JP2005258893A - Response analysis system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a response analysis system which analyzes a part to be adjusted by comparing the response time of processing of a program which works in an information processor with the load of its information processor. <P>SOLUTION: The response analysis system 1, which analyzes the process of the program by analyzing a process information log 2 which outputs a time when the program has been processed by a server unit 21 and a load information log 3 of the server unit 21 in the server unit 21 for performing the process by operating a program, includes an information extracting function 4 for extracting the response time RT of the process from the process information log 2 to generate the response information log 5 aligned in a time series, a response analysis function 6 for bringing the response information log 5 into coincidence with the load information log 3 at a time to analyze, and a display function 7 for displaying simultaneously the response information log 5 and the load information log 3 as graphs G51-G54. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a response analysis system that analyzes a process that causes a decrease in response from response information that is a response time of a program running on an information processing apparatus and load information of the information processing apparatus.

  For example, in an automobile production line, production is performed by assembling parts into a workpiece flowing on a transfer line. However, because many vehicles and their derivatives are produced on the same production line, order information is also included. Since production is performed based on this, it is very important to manage the work flowing through the production line. In order to manage the workpieces transported in this production line, a unique identification number is assigned to each workpiece, the number is recorded on a barcode or an IC tag, and the identification number is assigned at a predetermined location on the transportation line. It is configured to perform management of the workpieces that are read and currently flowing by a computer (assembly line control device).

  In such a production line, in order to improve productivity, the conveyance speed is set so that the workpiece is conveyed at a speed at which an operator can work without waste. For this reason, in the assembly line control device, for example, if the response time of processing for reading and storing the workpiece identification number is delayed, a waiting time is generated for the worker or the assembly robot, and the work time is increased, thereby affecting the entire production line. Will reduce productivity.

  Usually, in such an assembly line control apparatus, a database management system (hereinafter referred to as “DBMS”) is used for management of identification numbers, etc., so that the system can be adjusted by adjusting the setting and inquiry processing of the DBMS. It is possible to improve the response time as a whole (for example, see Patent Document 1).

JP-A-5-2515

  However, in the case of the assembly line control device exemplified here, it is not sufficient that the response time averaged over a predetermined period (for example, one day) is within the threshold value, and any processing is performed in a predetermined manner. There is a restriction that it must be completed in response to time (within threshold). For this reason, there is a problem that the process and the generation time when the response is delayed must be identified, and the cause must be investigated by comparing with the load status of the information processing device (assembly line control device) when the response is delayed. It was.

  The present invention has been made in view of such problems, and is a response analysis system that analyzes a portion to be adjusted by comparing a response time of processing of a program running on an information processing device and a load of the information processing device. The purpose is to provide.

  In order to solve the above problems, a response analysis system according to the present invention analyzes processing of a program in an information processing apparatus (for example, the server apparatus 21 in the embodiment) in which the program operates and performs processing. Processing time counting means (for example, processing time counting function 25 in the embodiment) that measures the processing time of processing executed by the program, and processing time recording means (for example, processing in the embodiment) that stores the measured processing time Information log 2), load information collecting means (for example, load information collecting function 24 in the embodiment) for collecting load information of the information processing apparatus when the program is operating, and the collected load information are stored Extract the response time of the process from the load information recording means (for example, the load information log 3 in the embodiment) and the process time. Then, the information extraction means (for example, the information extraction function 4 in the embodiment) for generating the response information arranged in time series (for example, the response information log 5 in the embodiment), the response information and the load information are made to coincide with each other in time. It is comprised from the response analysis means (For example, the response analysis function 6 in embodiment) which analyzes a process by comparing.

  At this time, the response analysis system according to the present invention is preferably configured to include display means (for example, the display function 7 in the embodiment) that displays the response information and the load information as a graph by matching the time. .

  The response analysis means compares the response information with the load information at the same time, and the load information corresponding to the time when the response information exceeds the predetermined threshold and the response information exceeding the threshold is recorded. It is configured to determine whether or not a predetermined threshold is exceeded, and in this response analysis means, it is determined that the response information exceeds the threshold or the response information exceeds the threshold and the load information exceeds the threshold It is preferable to be configured to have warning means (for example, a warning function 8 in the embodiment) that sometimes issues a warning.

  Furthermore, a selection unit (for example, the selection function 9 in the embodiment) that selects a part of the graph displayed on the display unit, and a response log display unit (for example, a response log display unit that displays response information of the time corresponding to the selected portion) It is preferable to be configured to have the response log display function 10) in the embodiment.

  When the response analysis system according to the present invention is configured as described above, the state of the response time can be compared with the load state of the information processing apparatus when the response time is delayed. It becomes easy. By setting thresholds for response time and load information in advance, and issuing a warning when the threshold is exceeded, the relationship between the response time delay and its cause can be understood more clearly. can do.

  In addition, when a part of the displayed graph is selected, the response information at the corresponding time is displayed, so that it is possible to analyze the processing that requires a response time in more detail.

  Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings. First, an assembly line control system 20 that manages a workpiece 40 that is transported and assembled by the transport line 31 of the production line 30 will be described as a system that performs response time analysis by the response analysis system 1 according to the embodiment. This assembly line control system 20 is used when a workpiece 40 is introduced into the transfer line 31 (for example, check point 32), when assembly is completed (for example, check point 33), or when inspection of a completed vehicle is completed (not shown). 2) using the terminal 22 to read the barcode printed on the management table 41 affixed to the work 40 and manage the vehicle identification number. The vehicle identification number read by the terminal 22 is sent to the server device 21 via a network (factory LAN 23) extending in the factory, and the reading location and reading time etc. are stored in the DBMS managed by the server device 21. Is added and stored and managed (executed as so-called transaction processing).

  In this server device 21, a process in which the vehicle identification number is read by the terminal 22 and stored in the DBMS is executed as a program. This program stores the time as time stamp information before and after the process (for example, a storage device (for example, A processing time count function 25 stored in the hard disk), and the processing time count function 25 stores time stamp information as the processing information log 2. The server device 21 also has a load information collection function 24. When the program is executed by the load information collection function 24, an access load to the storage device and a CPU load for processing. The access load to the memory device and the like are stored as the load information log 3.

  As shown in FIG. 3, the process information log 2 divides the process for storing the vehicle identification number in the DBMS into finer functions constituting the process, and the process time T (the execution of each function is performed for each function). Time of completion) and the vehicle identification number G to be processed are output in time series. Therefore, the response time of each process can be obtained from the process time of the process with the same vehicle identification number G. For example, in FIG. 3, the response time RT of the function S2 can be obtained from the end time of the function S1 and the end time of the function S2 having the same vehicle identification number. Similarly, although the load information log 3 is not shown, the load of each resource (storage device, CPU, memory, etc.) acquired in the server device 1 at predetermined time intervals and the acquired time are paired in time series. It is remembered.

  Then, the processing information log 2 and the load information log 3 stored in this way are transferred to the response analysis system 1 every predetermined period (for example, one day). This transfer can be realized, for example, by connecting the server device 21 and the response analysis system 1 using the network 23 or the like and transferring them by nighttime batch processing (in FIG. 2, response analysis is performed). The connection of the device 21 to the network 23 is omitted).

  Now, a method for analyzing the processing information log 2 and the load information log 3 collected in the server device 21 with the response analysis system 1 and identifying the part with the slow response time and the cause thereof will be described with reference to FIG. The response analysis system 1 extracts necessary information from the processing information log 2, obtains response time information for each process (the response time RT shown in FIG. 3 as described above), and outputs it as a response information log 5. 4, a response analysis function 6 for comparing the data of the response information log 6 and the load information log 3 to analyze the time and cause of the response time delay, and a graph G51 to be described later from the response information log 5 and the load information log 3 A display function 7 that creates G54 and displays it on the display device of the response analysis system 1, and a warning that issues a warning to the display device described above, for example, when the response analysis function 6 finds a time and cause that the response time is delayed And function 8.

  When the time of response time is taken from the response information log 5 extracted by the information extraction function 4 on the horizontal axis and the response time is taken on the vertical axis, it is displayed as a graph G51 shown in FIG. In this response analysis system 1, a threshold value TH1 of the response time required for the system to be analyzed (the assembly line control system 10 in this embodiment) is set, and is displayed over the graph G51. In this way, by displaying the threshold value TH1 and the response time so as to overlap each other on the graph, it is possible to extract a time when the response time is late (for example, time T1 or T4 in the graph G51). On the other hand, in the load information log 3, when the CPU load is plotted on the vertical axis and the time is plotted on the horizontal axis, it is displayed as a graph G52 shown in FIG. Also from this graph G52, the time when the CPU load increases can be easily found (for example, times T1, T2, T3). Similarly, the load on the storage device and the memory can be graphed (G53, G54).

  However, even if these response time graph G51 and load information log 3 graphs G52 to G54 are observed separately, the cause of the response time delay cannot be specified. However, as shown in FIG. The correlation between the two graphs can be easily specified by displaying them side by side so that the times coincide (the same effect can be obtained even if these graphs are displayed in an overlapping manner). In the case of FIG. 4, since the CPU load is also increased at the time T1 when the response time is delayed, there is a high possibility that the CPU load causes a delay in the response time. I understand. At time T4, the response time is delayed, but no change in the load on the graphs G52 to G53 of the load information log 3 is seen, so it is necessary to investigate causes other than the load on the server device 21. Is shown.

  The cause of the slow response time can be easily analyzed by displaying the graph as described above. However, the response analysis apparatus 1 can automatically detect and warn about the cause. For example, thresholds TH2 to TH4 for the load of each resource are provided in advance in each of the graphs G52 to G54 in FIG. 4, and the response analysis function 6 uses the thresholds TH2 to TH4 to make a determination as shown in FIG. I do. Specifically, it is determined whether or not there is a time at which the response time exceeds the threshold value TH1 in step S101. If the response time exceeds the threshold value TH1, the process proceeds to the next step S102. In step S102, it is determined whether or not the load on the CPU exceeds the threshold value TH2 at that time (for example, time T1 in FIG. 4). If the threshold value TH2 is exceeded, the process proceeds to the first warning processing S105, and the response time is set to the threshold value TH1. The warning function 9 shown in FIG. 1 gives a warning that the CPU resource load of the server device 11 has exceeded the threshold value TH2. For example, the warning can be configured to output a message such as a time that occurred in the display device of the response analysis system 1 or a resource with a high load. If the CPU load does not exceed the threshold value TH2 in step S102, it is determined whether or not the load on the storage device exceeds the threshold value TH3 in the same procedure (step S103), and further whether or not the memory load exceeds the threshold value TH4. (Step S104), if the threshold value is exceeded, a warning is given in the first warning process S105. If the response time does not exceed the threshold value TH1 in step S101, no warning is given, and the response time exceeds the threshold value TH1, but the load on the server device 10 does not exceed the threshold values TH2 to TH4 (for example, FIG. 4 at time T4), a warning to that effect is given by the second warning process S106.

  As described above, the response analysis system 1 automatically determines the delay of the response time and the resource load that causes the response time, thereby slowing down the response speed in the analysis target system (for example, the assembly line control system 20). A warning can be issued at the same time, and when there is an abnormality in the load of the server device 21, a warning to that effect can be given. Can be done quickly.

  In the response analysis system 1, when the graph shown in FIG. 4 is displayed on the display device, a point on the graph is selected by the selection function 9 (for example, selected using a mouse or the like) Work for investigating the cause of delaying the response time by displaying the corresponding time and having the response log display function 10 for displaying the corresponding part of the processing information log 2 as shown in FIG. 3 on the display device Can be made easier.

It is a block diagram which shows the response analysis system which concerns on this invention. It is explanatory drawing which shows the analysis object of the response analysis system which concerns on this invention. It is explanatory drawing which shows the structure of a process information log. It is a graph which shows the response time and the load information of a server apparatus. It is a flowchart which shows the process sequence of a judgment function.

Explanation of symbols

1 Response analysis system 2 Processing information log (processing time information recording means)
3 Load information log (load information recording means)
4 Information extraction function (information extraction means)
5 Response information log (response information)
6 Response analysis function (response analysis means)
7 Display function (display means)
8 Warning function (Warning means)
9 Selection function (selection means)
10 Response log display function (response log display means)
21 Server device (information processing device)
24 Load information collection function (load information collection means)
25 Processing time counting function (processing time counting means)
RT response time

Claims (4)

A response analysis system that analyzes the processing of the program in an information processing apparatus that performs processing by running the program,
Processing time counting means for measuring the processing time of processing executed by the program;
Processing time recording means for storing the measured processing time;
Load information collecting means for collecting load information of the information processing apparatus when the program is operating;
Load information recording means for storing the collected load information;
Information extraction means for generating response information arranged in time series by extracting the response time of the process from the processing time;
A response analysis system comprising: response analysis means for analyzing the processing by comparing the response information and the load information with matching time.   The response analysis system according to claim 1, further comprising display means for displaying the response information and the load information as a graph by matching the time. The response analysis means compares the response information with the load information at the same time, and the response information exceeds a predetermined threshold and corresponds to the time when the response information exceeding the threshold is recorded. Configured to determine whether the load information exceeds a predetermined threshold;
A warning means for issuing a warning when the response analysis means determines that the response information exceeds the threshold or the response information exceeds the threshold and the load information exceeds the threshold; The response analysis system according to claim 1, wherein the response analysis system is configured.   A selection unit that selects a part of the graph displayed on the display unit, and a response log display unit that displays the response information at a time corresponding to the selected part, The response analysis system according to claim 2 or 3.

Images