JPH10260870A - System starting and termination controller - Google Patents

Abstract

PROBLEM TO BE SOLVED: To detect an occurred error and a factor causing the last time machine stop, and to inform a remote maintenance place of it, by acquiring execution log showing the history of the starting processing and the terminating processing of a system and generating a log file. SOLUTION: When the system starting processing is called, a log saving processing part 11 saves the execution log files 16 and 17 in different names as saving log files 16a and 17a. A starting log output part 12 stores the execution process in the execution log file 16. An RAS information processing part 13 checks an error occurring situation from the contents of the execution log files 16 and 17 and checks the previous machine stop cause by a stop cause file 18. An RAS diagnosis part 20 outputs the fault content to the RAS log file 40 and pervious machine stop information. Then, it is RAS-informed to the remote maintenance place or the like through an online transmission part 30.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a system start / end control device, and more particularly to a system start / end control device for collecting a history of processing operations at the time of starting and ending a system.

[0002]

2. Description of the Related Art In a system in which a system operation or maintenance person is not resident, for example, in a terminal system for a store distributed in a wide area, various kinds of systems are prepared in case of a failure. Measures have been taken.

For example, according to Japanese Patent Application Laid-Open No. 02-231650, a maintenance countermeasure program for monitoring the execution status of the system and a program for displaying the cause of the system trouble are provided to collect the cause of the trouble. Further, according to Japanese Patent Application Laid-Open No. 08-65387, a method is proposed in which when a software failure occurs, the cause of the failure is analyzed, and when the system is restarted, it is quietly operated.
Further, according to Japanese Patent Application Laid-Open No. 06-348546, a method is proposed in which logs are collected for each job and the operation status is grasped, regardless of the configuration of the program.

[0004] However, all of the conventional systems described above relate to the execution status of the business system, and do not relate to the startup processing or the termination processing of the terminal system.

[0005]

The first problem with the conventional system as described above is that some trouble occurs during the start-up and end processes of the system, and the process cannot be continued. In this case, it is restored by pressing the reset switch or the dump switch. At this time, there is no information collecting means for analyzing which process has caused the failure, how far the process has been executed, and the like. The reason is that even if an error message is output on the screen, the screen information is rarely saved due to a reset operation etc. for quick recovery, and it is troublesome to secure the information manually. This is because it takes time and reliability is low.

A second problem is that when a system stall or panic occurs during the system start-up process and the end process, the execution log file being collected by pressing a reset switch or a dump switch or automatically executing a dump routine. Will be overwritten, where the trouble occurred during the previous system startup and shutdown processing,
The factor cannot be identified. The reason is that the execution log file needs to be overwritten as needed to avoid the simple increase of the file procedure, but if the previous execution history was overwritten, it would no longer be meaningful to collect it. This is because

A third problem is that when a failure is detected during the system startup processing and the termination processing, the fact is output to a screen or a printer. Is not possible. The reason is that, in a system in which a machine operator does not reside, particularly in a store system, it is meaningless to show general users trouble contents such as an error message of a computer machine, but only to be confused.

An object of the present invention is to improve the maintainability of the system by collecting an execution history during system startup processing and termination processing and storing the log file for one generation, in order to improve the above problems. An object of the present invention is to provide a system start / end control device that detects an error that occurred during system start processing and end processing and a factor that caused the previous machine stop and notifies a remote maintenance base or the like.

[0009]

SUMMARY OF THE INVENTION According to the present invention, there is provided a method for starting a system.
The termination control device is configured to collect an execution log indicating a history of system startup processing and termination processing and create a log file.

The system start / end control device according to the present invention includes a log save processing unit for saving an execution log of one generation before when a power ON request is received, and a system start processing according to the power ON request. A start log output unit for executing and sending the history as an execution log; an end log output unit for executing a stop process of the system according to a power-off request and sending the history as an execution log; An execution log file that holds an execution log sent by the end log output unit and a save log file that saves the previous generation execution log held in the execution log file are provided.

In addition, the system start / end control device of the present invention, when receiving a power-off request, detects a stop factor and notifies the end log output unit and stores the stop factor in a stop factor file. And a stop factor file holding the stop factor, wherein the start log output unit checks the error occurrence status in the start process when the system start process is completed, and checks the last time held in the stop factor file. And a RAS notification processing unit for checking the cause of the stop and sending it out as RAS log information.

Further, the system start / end control device of the present invention analyzes the RAS log information and reports a failure content, a RAS log file storing the failure content, and transmits the failure content to the outside. And an online transmission unit for notifying.

That is, the system start / end control device according to the present invention collects the execution log files at the time of the system start-up processing and the end processing, thereby determining what kind of processing has caused the failure during the execution, and how far the processing has been performed. Provides information collection means for analyzing whether the execution has been performed.

As an initial process at the time of starting the system,
If the system stalls or panics during the system startup and termination processing by saving the execution history of the previous system startup and termination processing for one generation, the reset switch or dump switch is pressed, or the dump routine is automatically executed. It is possible to prevent the execution log file being collected from being overwritten by the execution, and to identify the trouble occurrence location and the cause during the previous system start-up processing and end processing.

An error occurrence situation is detected from the log files collected at the time of system startup processing and termination processing, and the contents of the failure are reported to a remote maintenance base or the like. It can be executed, and in the case of an urgent report, a maintenance person can be arranged and dealt with immediately, thereby improving maintainability.

Determining the cause of the previous machine stop may provide clues for determining the need for automatic recovery processing.

Further, in a store system or the like in which a machine operator is not always resident, by notifying the contents of a fault or a previous machine stop cause to a remote maintenance base or the like,
The report information can be centrally managed, and preventive maintenance work by maintenance personnel is enabled.

[0018]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of the present invention. In FIG. 1, a system start / end control device 10 according to the present invention includes a log save processing unit 11 that saves an execution log of one generation before when a power ON request is received,
A start log output unit 1 that executes a system start process in accordance with the power-on request and sends the history as an execution log.
2, an end log output unit 15 that executes the stop processing of the system according to a power-off request and sends the history as an execution log, and holds an execution log sent by the start log output unit and the end log output unit. The system includes execution log files 16 and 17 and save log files 16a and 17a for saving the previous generation execution log held in the execution log file.

Further, in the system start / end control device 10, when a power-off request is received, the stop factor is detected, notified to the end log output unit, and stored in a stop factor file. And a stop factor file 18 for holding the stop factor.
Upon completion of the system startup processing, the startup log output unit 12 checks an error occurrence situation in the startup processing, checks a previous stop factor stored in the stop factor file, and sends out the RAS message as RAS log information. A processing unit 13 is included.

In FIG. 1, a system activation / termination control device 10 is included in the information processing device 1, and the information processing device 1 further includes an RAS diagnosis unit 20 that analyzes the RAS log information and reports a failure content. R for storing the failure content
An AS log file 40 and an online transmission unit 30 for notifying the failure contents to the outside are provided. Note that the online transmission unit 30 is connected to the external maintenance information processing device 2 and reports the content of the failure described above.

FIG. 2 is a flowchart showing a starting operation in the system start / end control device 10 described above. In the figure, when the system startup processing is called, first, the log save processing unit 11 saves the execution log files 16 and 17 as save log files 16a and 17a, respectively, under different names (S11). After that, various startup processes are executed, and the execution log output unit 12 stores the execution process in the execution log file 16 (S12).

At the end of the various startup processes, the RAS notification processing unit 13 checks the error occurrence status from the contents of the execution log files 16 and 17 (S13), and further checks the previous machine stop factor from the stop factor file 18. (S14). Then, the RAS diagnostic unit 20 is requested to store the information to be stored in the RAS log file 40 and the information to be reported to a remote maintenance base or the like (S1).
5, S16).

The RAS diagnosis unit 20 responds to the request by
The failure content and the previous machine stop cause information are output to the RAS log file 40 (S17). In addition, RAS notification is made to a remote maintenance base or the like via the online transmission unit 30 (S18).

FIG. 3 is a flowchart showing an end operation in the system start / end control device 10 described above. In the figure, when the system termination processing is called, the stop cause monitoring unit 14 stores the fact in the stop cause file 18 (S21). An initial value is assigned to the stop factor file 18, and the stop factor is determined by changing the contents of the initial value to the detected stop factor.
If not detected due to CPU stall, etc.,
Since the content of the stop cause file 18 is kept at the initial value, in this case, it is determined that the CPU has been stalled.

Thereafter, various stop processes are executed, and the end log output unit 15 stores the execution process in the end process execution log file 17.

FIG. 4 is a block diagram showing a specific example of the system start / end control device. FIG. 1 shows an example where the platform is UNIX.

Referring to FIG. 4, the store system 51 includes:
System start / end control device 60, RAS daemon 70 and online daemon 80 which are resident processes
And an RAS log file 90. And it is connected to a remote maintenance base system 52.

Further, the system start / end control device 60
Is a log save processing unit 61 using a copy command,
A startup log output unit 62 using the standard output file function of the shell, an RAS report processing unit 63 combining an awk or grep command or an original RAS command, and a stop factor monitoring daemon 64 which is a resident process for receiving a stop factor signal. An end log output unit 65 also using the standard output file function of the shell, an execution log file 66 of a start process, which is a standard UNIX file and its save log file 66a, and an execution log file 6 of the end process
7 and its save log file 67a, and a stop cause file 68.

FIG. 5 is a flowchart showing a starting operation in the system start / end control device 60 described above. In the figure, when the system startup processing is called by turning on the power by pressing the DC switch or pressing the reset switch and calling the reboot processing, first, the log save processing unit 61 first copies the existing execution log files 66 and 67 by the copy command. Backup log file 6 with another name
Save as 6a and 67a (S61). afterwards,
Various startup processing shell scripts (/etc/rc2.d
/ And / etc / rc. The shell script starting with a capital letter "S" under the d / directory is executed, and the execution log output unit 62 stores the execution process in the startup processing execution log file 66 using the standard output file function of the shell (S62).

At the very end of the execution of the various startup processing shell scripts, the RAS notification processing unit 63 is executed. Here, as from the contents of the execution log files 66 and 67B,
The error occurrence status is checked by combining k and grep commands (S63).
Then, the previous machine stop factor is read from 8 and checked (S64). Further, a request is made to the RAS daemon 70 for an item to be stored in the RAS log file 90 and for an item to be reported to a remote maintenance site or the like using unique RAS commands (APIs) (S65, S66).

In response to the request, the RAS daemon 70 outputs the contents of the failure and the previous machine stop cause information to the RAS log file 90 (S67). Further, the RAS is notified to a remote maintenance base or the like via the online daemon 80 (S68).

FIG. 6 is a flowchart showing an end operation in the system start / end control device 60 described above. In the figure, when the system termination is called by the SW / DC switch being pressed, the UPS being detected by a power failure, the occurrence of a panic, the occurrence of a stall, etc. during the operation of the machine, the stop factor monitoring daemon 64 stores the fact in the stop factor file 68. (S71). An initial value is assigned to the stop factor file 68, and the stop factor is determined by changing the contents of the initial value of the software to the detected stop factor.
If not detected due to CPU stall, etc.,
Since the contents of the stop cause file 68 are kept at the initial values, in this case, it is determined that the CPU has been stalled. After that, various stop processing shell scripts (shell scripts starting with a capital letter "K" under the /etc/rc0.d/ directory) are executed, and the execution process is terminated by the log output unit 65 using the standard output file function of the shell. Stored in the execution log file 67 (S7
2).

[0034]

As described above in detail, according to the system start / end control device of the present invention, an execution log at the time of system start processing and end processing is collected and further stored for one generation. Therefore, there is an effect that information for analyzing the cause of failure in the system start-up process and the end process can be collected.

Further, by determining the cause of the previous machine stoppage, it is possible to obtain clues for determining the necessity of the automatic recovery process.

Further, in a store system or the like where a machine operator does not reside, it is possible to report the contents of a failure to a remote maintenance base or the like, so that remote maintenance from a remote machine is executed by using the report as a trigger. In the case of an urgent report, maintenance personnel can be arranged and dealt with immediately, and maintainability is improved.

Further, by notifying the contents of the fault and the cause of the previous machine stop to a remote maintenance base or the like, the report information can be centrally managed, thereby enabling preventive maintenance work by maintenance personnel.

[Brief description of the drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention.

FIG. 2 is a flowchart showing a starting operation of the present invention.

FIG. 3 is a flowchart showing an end operation of the present invention.

FIG. 4 is a block diagram showing a specific example of an embodiment of the present invention.

FIG. 5 is a flowchart showing a starting operation of the example of FIG. 4;

FIG. 6 is a flowchart showing an end operation of the example of FIG. 4;

[Explanation of symbols]

 Reference Signs List 10 system start / end control device 11 log save processing unit 12 start log output unit 13 RAS notification processing unit 14 stop factor monitoring unit 15 end log output unit 16, 17 execution log file 16a, 17a save log file 18 stop cause file 20 RAS Diagnosis unit 30 Online transmission unit 40 RAS log file

Claims (5)

[Claims] 1. A system start / end control device, which collects an execution log indicating a history of system start processing and end processing and creates a log file. 2. The system start / end control device according to claim 1, wherein an execution log of a previous generation indicating the history of the previous start process and end process is saved. 3. The system according to claim 1 or 2,
A log save processing unit that saves an execution log of one generation before when a power-on request is received, and a start-up that executes a system startup process in accordance with the power-on request and sends the history as an execution log. A log output unit, an end log output unit that executes a stop process of the system in accordance with a power-off request, and sends the history as an execution log, and holds an execution log sent by the start log output unit and the end log output unit. And a save log file for saving an execution log of one generation before stored in the execution log file. 4. A stop factor monitor according to claim 3, wherein when a power-off request is received, a stop factor is detected and notified to said end log output unit and stored in a stop factor file. Unit, and a stop factor file for holding the stop factor, wherein the start log output unit checks an error occurrence state in the start process when the system start process is completed and is held in the stop factor file. A system start / end control device characterized by including a RAS notification processing unit for checking a previous stop cause and sending it as RAS log information. 5. The system start / end control device according to claim 4, wherein the RAS diagnostic unit analyzes the RAS log information and reports a failure content, and the RA stores the failure content.
An S-log file and an online transmission unit for notifying the contents of the failure to the outside;
Termination control unit.