JP2001056772A - Fault monitoring system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress the occurrence of a fault in client operation type unmanned terminal system operating for 24 hours by judging the presence/absence of abnormality by comparing acquired supervisory information with set supervisory information, judging the presence of abnormality when the two supervisory information are different and restarting a process. SOLUTION: The used memory capacity, used handle number and CPU using rate of a supervisory object process are acquired (S1). The acquired used memory capacity, handle number and CPU using rate are compared with values acquired up to the last time and it is judged whether resources are abnormally consumed or not (S2). When abnormality is related to hang or loop, the state of disabling the provision of services to a client is judged and restart processing is immediately started (S5). When abnormality is resource leak, it is judged the service provision to the client can be continued (S6), and all the processes under managing object are finished (S7). A Window NT is rebooted and recovery from abnormality is attained (S8).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a fault monitoring system and, more particularly, to a system that requires a 24-hour continuous operation of a customer-operated unattended terminal system that directly operates a computer system in order for a customer to use a service. Useful for operation.

[0002]

2. Description of the Related Art In a conventional computer system, it has been possible to immediately detect a failure in which service provision is stopped and restart the system. However, due to a memory leak due to a bug in the application or the like, it does not cause an immediate failure, but it is not possible to detect in advance the occurrence of an abnormality that will eventually become a failure when the operation is continued.

For this reason, an operator has to respond after an actual failure has occurred, or the operation manager has monitored the memory usage rate or the like.

[0004]

However, as the operation time is extended and the number of systems that provide continuous service for 24 hours is increasing, there is a limit to manned coping, and it is anticipated that failures will occur. Need to be proactively maintained to improve availability. In addition, current terminal systems are often created on a Web (Web) basis. In such a case, it is necessary to detect and deal with an abnormality specific to a Web (Web) browser. The conventional method described above has the following problems.

[0005] (1) Although a failure does not occur immediately, an abnormality that causes a failure if the operation is continued cannot be dealt with until a failure occurs. (2) Abnormalities in the Web browser are not limited to general resources, hangs, and loops, but can also be caused by service suspension due to an error dialog displayed by the Web browser.
The state cannot be detected as abnormal.

(3) Among the above error dialogs, there are those that are restarted as errors and those that should not be processed as errors, but cannot be separated. (4) In the case of restarting, if all are restarted from the OS (operating system) as in the past, the service suspension time becomes longer. (5) Depending on the business hours, even if the customer is likely to be using it, a restart will occur unconditionally,
The customer is waiting for the use of the terminal.

[0007] Therefore, due to the above problems, it was not technically satisfactory.

[0008]

According to the present invention, monitoring information of a process to be monitored in a normal state is set in advance, monitoring information of the process is obtained, and the monitoring information set with the obtained monitoring information is set. The present invention provides a failure monitoring system characterized in that the presence or absence of an abnormality is compared by comparing the two pieces of monitoring information, and when the two pieces of monitoring information are different, it is determined that there is an abnormality, and the process is restarted. As a result, it is possible to detect the occurrence of an error that causes a failure based on the monitoring information based on the monitoring information. Automatically restarts when a failure is detected, and the occurrence of a failure can be suppressed in advance. For this reason, it is possible to suppress a failure of the customer-operated unattended terminal system that operates 24 hours a day.

A web browser that displays an error dialog when an error exists is set as a monitoring target, and when the error dialog is displayed when the monitoring information is obtained, it is determined that there is an abnormality, and the web browser is restarted. You may make it start. As a result, the web browser itself detects a failure and displays an error dialog, thereby detecting a stop of the application that occurs, and automatically recovering by restarting the application.

A list of abnormalities requiring restart may be set in advance, and if the recognized abnormality is in the list, the system is restarted, and if not, the system is not restarted. This makes it possible to automatically distinguish abnormalities that require a restart from abnormalities that do not, and automatically recover by restarting when an abnormality that requires a restart is detected.

Further, a list of abnormalities requiring restart of the operating system is set in advance, and if the recognized abnormality is in the list, the operating system is restarted. A restart may be performed. Thus, in the case of an error that can be recovered only by restarting the process, only the related process can be restarted, and the unnecessary processes can not be restarted, so that the time required for restart can be reduced. .

Furthermore, a time zone in which restart is possible and a time zone in which restart is not possible are set in advance, a time zone when an abnormality is recognized is grasped, and restart is performed only in a time zone in which restart is possible. It may be. Thus, in the case of a failure that does not need to be restarted immediately, the restart can be performed according to the usage status of the terminal. For this reason, if the restart is not possible, the execution of the restart can be made to wait, and the time during which the user cannot operate the terminal due to the restart can be reduced.

[0013]

Embodiments of the present invention will be described below with reference to the drawings. It should be noted that the present invention is not limited by this. First Embodiment FIG. 1 is a functional block configuration diagram of the present invention. In the figure, reference numeral 1 denotes a monitoring function, which has a function of monitoring various processes and application processes constituting a Web (Web) browser and a platform, and detecting an abnormality.
Reference numeral 2 denotes an operation management function, which has a function of managing a dependency between various processes and application processes constituting a Web browser and a platform. Reference numeral 3 denotes a Web browser, which has a function of displaying a business screen using normal HTML or script. Reference numeral 4 denotes various processes and application processes constituting the platform, which are processes for providing platform functions and business functions. Here, PF process 4 (1), PF process 4
(2), AP process 4 (3), AP process 4 (4)
Is shown. In general, there is often a dependency between the processes, and therefore, it is assumed here that there is a dependency. Reference numeral 5 denotes an operation management information table, which describes dependencies between various processes and application processes constituting a Web browser and a platform. Reference numeral 6 denotes an operating system. Here, Windows NT (manufactured by Microsoft Corporation (US)) is assumed. Each function can be realized by the CPU executing a program stored in a storage medium such as a ROM or a RAM, but a detailed description thereof will be omitted.

Next, a flow of a smart restart process for preventing the system from being stopped due to a memory leak or a leak of system resources in advance will be described. FIG. 2 shows a flowchart thereof. S1: The monitoring function 1 acquires the amount of used memory, the number of used handles, and the CPU usage rate of the monitored process. Furthermore,
Sends a Windows message to the target process and checks if there is a response.

S2: The monitoring function 1 compares the obtained used memory amount, the number of handles, and the CPU usage with the values obtained up to the previous time to determine whether or not abnormal resource consumption has occurred. S3: The monitoring function 1 repeats S1 and S2 until all the monitoring target processes have been checked. S4: When all the monitoring target processes have been checked, the monitoring function 1 sleeps until the next monitoring.

S5: If the abnormality is a hang or a loop, the operation management function 2 determines that the service provision to the customer is already impossible, and immediately starts the restart processing. S6: If the abnormality is not a hang or a loop but a resource leak, the operation management function 2 determines that the service provision to the customer can be continued, and if the service is provided to the customer, the service management function 2 proceeds to the service. Wait for the service to end without any impact.

S7: The operation management function 2 terminates all processes under management. S8: The operation management function 2 reboots Windows NT6 to recover from the abnormality. All processes to be managed are automatically started when Windows NT is started, and services can be resumed for customers. According to the first embodiment, the severity of a failure (hang, loop,
In response to a resource leak, appropriate appropriate processing can be automatically performed. Furthermore, at the present time, such as a resource leak, which does not lead to the suspension of service to the customer, but if left unchecked, it is possible to detect in advance and deal with an abnormality that may cause a serious failure in the future. Also, instead of immediately restarting when a resource leak is detected, the service status (whether the service is being provided to the customer) is determined and the restart process is implemented without affecting the customer in use. be able to.

Second Embodiment Since the functional configuration of the second embodiment is the same as that of the first embodiment, a description thereof will be omitted, and processing different from that of the first embodiment will be omitted. Only the explanation will be given. In the second embodiment, when a Web browser is to be monitored,
An operation for realizing smart restart by detecting a state in which the service cannot be provided to the customer due to the operation state specific to the browser b will be described. FIG. 3 shows the flowchart. Note that the same steps as those in FIG. 2 are denoted by the same reference numerals, and description thereof will be omitted.

S9: The monitoring function 1 checks that the monitoring target is Web
It is determined whether or not the browser 3 is used. If the target is a Web browser, information unique to the Web browser 3 is monitored. The operation management function 2 can determine whether the target is the Web browser 3 by referring to the operation management information table 5. The operation management information table 5 describes the types of all processes (platform processes, application processes, third-party processes such as Web, etc.) managed by the present system, and the dependencies between the processes. Information on whether or not a process is a monitoring target may be managed in the management information table 5.

S10: The monitoring function 1 is the Web browser 3
Is displayed by acquiring information of the Web browser to determine whether or not an error dialog is displayed. If the error dialog is not displayed, it is judged to be normal and the next monitored process is checked. If an error dialog is displayed, it is determined by the display of the dialog that the business screen transition flow has been interrupted, and restart processing is executed.

According to the second embodiment, when a customer-operated system for controlling a screen with a Web browser is provided, a function for detecting the occurrence of an error dialog automatically displayed by the Web browser is provided. In addition, it is possible to detect a failure in which a screen transition is stopped because an application operating on a Web browser cannot be detected. Therefore, it is possible to automatically deal with a failure that has been handled by a man, and cost reduction can be expected.

Third Embodiment The functional configuration of the third embodiment is the same as that of the first embodiment, and therefore the description is omitted, and the processing is different from that of the second embodiment. Only the explanation will be given. In the third embodiment, the process of detecting the stop of the service due to the problem peculiar to the Web browser and restarting it according to the second embodiment can be further dealt with according to the cause of the failure detected by the Web browser. The operation in the case of the above will be described. FIG. 4 shows a flowchart of the operation. 2 or 3
The same steps as those described above are denoted by the same reference numerals, and description thereof will be omitted.

S11: The monitoring function 1 is the Web browser 3
In order to find out what kind of trouble is detected and displayed by the user, the title of the error dialog currently displayed by the Web browser 3 is acquired from the Web browser 3. S12: The monitoring function 1 executes a smart restart process according to the title of the error dialog registered in the operation management information table 5 from the title of the output error dialog. If an error dialog not to be restarted is displayed, the customer continues processing as it is. If an error dialog for restart is displayed, restart processing is executed.

According to the third embodiment, when a customer-operated system for controlling a screen with a Web browser is provided, a failure detected by the Web browser is recognized from an error dialog automatically displayed by the Web browser. Since the necessity of restart can be determined, it is possible to reduce occurrence of system prevention due to Web browser abnormality detection according to the second embodiment. Therefore, the effect of reducing the number of service stoppages due to system restart due to abnormality detection can be expected.

Fourth Embodiment Since the functional configuration of the fourth embodiment is the same as that of the first embodiment, a description thereof will be omitted, and processing different from that of the third embodiment will be omitted. Only the explanation will be given. In the fourth embodiment, there is provided a function of isolating whether or not WindowsNT needs to be rebooted and a function of automatically determining a process that requires restart and restarting in the case of an abnormality that does not require rebooting. The operation of the restart process according to will be described. FIG. 5 shows a flowchart of the operation. The same steps as those in FIG. 2, FIG. 3, or FIG. 4 are denoted by the same reference numerals, and description thereof will be omitted.

S13: The monitoring function 1 determines that the detected abnormality is W
It is determined whether it is necessary to reboot Windows NT6. If it is necessary to reboot Windows NT6, the process of S7 is executed. Windows
A failure that does not require the sNT6 to be rebooted restarts only the process associated with the monitored process that has detected the abnormality. The relation between the cause of the failure and the information of whether or not to reboot the NT is stored in the monitoring function in advance.

S14: The operation management function 2 reads information from the operation management information table 5 in order to determine a process related to the process to be monitored. S15: In the operation management function 2, since the operation management information also describes information on the dependency between processes,
The restart target processes are sequentially terminated according to the dependency. The dependency between processes describes the start order and the end order when a process that depends on the start and end order exists.

S16: After the termination process of the restart target process is completed, the operation management function 2 sequentially restarts the target processes according to the dependency in the operation management information. According to the fourth embodiment, when restarting after detecting an abnormality, it is possible to automatically detect whether or not to reboot the NT according to the cause of the abnormality. further,
Only processes related to the process that detected the abnormality can be restarted, and unrelated processes can be left as they are, so that the time required for restart can be reduced. Therefore, an effect of further reducing the restart time of the system due to the abnormality detection can be expected.

Fifth Embodiment Since the functional configuration of the fifth embodiment is the same as that of the first embodiment, a description thereof will be omitted, and processing different from that of the fourth embodiment will be omitted. Only the explanation will be given. In the fifth embodiment, an operation for realizing a process for determining whether or not restart is possible according to business hours will be described. FIG. 6 shows a flowchart of the operation. In addition, FIG. 2, FIG. 3, FIG.
The same steps as those described above are denoted by the same reference numerals, and description thereof will be omitted.

S17: In the case of a failure that does not need to be restarted immediately, the monitoring function 1 reads the business time information in order to automatically determine whether or not to execute the restart according to the business time, Check which time zone the current time is classified into. The business hours need to be registered separately in advance in a time zone in which immediate restart is possible and in a time zone in which restart is not possible (time zone in which terminals are frequently used).

S18: If the time when the abnormality is detected is not in the restartable time period, the monitoring function 1 causes the restart to wait until the restartable time period. When the restartable time period is reached, restart processing is executed. According to the fifth embodiment, it is possible to execute a restart process by detecting an abnormality that does not need to be immediately restarted such as a resource leak in advance according to a preset use time period of the terminal. It is possible to improve the convenience so that the service stop due to the restart is not affected by the user.

[0032]

As described above, according to the fault monitoring system of the present invention, a fault does not occur immediately, such as a resource leak, a hang, or a loop of a monitored process, but if the operation is continued as it is, the fault will not occur. A customer operation type that detects the occurrence of the cause of the abnormality based on the monitoring information, automatically restarts the system when an abnormality is detected, and suppresses the occurrence of the failure in advance. The effect of being able to suppress the occurrence of a failure in the unattended terminal system is obtained.

It should be noted that a web browser displaying an error dialog when an error exists is set as a monitoring target, and when the error dialog is displayed when the monitoring information is obtained, it is determined that there is an abnormality, and the web browser is restarted. When the application is started, an effect is obtained in which the web browser itself detects a failure and displays an error dialog to detect a stop of the application that occurs, and can automatically recover by restarting.

Also, a list of abnormalities that need to be restarted is set in advance, and if the recognized abnormality is in the list, the system is restarted. If not, the system is not restarted. Is automatically distinguished from abnormalities that need to be restarted, and when an error that requires restarting is detected, the effect of automatically recovering by restarting is obtained.

Further, a list of abnormalities requiring restart of the operating system is set in advance, and if the recognized abnormalities are in the list, the operating system is restarted. By restarting, if an error can be recovered only by restarting the process, only the related process can be restarted, and unnecessary processes can not be restarted. The effect that can be shortened is obtained.

Further, a time zone in which restart is possible and a time zone in which restart is not possible are set in advance, a time zone when an abnormality is recognized is grasped, and restart is performed only in a time zone in which restart is possible. If the failure does not need to be restarted immediately,
Restart can be performed according to the usage status of the terminal, and if the restart is not possible, the restart can be made to wait, reducing the time during which the user can not operate the terminal by restarting The effect that can be obtained is obtained.

[Brief description of the drawings]

FIG. 1 is a functional block configuration diagram of the present invention.

FIG. 2 is a flowchart according to the first embodiment;

FIG. 3 is a flowchart according to a second embodiment;

FIG. 4 is a flowchart of a third embodiment.

FIG. 5 is a flowchart according to a fourth embodiment;

FIG. 6 is a flowchart according to a fifth embodiment;

[Explanation of symbols]

 1 monitoring function 2 operation management function 3 web browser 4 process 5 operation management information table

Claims (5)

[Claims] 1. Monitoring information of a process to be monitored at normal time is set in advance, monitoring information of the process is obtained, and the obtained monitoring information is compared with the set monitoring information to determine whether there is an abnormality. A failure monitoring system that determines that there is an abnormality when the two pieces of monitoring information are different from each other, and restarts the process. 2. The method according to claim 1, wherein a web browser that displays an error dialog when an error exists is set as a monitoring target, and when an error dialog is displayed when the monitoring information is obtained, it is determined that there is an abnormality. A fault monitoring system, wherein the web browser is restarted. 3. The method according to claim 1, wherein a list of abnormalities requiring restart is set in advance, and if the recognized abnormality is in the list, the system is restarted. A failure monitoring system characterized by not being restarted. 4. A method according to claim 1, wherein a list of abnormalities requiring restart of the operating system is set in advance. A failure monitoring system that performs a restart and restarts a process if the process is not on the list. 5. A time zone according to claim 1, 2, 3, or 4, wherein a time zone in which restart is possible and a time zone in which restart is not possible are set in advance, and a time zone when an abnormality is recognized. A failure monitoring system characterized in that the system is restarted only in a time zone in which it can be restarted.