acquisition process when executing the method corresponding to the value set as the sampling period.

  In this way, by making it possible to arbitrarily set the frequency of log acquisition, for example, usually set a large number of samplings and set a small number of samplings only when a detailed test is desired. Thus, operation such as increasing the frequency of log acquisition becomes possible. In addition, acquiring a log for each execution of each method places a load on the entire system in a process different from the original purpose of the system (for example, monitoring control process), leading to an increase in processing amount and processing time. Therefore, the amount of log acquisition processing in the entire system can be reduced as compared with the case where log acquisition processing is performed every time, so that overhead time can be minimized.

  In the sampling determination process, first, the counter value is incremented by 1 (S201), and it is determined whether the sampling period and the counter value match (S202).

  If the sampling period and the counter value match (S202: Yes), the target join point is updated to the current join point (S203), and the target thread is further updated to the currently executing thread (S203). S204) and the process proceeds to S205. On the other hand, if the sampling period and the counter value do not match (S202: No), the process proceeds to S205.

  In step S205, it is determined whether the target join point is not empty and the target thread is a thread calling this procedure (S205). Specifically, it is determined whether the join point is not a null value and whether the object representing the target thread matches.

  When the condition of S205 is satisfied (S205: Yes), true is returned (S206), and when not applicable (S205: No), false is returned (S207), and the sampling determination process ends. In other words, in this process, the true value is the set sampling timing, and the false value is the return value otherwise.

  FIG. 9 is a flowchart illustrating an example of a log acquisition process at the end of a transaction.

  First, it is determined whether or not the log acquisition process start condition is satisfied (S301). If it is satisfied (S301: Yes), it is determined whether or not the current join point is the same as the join point stored at the start (S302). . If they are the same (S302: Yes), the sampling data (counter, etc.) are reset (S303). On the other hand, when the log acquisition process end condition is not satisfied (S301: No), or when the current join point does not match the join point stored at the start (S302: No), the log acquisition process at the end of the transaction ends. .

  Next, the identification time corresponding to the current join point is acquired (S304). Here, the identification time is acquired from the “pair having the current join point as a key and the identification time as a value” stored in S104. As a result, even when the same method is called in parallel, the time can be measured independently.

  Here, it is determined whether or not the identification time is empty, and if it is empty (S305: Yes), the log acquisition process ends as invalid data.

  On the other hand, when there is an identification time (S305: No), the log acquisition target information is acquired (S306). Note that the content of the process is the same as that in S105, and thus the description thereof is omitted.

  Next, an end time is acquired (S307). The acquisition method of the end time is not particularly limited, but is acquired from a timer (not shown) of the central monitoring and control facility 2 in the same manner as the identification time.

  Finally, log acquisition target information, identification data, call stack and end time are stored (S308), and the process ends. As the identification data, the combination of the method name acquired in S306 and the pair having the identification time as a value using the current join point acquired in S304 as a key is stored as in S108. Note that the acquisition order of the log acquisition target information, the identification data, and the end time illustrated in FIG. 9 is an example and is not limited thereto. Moreover, the information acquired as the above-mentioned log at the end time is an example, and necessary information may be acquired as appropriate according to the necessity of business.

  The log acquisition target information, identification data, call stack, and start time acquired by the log acquisition process described above are the log at the start of the process, and the log acquisition target information, identification data, and end time are the log at the end of the process And stored in the log file of the log storage unit 13. In addition, the display control unit 14 can output a log to the output device 20.

  As a result, the operator 9 can analyze the log by displaying, printing, and the like of the log file. According to the log 8 provided by the log acquisition system 10 of the present embodiment, the operator 9 determines which access is made based on the identification data even when a plurality of the same transactions are executed in parallel. Is possible.

  In general, in developing a log acquisition system, in order to realize special processing, many parameters are used and the processing procedure tends to be complicated. According to the log acquisition method of the present invention, As described above, the processing procedure itself and the types of parameters to be used can be kept to a minimum, and the overhead of the log acquisition processing can be minimized.

  The above-described embodiment is a preferred embodiment of the present invention, but is not limited thereto, and various modifications can be made without departing from the gist of the present invention.

It is a figure which shows an example of the hardware constitutions of a monitoring control system. It is a figure which shows an example of the hardware constitutions of a central monitoring control apparatus. It is a figure which shows the relationship between the software library in a central supervisory control apparatus, and the supervisory control application. It is a figure which shows an example of a function structure of a log acquisition system. It is the figure which showed the image at the time of compiling the aspect which defined the object class and log acquisition which comprise a monitoring control library, and producing | generating an execution code. It is a figure which shows an example of log acquisition process start conditions. It is a flowchart which shows an example of the log acquisition process at the time of transaction start. It is a flowchart which shows an example of a sampling determination process. It is a flowchart which shows an example of the log acquisition process at the time of transaction end.

Explanation of symbols

1 Monitoring and Control System 10 Log Acquisition System 11 Log Acquisition Unit 13 Log Storage Unit

Claims (8)

  In the aspect-oriented method of acquiring a log, at the start of a transaction for which a log is to be acquired, the time at the time is acquired as an identification time, and the identification time is stored in association with the join point at the time And at least a process of storing log acquisition target information of the transaction, and storing the join point, the identification time, and the log acquisition target information as a log at the start of the transaction, and ending the transaction Sometimes, the process of acquiring the identification time using the join point at the time as a key and storing them, the process of storing the log acquisition target information of the transaction and the time at the time as the end time, At least a process of storing the end time is performed, and the traffic is As at the end of log-transactions, the join points, the identification time, log acquisition method characterized by storing the log acquisition target information and the end time.   After performing the process of storing the identification time in association with the join point at the time and the process of storing the log acquisition target information of the transaction at the start of the transaction, the time at the time is further set to the start time The log acquisition method according to claim 1, wherein the log acquisition method is stored as a log at the start of the transaction.   The log acquisition method according to claim 1 or 2, wherein a call stack indicating in what order the methods are called at the start of the transaction is also stored and saved as a log at the start of the transaction. .   The log acquisition method according to any one of claims 1 to 3, wherein an object class and a method for acquiring the log can be arbitrarily set as parameters in advance.   5. The counter is incremented at the start of the transaction, and a log is acquired only when a preset value of a sampling period matches the value of the counter. The log acquisition method according to any one of the above.   6. A log acquisition method according to claim 1, wherein the log acquisition method according to claim 1 is applied to a software library used for a monitoring control application for realizing a monitoring control function in the monitoring control system.   A system for acquiring a log using aspect-oriented technology, and at the start of a transaction for which a log is to be acquired, the time at the time is read from a timer as an identification time, and the identification time is At least a process of storing in the storage device in association with the join point and a process of storing the log acquisition target information of the transaction in the storage device, and as the log at the start of the transaction, the join point, the identification time and Processing for storing the log acquisition target information in a log recording unit and processing for reading the identification time using the join point at the time as a key and storing them in a storage device at the end of the transaction, log acquisition of the transaction Processing to store the target information in the storage device and at the time Read the time from the timer as the end time, execute at least the process of storing the end time in the storage device, and as the log at the end of the transaction, the join point, the identification time, the log acquisition target information, A log acquisition system comprising a log acquisition unit that executes a process of storing the end time in a log recording unit.   A program for acquiring a log using aspect-oriented technology, and at the start of a transaction for which a log is to be acquired, the time at the time is read from a timer as an identification time, and the identification time is set at the time The process of storing in the storage device in association with the join point, the process of storing the log acquisition target information of the transaction in the storage device, and the log at the start of the transaction, the join point, the identification time, and the log acquisition target information The log recording unit, the process of reading the identification time using the join point at the time as a key and storing them in the storage device at the end of the transaction, and storing the log acquisition target information of the transaction Processing stored in the device, time at the time A log recording unit that reads from the timer as an end time, stores the end time in a storage device, and logs at the end of the transaction, the join point, the identification time, the log acquisition target information, and the end time A log acquisition program for causing a computer to execute processing to be stored in the computer.