JP2008003945A - Monitoring control system and computer management method and program thereof - Google Patents

Abstract

The present invention realizes stable operation of a system by reliably detecting occurrence of a memory leak during system operation.
A current free memory capacity of a computer main body is read (S201), the current free memory capacity is equal to or less than the previous value (NO in S202, NO in S203), and a count value of a change counter is set in advance. If it is greater than the specified value (NO in S204), it is determined that memory leak has been detected, and an alarm signal indicating memory leak detection is output (S205). When the count value of the change counter is less than or equal to the specified value (NO in S204), the change counter is counted up (S206), and when the current free memory capacity is larger than the previous value (NO in S202), the change The counter is cleared (S207). The current free memory capacity is recorded as the previous value (S208), the processing is stopped for 24 hours (S209), and then the memory leak detection process is resumed at the same time on the next day.
[Selection] Figure 2

Description

  The present invention relates to a technique for ensuring stable operation of a system in a monitoring control system that performs monitoring control of a target system.

  2. Description of the Related Art Conventionally, in a monitoring control system that performs monitoring control of various systems, a main control device is configured by a computer, and various processes are realized using programs according to the target system. In such a supervisory control system, from the viewpoint of enhancing the reliability of the supervisory control system itself, a computer constituting the main controller incorporates an abnormality monitoring device that detects internal abnormality of the computer main body by hardware. Exists (see, for example, Patent Document 1).

Specifically, recent industrial computers used in industrial monitoring and control systems are equipped with abnormality monitoring devices collectively referred to as RAS functions in order to detect internal abnormalities in the computer body. Here, the RAS function is a general term for functions intended to improve the reliability, availability, and serviceability of a computer system. As such a RAS function, for example, a watch dog timer (WDT) function for detecting a program operation disabled state is known.
JP2004-334548

  By the way, although the program operates using the primary storage area (hereinafter referred to as memory) mounted on the computer, it can be used without releasing the memory dynamically reserved due to inadequacies in the program or unexpected operations. Memory capacity may be insufficient. This phenomenon is generally called a memory leak.

  When this memory leak occurs, not only the operation of a specific program stops, but also the operation of the program of the entire computer system including the OS may stop. For this reason, if this memory leak occurs in the computer used in the supervisory control system, there is a possibility that the original system supervisory control cannot be performed.

  On the other hand, in the field of computer control system programming, compiler manufacturers, various debugger program development manufacturers, and the like have previously announced a mechanism that does not cause a memory leak for each program, a logic assembly method, and a debugging method. However, even if a program is created with care so as not to cause memory leaks, memory leaks may occur after system operation without the occurrence of memory leaks being detected by program verification.

  In other words, the memory leak is a problem caused by the completeness of each program, and it is a human factor non-conformity that can be prevented if sufficient program verification is performed, but in practice, sufficient program verification is generally performed. It is difficult to implement.

  First, the program developer verifies that various problems including memory leaks do not occur in the verification of the entire system including the program developed by himself, but in reality, it bypasses the logic of all abnormal processing. In many cases, the inspection cannot be performed, or all the conditions cannot be set at the logic branch point, and verification by the developer's own visual visual inspection is often performed.

  Some computer control systems operate continuously for 24 hours 365 days. In such a system, since continuous operation for a long time far exceeding the time spent for program verification is required, it is necessary to verify the occurrence of memory leaks due to long time as real time instead of calendar operation. Are often difficult in practice.

  In addition, recently, almost no system is configured only by a developer's own program, and a combination with a purchased program including an OS is forced. In a system in which such a plurality of programs are combined, it is difficult to verify the occurrence of a memory leak only by an operation test.

  As described above, since it is difficult to reliably verify the occurrence of a memory leak at the time of program verification, even if the computer of the supervisory control system cannot perform sufficient program verification, a memory leak will occur after system operation. May occur, and the system monitoring control that is the original purpose may not be performed.

  On the other hand, the watchdog timer (WDT) function, which is known as one of the RAS functions, only detects that the target program has become inoperable and cannot respond within a specific time. Therefore, it cannot be distinguished whether the program operation disabled state is due to the occurrence of a memory leak or other program abnormality.

  The present invention has been proposed in order to solve the above-described problems of the prior art. The object of the present invention is to make it possible to reliably detect the occurrence of a memory leak during system operation, thereby ensuring stable system operation. To provide a supervisory control system and a computer management method and program thereof.

  In order to achieve the above-mentioned object, the present invention monitors the memory change state of the computer main body, detects a memory leak, and outputs an alarm, thereby reliably detecting the occurrence of the memory leak during system operation. The system operation manager can be alerted.

  That is, the computer management method of the monitoring control system of the present invention is a computer of a monitoring control system including a computer main body constituting a main control device that performs monitoring control of a target system and an abnormality monitoring device that detects an internal abnormality of the computer main body. In the management method, the abnormality monitoring device monitors the memory change state of the computer main body, and performs memory leak detection processing for outputting an alarm indicating memory leak detection when a memory leak is detected.

  The monitoring control system and the computer management program of the present invention grasp the features of the computer management method from the viewpoints of the system and the program.

  According to the present invention, it is possible to reliably detect the occurrence of a memory leak during system operation, and therefore it is possible to provide a monitoring control system that realizes stable system operation, a computer management method thereof, and a program.

[First Embodiment]
[System configuration]
FIG. 1 is a block diagram showing a first embodiment of a supervisory control system to which the present invention is applied. As shown in FIG. 1, the monitoring control system 100 according to the present embodiment includes a computer main body 110 that constitutes a main control device that performs monitoring control of the target system 200, and an RAS function device that detects an internal abnormality of the computer main body 110 ( Anomaly monitoring device) 120.

  The computer main body 110 includes hardware resources such as a CPU 111, a memory 112, an auxiliary storage device 113, and an interface 114, and software resources such as a monitoring control program 115 and data 116 stored in the auxiliary storage device 113. .

  The RAS functional device 120 includes hardware such as a RAS processing unit 121, a RAS memory 122, a state detection unit 123 that detects the state of the computer main body 110, a hardware control unit 124 that controls hardware of the computer main body 110, and an input / output unit 125. Hardware resources and software resources such as a computer management program 126 and RAS data 127 stored in the RAS memory 122.

  Here, the state detection unit 123 counts the memory capacity reading unit 123b that reads the memory capacity of the memory 112 of the computer main body 110 and the number of times the memory capacity is reduced to the existing state detection unit 123a including a watchdog timer (WDT) and the like. In addition, the input / output unit 125 includes a signal input / output unit 125a and an alarm signal generation unit 125b. On the other hand, the computer management program 126 is a program specialized for computer management of the monitoring control system according to the present invention.

  The RAS function device 120 includes a memory leak detection unit 130 that monitors the change state of the memory 112 of the computer main body 110 using the resources described above and outputs an alarm indicating a memory leak detection when a memory leak is detected. It is configured.

[Operation during system operation]
During operation of the monitoring control system 100 according to the first embodiment as described above, the computer main body 110 constituting the main control device stores the monitoring control program 115 stored in the auxiliary storage device 113 on the memory 112. The CPU 111 operates in the work area, and the CPU 111 executes monitoring control processing. During such system operation, the memory leak detection means 130 configured in the RAS function device 120 performs memory leak detection processing.

[Memory leak detection processing]
FIG. 2 is a flowchart illustrating a memory leak detection process performed by the memory leak detection unit 130 according to the first embodiment.

  As shown in FIG. 2, the memory leak detection means 130 reads the current free memory capacity in the memory 112 of the computer main body 110 by the memory capacity reading unit 123b (S201), and the current free memory capacity is less than the previous value. (NO in S202, NO in S203), it is determined whether or not the count value of the change counter 123c is equal to or less than a preset specified value (S204). In this case, a specific value of the specified value can be freely selected. For example, the specified value “60 counter” can be set.

  When the count value of the change counter 123c is larger than the specified value (NO in S204), it is determined that the memory leak is detected, the alarm signal generation unit 125b generates an alarm signal indicating the memory leak detection, and the signal input / output unit 125a The alarm signal is output from (S205). In this case, the alarm output method can be selected and combined as appropriate, such as visual display of alarm messages via screens and LEDs, alarm sound output and alarm message audio output, transmission to external devices such as system operation manager terminals, etc. It is.

  On the other hand, when the count value of the change counter 123c is equal to or less than the specified value (YES in S204), the change counter 123c is counted up (S206). If the current free memory capacity is greater than the previous value (YES in S202), the change counter 123c is cleared (S207).

  Then, after an alarm signal for memory leak detection is output (S205), the change counter 125c is counted up or cleared (S206, S207), or when the current free memory capacity does not change from the previous value (YES in S203). ) Records the current free memory capacity as the previous value (S208), stops the processing for 24 hours (S209), and then restarts the memory leak detection process at the same time on the next day.

[Function and effect]
According to the first embodiment as described above, it is possible to output an alarm and urge the system operation manager to warn when a free memory area of the computer main body tends to decrease in system operation. Become.

  That is, as described above, even in a program that has undergone sufficient program verification, there is a possibility that an unexamined timing may occur or a combination of logics that are not expected during continuous operation for a long time may pass. Although there is a case where the opening is not performed, according to the present embodiment, even when a memory leak occurs, an alarm can be urged to the operation manager in advance before the system goes down. Therefore, by performing the memory leak detection process of the present embodiment after the system operation is started, it is possible to investigate the memory leak factor and take a countermeasure before the system is stopped, thereby realizing the stable operation of the system. Can be ensured.

  Also, the detection of memory leaks is not performed by a specific program on the computer main body, but is realized not by detecting only nonconformity of a specific application in the system by being realized by a RAS function device attached to the computer system. As a result, the safety and reliability of the entire system can be ensured.

[Second Embodiment]
[System configuration]
FIG. 3 is a block diagram showing a second embodiment of the supervisory control system to which the present invention is applied. As shown in FIG. 3, the monitoring control system 100 according to the present embodiment adds a dummy program 117 to the computer main body 110 of the monitoring control system 100 according to the first embodiment. 117 is configured to start and stop. Other configurations are the same as those of the first embodiment.

[Operation during system operation]
When the monitoring control system 100 according to the second embodiment as described above is in operation, the computer main body 110 constituting the main control device stores the monitoring control program 115 stored in the auxiliary storage device 113 on the memory 112. The CPU 111 executes the monitoring control process. In this case, at the start of the system operation, a part of the work area on the memory 112 in the computer main body 110 is moved by the dummy program 117 for securing the memory capacity by the initialization process by the memory leak detection unit 130 as shown in FIG. Reserved as a dynamic data area.

[Initialization]
FIG. 4 is a flowchart showing an initialization process by the memory leak detection unit 130 of the second embodiment.

  As shown in FIG. 4, the memory leak detection unit 130 reads the current total memory capacity in the memory 112 of the computer main body 110 by the memory capacity reading unit 123b at the start of system operation (S401). On the other hand, the dummy program 117 for securing a memory capacity of a memory securing ratio set in advance as a dynamic data area is activated (S402). In this case, the specific value of the memory allocation ratio can be set freely. For example, as shown in FIG. 4, it can be set to 10% of the current total memory capacity.

  When the dummy program 117 is activated, the memory allocation function (for example, malloc) may not acquire the real memory only by securing the area, and therefore sets random number data in the entire area.

[Memory leak detection processing]
FIG. 5 is a flowchart showing a memory leak detection process performed by the memory leak detection unit 130 according to the second embodiment.

  In the memory leak detection process shown in FIG. 5, when the current free memory capacity is read and the current free memory capacity is lower than the previous value, it is determined that the memory leak is detected if the count value of the change counter 123c is larger than the specified value. The series of processing (S501 to S507) until the alarm signal is output or the change counter 125c is counted up or cleared (S501 to S507) is the same as that in FIG. 2 (S201 to S207). It is the same.

  In the memory leak detection process in the present embodiment, a free memory capacity check process (S510) is performed following this series of processes (S501 to S507). In this free memory capacity check process (S510), the memory leak detection means 130 checks whether or not the current free memory capacity has decreased below a preset free memory ratio with respect to the total memory capacity (S511). In this case, the specific value of the free memory ratio can be set freely, but for example, as shown in FIG. 5, it can be set to 5% of the current total memory capacity.

  When the current free memory capacity is lower than the free memory ratio (NO in S511), the memory leak detection unit 130 generates an alarm signal indicating that the memory capacity is insufficient by the alarm signal generation unit 125b. The alarm signal is output from the signal input / output unit 125a, and if the dummy program 117 is operating on the computer main body 110, the hardware control unit 124 stops the dummy program 117 (S512).

  In this case, the alarm output method indicating that the total memory capacity is insufficient is the same as the alarm output when a memory leak is detected, such as visual display of an alarm message on the screen or LED, alarm sound output or alarm message sound output, etc. Transmission to an external device such as an administrator terminal can be selected and combined as appropriate.

  Then, after performing this free memory capacity check process (S510), the current free memory capacity is recorded as the previous value (S508), the process is stopped for 24 hours (S509), and the memory is stored at the same time on the next day. Restart the leak detection process.

[Function and effect]
According to the second embodiment as described above, in addition to the same functions and effects as those of the first embodiment, the following functions and effects can be obtained.

  First, according to the second embodiment, a dummy program that secures a specific memory area is operated in advance at the start of system operation, and the dummy program is terminated when memory exhaustion occurs due to a memory leak. As described above, when the memory decrease occurs, the memory area can be released by terminating the dummy program that has secured a certain amount of memory area.

  In this case, the memory area reserved by the dummy program is not changed to be used for an illegal program after that, but by temporarily increasing the memory area required by the OS of the computer main body, the memory area is exhausted. Sufficient time to stop the system can be secured. Therefore, the system can be operated more stably.

  Further, specific values of the memory securing ratio and the free memory ratio can be freely set. For example, in the second embodiment, the free memory capacity check process is performed using the same free memory ratio without distinction between the operation and stop of the dummy program. However, as a modification, in the free memory capacity check process, The first free memory ratio during the operation of the dummy program and the second free memory ratio after the dummy program is stopped may be set to different values.

  In either case, the specific values of the memory allocation ratio and the free memory ratio are not sufficient if a small value is obtained, but it is useless to set an excessively large value. It is desirable that the memory securing ratio is 10% and the free memory ratio is 5%.

[Third Embodiment]
As a third embodiment of the monitoring control system to which the present invention is applied, the monitoring control system shown in FIG. 3 having the same configuration as that of the second embodiment is replaced with the processing shown in FIGS. As shown in FIG. 6, an initialization process and a memory leak detection process may be performed in consideration of a memory change due to a normal operation process. Hereinafter, these processes will be described.

[Initialization]
FIG. 6 is a flowchart showing an initialization process by the memory leak detection unit 130 of the third embodiment.

  As shown in FIG. 6, in the initialization process of the third embodiment, the current total memory capacity is read (S601), and the memory capacity of the memory allocation ratio set in advance with respect to the total memory capacity is set as dynamic data. The point that the dummy program 117 secured as an area is started (S602) is the same as the initialization process of FIG. In the present embodiment, from the viewpoint of taking into account memory changes due to normal operation processing, a memory capacity with a memory leak exclusion ratio set in advance with respect to the total memory capacity is set as a memory leak exclusion constant (S603).

  Note that specific values of the memory allocation ratio and the memory leak exclusion ratio can be freely set. However, in the present embodiment, as an example, the memory allocation ratio is the same as the initialization process of FIG. The memory leak exclusion ratio is set to 1% of the current total memory capacity. In general, these values are desirable.

[Memory leak detection processing]
FIG. 7 is a flowchart showing a memory leak detection process performed by the memory leak detection unit 130 according to the third embodiment.

  In the memory leak detection process shown in FIG. 7, the processes (S701 to S703) until the current free memory capacity is read and the current free memory capacity is compared with the previous value are the same as those in FIG. This is the same as the series of processing (S501 to S503).

  In the memory leak detection process in the present embodiment, following this series of processes (S701 to S703), the amount of change of the current free memory capacity from the previous value is compared with the memory leak exclusion constant (S720). This is different from the memory leak detection process of FIG. If the amount of change is smaller than the memory leak exclusion constant (NO in S720), it is determined that the memory change is not caused by normal operation processing, and the process proceeds to determination of the count value of the change counter (S704).

  On the other hand, if there is a change equal to or greater than the memory leak exclusion constant (YES in S720), it is determined that the memory has changed due to normal operation processing, and free memory capacity check processing (S710) is performed. This free memory capacity check process (S710) is the same process as the free memory capacity check process (S710) of FIG.

[Function and effect]
According to the third embodiment as described above, in addition to the same functions and effects as those of the first and second embodiments, the following functions and effects can be obtained.

  First, according to the third embodiment, in the memory leak detection process, it is possible to distribute the memory depletion phenomenon due to a factor other than the memory leak and the memory leak. Alarm output indicating leak detection can be prevented from being performed.

  In other words, a phenomenon that is difficult to find such as a memory leak often decreases in memory over a long period of time. The processing in the present embodiment is based on this fact. When the memory gradually decreases over a long period of time, it is determined as a memory leak, and when the memory decreases (fluctuates) for a short time, it is not a memory leak. The alarm output for detecting the memory leak is not performed.

  However, even if it is a factor other than a memory leak, a memory decrease due to a program mistake may occur continuously, resulting in a memory exhaustion. Such a memory exhaustion itself may cause a program operation stop of the entire system in a computer system. Many cases lead to important problems. In this embodiment, when such a memory depletion occurs, the memory depletion is detected by the free memory capacity check process, and an alarm indicating a shortage of the total memory capacity is output, whereby a memory depletion other than a memory leak occurs. With regard to the above, it is possible to alert the system operation manager to a warning.

  As described above, in this embodiment, in the case of a temporary memory decrease change, an alarm output for detecting a memory leak is not performed, so that the possibility of a false alarm can be reduced in actual system operation. However, assuming that there is a continuous decrease in memory due to program mistakes, even if memory depletion occurs, an alarm indicating that the total memory capacity is insufficient will be output before the program stops. It is possible to alert the system operation manager to a warning. Therefore, the system can be operated more stably.

[Other Embodiments]
It should be noted that the present invention is not limited to the above-described embodiments, and various other variations can be implemented within the scope of the present invention. For example, specific configurations of the computer main body, the RAS function device, the memory leak detection means, and the like can be freely selected over both hardware resources and software resources. Further, the specific procedure of the memory leak detection process and the initialization process therefor can be freely selected.

1 is a block diagram showing a first embodiment of a monitoring control system to which the present invention is applied. 6 is a flowchart illustrating memory leak detection processing by the memory leak detection unit according to the first embodiment. The block diagram which shows 2nd, 3rd embodiment of the monitoring control system to which this invention is applied. 9 is a flowchart illustrating initialization processing by a memory leak detection unit according to the second embodiment. 9 is a flowchart showing memory leak detection processing by a memory leak detection unit of the second embodiment. 10 is a flowchart illustrating initialization processing by a memory leak detection unit according to the third embodiment. 10 is a flowchart illustrating memory leak detection processing by a memory leak detection unit according to the third embodiment.

Explanation of symbols

100: Monitoring control system 110 ... Computer main body (main control device)
111 ... CPU
112 ... Memory 113 ... Auxiliary storage device 114 ... Interface 115 ... Monitoring control program 116 ... Data 117 ... Dummy program 120 ... RAS functional device 121 ... RAS processing unit 122 ... RAS memory 123 ... State detection unit 123a ... Existing state detection unit 123b ... Memory capacity reading unit 123c ... Change counter 124 ... Hardware control unit 125 ... Input / output unit 125a ... Signal input / output unit 125b ... Alarm signal generation unit 126 ... Computer management program 127 ... Data 130 ... Memory leak detection means 200 ... Target system

Claims (8)

In a computer management method of a monitoring control system comprising a computer main body constituting a main control device for performing monitoring control of a target system and an abnormality monitoring device for detecting an internal abnormality of the computer main body,
By the abnormality monitoring device,
A computer management method for a monitoring control system, wherein a memory leak detection process is performed for monitoring a memory change state of the computer body and outputting an alarm indicating a memory leak detection when a memory leak is detected. By the abnormality monitoring device,
A dummy program for detecting the current total memory capacity of the computer main body at the time of starting the computer main body, and ensuring a memory capacity of a memory allocation ratio set in advance with respect to the detected total memory capacity as a dynamic data area Initialization process to start up on the computer,
The memory leak detection process includes
It is checked whether the free memory capacity of the computer main body has dropped below a preset free memory ratio with respect to the total memory capacity. 2. A computer management method for a supervisory control system according to claim 1, further comprising a free memory capacity check process for outputting and stopping the dummy program if the dummy program is operating. The memory leak detection process includes
Check whether the free memory capacity of the computer main body has dropped below a first free memory ratio set in advance with respect to the total memory capacity, and if it falls below the first free memory ratio, A first free memory capacity check process for outputting an alarm indicating insufficient capacity and stopping the dummy program;
After stopping the dummy program, it is checked whether or not the free memory capacity of the computer main body has dropped below a second free memory ratio set in advance with respect to the total memory capacity, and the second free memory ratio 3. The computer management method for a monitoring control system according to claim 2, further comprising a second free memory capacity check process for outputting an alarm indicating a shortage of capacity when the capacity drops below. 4. The computer management method for a supervisory control system according to claim 2, wherein the memory allocation ratio is 10% and the free memory ratio is 5%. The initialization process includes a process of setting a memory capacity of a memory leak exclusion ratio set in advance with respect to the detected total memory capacity as a memory leak exclusion constant,
The memory leak detection process includes a process of determining that it is not a memory leak but a normal capacity change when the amount of change in the free memory capacity of the computer main body is equal to or greater than a previous period memory leak exclusion constant. The computer management method of the supervisory control system according to any one of claims 2 to 4. 6. The computer management method for a supervisory control system according to claim 5, wherein the memory leak exclusion ratio is 1%. In a monitoring control system including a computer main body that constitutes a main control device that performs monitoring control of a target system and an abnormality monitoring device that detects an internal abnormality of the computer main body,
The abnormality monitoring device is:
A monitoring control system, comprising: a memory leak detecting means for monitoring a memory change state of the computer body and outputting an alarm indicating a memory leak detection when a memory leak is detected. In a computer management program used for an abnormality monitoring device of a monitoring control system including a computer main body constituting a main control device that performs monitoring control of a target system and an abnormality monitoring device that detects an internal abnormality of the computer main body,
By the abnormality monitoring device,
A computer management program for a supervisory control system, which monitors a memory change state of the computer main body and performs a memory leak detection process for outputting an alarm indicating a memory leak detection when a memory leak is detected.

Images