JP2001282554A - Monitoring method and monitoring device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To periodically monitor the respective monitoring items of a computer system so that the influence exerted on an information system becomes small. SOLUTION: Monitoring job modules monitoring the respective monitoring items are installed for the respective monitoring items. A monitoring job schedule generation program 3045 decides the execution time bands of the respective monitoring job modules so that the load of monitoring in the respective time bands becomes not more than a threshold and the execution period of the respective monitoring modules becomes the monitoring period of the corresponding monitoring item from the monitoring periods of the monitoring items, which are mentioned in a monitoring job list table 3041, and the required execution time of the respective monitoring job modules mentioned in a monitoring job execution time table 3042, and registers it in a monitoring job periodical start time table 3043 as the execution schedule of the monitoring job module.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technique for monitoring the operation status of an information system.

[0002]

2. Description of the Related Art As a conventional technology for monitoring the operation status of an information system, the acquisition and setting of the status of an information network system composed of server computers and communication devices are performed by using SNMP (Simple Network Management Protocol) or the like. 2. Description of the Related Art There is known a "network management system" that is remotely performed using a protocol.

Here, in such an SNMP-based network management system, as a method of monitoring the operating status of each monitoring target of the network information system,
A method called "trap" in which the monitoring agent notifies the manager of the operating status of the monitoring target as appropriate according to the operating status of the inspection target, and the manager reports the operating status to the monitoring agent that operates on the monitoring target at regular intervals. A method called "polling" is known in which a monitoring agent reports an operation status of a monitoring target in response to the request.

[0004]

Now, with the development of the Internet, an information system for providing a high-performance Internet service such as an electronic mall has been realized. This type of information system is a complex system including various elements such as a network application that receives a request from a user via a network, and a back-end application for identifying and authenticating the user. The targets of the monitoring items that should be periodically monitored in such an information system include a computer constituting the information system, hardware in a network device,
Network, operating system, middleware, application software, and so on.

[0005] On the other hand, an information system that provides an Internet service provides a service to an unspecified number of users, and therefore has a characteristic that a load level due to a service request varies greatly with time. Due to this, if these various monitoring items are monitored at random, the load provided by the monitoring may affect services provided by the information system. Therefore, it is important to determine the timing of monitoring each monitoring item.

Conventional techniques related to scheduling include, for example, US Pat. No. 5,669,955 “Method of scheduling successive data” applied to a real-time OS or the like.
A technique described in “skssubject only to timing constraints” is known. In this technique, an upper limit value and a lower limit value of the execution time of the task are set in advance for each task, and the first time the task is set to the lower limit value. 2 in ascending order
In the third round, each task is scheduled in the order of increasing the upper limit value, and tasks that do not satisfy the constraints are adjusted, thereby scheduling the tasks.

[0007] The present invention has been made in view of the above circumstances, and an object of the present invention is to schedule a plurality of jobs (tasks) that are repeatedly executed periodically so as to reduce the load on the system. Is to do. In particular,
An object of the present invention is to schedule the periodic monitoring of each monitoring item of the information system so that the effect on services provided by the information system is reduced.

[0008]

In order to achieve the above object,
According to the monitoring method of the present invention, for each of a plurality of monitoring processing programs prepared in advance for each of a plurality of items and monitoring a monitoring target of the item, a time required for executing the monitoring processing program is obtained. Further, for each of the plurality of monitoring processing programs prepared in advance for each of the plurality of items, a cycle for executing the monitoring processing program is determined. Then, based on the obtained time required for execution of each monitoring processing program and the determined cycle at which each monitoring processing program is to be executed, the load caused by the execution of each of the plurality of monitoring processing programs prepared in advance for each of the plurality of items is determined. Are scheduled in time so that each of the plurality of monitoring processing programs is started at the corresponding cycle determined in the second step. I do.

In the present invention, a monitoring processing program is prepared for each of a plurality of items. By doing so, the execution time of each monitoring process can be shifted for each monitoring item. Therefore, utilizing this fact, the execution start time of the monitoring processing program is scheduled so that the cycle in which the monitoring processing program should be executed is kept and the load is temporally distributed. As a result, by executing each monitoring processing program according to this scheduling, periodic monitoring of each monitoring item of the information system is performed.
It is possible to reduce the influence on the service provided by the information system.

[0010]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below.

FIG. 1 shows a configuration of a monitoring system according to an embodiment of the present invention.

In FIG. 1, a computer system 103 is an information system to be monitored, and includes a plurality of service computers 1031 for providing services and a service system network 1033. Computer system 103
Provides a service to the user. The user
Using the user terminal 101, a service is received from the computer system 103 via the network 102.

In order to normally operate and manage such a computer system 103, a system administrator of the computer system 103 uses a system administrator terminal 203 to manage the management computer 202 connected to the service system network 1033. , The operation status is constantly monitored for a failure in the computer system 103. Here, even if the system administrator terminal 203 accesses the management computer 202 via the service system network 1033, the system administrator terminal 203 ′
As described above, the management computer 202 may be accessed via another network 110.

Next, in such a monitoring system,
The configuration for performing the monitoring is changed to an arbitrary service computer 103.
FIG. 2 (a) shows an example of monitoring the monitoring target n on 1.

In FIG. 2A, the service computer 10
A service providing unit 2012 for providing a service to the user is provided inside 31. Also, the management computer 202
Is a computer for supporting the management of the computer system 103, and is a service providing unit 2 of the service computer 1031.
A monitoring processing unit 2022 for periodically checking the operation status of the 012 from the outside via a network, and a monitoring processing unit 202
There is a result report processing unit 2021 that receives the monitoring result of No. 2 and reports the monitoring result to the system administrator terminal 203. Further, the system administrator terminal 203 has a monitoring result display processing unit 2031, and the monitoring result display processing unit 2031.
Displays the monitoring result data from the result report processing unit 2021. The system administrator uses this system administrator terminal 2
03, the result of monitoring the operating status of the service computer can be obtained. Also, the system administrator terminal 203
The monitoring processing unit 2022 has a monitoring content designating unit 2032 for designating the monitoring content to the management computer 202. The monitoring processing unit 2022 creates an inspection schedule according to the monitoring content designated by the monitoring content designating unit 2032, and provides a service in accordance with the schedule. Part 2
The inspection of 012 is performed.

However, as shown in FIG. 2 (b), this configuration is similar to that of the service providing unit 201 of the service computer 1031.
The service computer 1031 may be provided with the monitoring processing unit 2011 for monitoring the operation status of the second and determining the monitoring schedule. In this case, the result report processing unit 2
Reference numeral 021 reports the monitoring result of the monitoring processing unit 2011 to the system administrator terminal 203. The monitoring content specifying unit 2032 specifies the monitoring content to the monitoring processing unit 2011 via the management computer 202.

Hereinafter, such a monitoring processing unit 2011/2 will be described.
022 will be described in detail.

FIG. 3 shows the configuration of the monitoring processor 2011/2022.

In the figure, reference numeral 301 denotes a monitoring processing unit 2011/2.
Reference numeral 022 denotes a computer provided, and the service computer 1031 corresponds to the computer shown in FIG. 2B, and the management computer 202 corresponds to the computer shown in FIG. 2A.

The computer 301 has, for example, a main memory 302
, A bus 303, a fixed disk 304, a processor 305, an I / O control unit 306, and a communication control unit 307. The processor 305 is a computer
1 is a device for controlling between hardware and processing programs. The fixed disk 304 is an auxiliary storage device that stores programs and tables. The main memory 302 is a device that loads a program and stores temporary data. The I / O control unit 306 controls output to a monitor and input from a keyboard, and the communication control unit 307 controls a line to another computer.

The fixed disk 304 stores an operating system (OS) program 3021 and various programs and tables for realizing the monitoring processor 2011/2022. Monitoring processing unit 2
As programs for realizing 011/2022, a monitoring program 3044, a monitoring schedule creation program 3045, and a monitoring job list table setting program 3046 are stored. Monitoring processing unit 201
As a table for realizing 1/2022, a monitoring job list table 3041, a monitoring job execution time table 3042, and a monitoring job regular activation time table 3043 are stored.

Here, the monitoring program 3044 corresponds to FIG.
As shown in the figure, the monitoring job modules 402 to 40 are programs provided to correspond to the monitoring items, and are programs for realizing a monitoring job process for the corresponding monitoring items.
4 and a monitoring job module activation module 401 that activates the monitoring job modules 402 to 404.

The monitoring job list table 3041
Is configured as shown in FIG.

In the figure, reference numeral 501 denotes a job ID field describing the ID of a monitoring job, and reference numeral 502 denotes a job ID.
A job name field describing the name of the monitoring job described in the field 501, a reference numeral 503 is a cycle field describing a start cycle of the monitoring job described in the job ID field 501, and a reference numeral 504 is a job ID.
This function field describes the names of the functions (monitoring job modules 402 to 404) that execute the processing of the monitoring job described in the field 501.

The monitoring job execution time table 304
2 is configured as shown in FIG.

In the figure, reference numeral 601 denotes a job ID field describing the ID of a monitoring job, and reference numeral 602 denotes a job ID.
This is an execution time field that describes the required execution time of the monitoring job described in the field 601.

The monitoring job periodic start time table 3
043 is configured as shown in FIG.

In the figure, reference numeral 701 denotes the monitoring program 302.
This is a time field provided for each time zone between the period in which the first monitoring job module activation module 401 is activated and the synchronization period. In the present embodiment, since the cycle for activating the monitoring program 3021 is one minute, the length of each time zone is one minute. Here, the same minutes at different times (hours) are assumed to be in the same time zone, that is, the same schedule is repeated every hour (hour) using one hour as a unit of schedule. Therefore, the number of time fields 701 is 60. In the present embodiment, the time zone is represented by the minute at which the time zone starts.
Is from 0 to 59. Reference numeral 702 denotes a time field 7
I of the monitoring job executed in the time zone corresponding to 01
It is a time job ID list field in which D is stored as a list. Reference numeral 703 denotes a time field 7
The time job execution time total field stores the total execution time of the monitoring job included in the time job ID list 702 corresponding to “01”.

The processor 305 stores the OS program 3021 stored in the fixed disk 304 in the main storage 3.
Under the control of an operating system (OS) realized by reading and executing the monitoring program 3002, the monitoring program 3044, the monitoring schedule creation program 3045, and the monitoring job list table setting program 3046 stored in the fixed disk 304 are read. By executing the monitoring processing unit 2011/2022 on the computer 301, the monitoring job list table setting program 3046 is a program for rewriting the monitoring job list table 3041.
The system administrator can appropriately set the monitoring items of the computer system 103 in the monitoring job list table 3041 by a process of the monitoring job list table setting program 3046. After rewriting the monitoring job list table 3041, the monitoring job list table setting program 3046 activates the monitoring job schedule creation program 3045. The process of the monitoring job schedule creation program 3045 is based on the set monitoring job list table 30
41 and the contents of the monitoring job execution time table 3042 that describes the required execution time of each monitoring monitoring job,
The monitoring job regular start-up time table 3043 that defines the inspection item inspection schedule is updated.

The process of the monitoring job module activation module 401 of the monitoring program 3044
Is activated at regular intervals. In the present embodiment, hereinafter, this cycle will be described as one minute. The process of the monitoring program 3044 determines the monitoring job to be started in that time zone based on the contents of the monitoring job regular startup time table 3043, and starts the monitoring job modules 402 to 404 corresponding to the monitoring job.

Hereinafter, details of such processing will be described.

First, the processing of the monitoring job list table setting program 3046 will be described.

The process of the monitoring job list table setting program 3046 receives an input of a new startup cycle of each monitoring job from the system administrator who monitors the computer system 103, and stores the new startup cycle in the monitoring job list table 3041. It is a program that writes and starts the monitoring job schedule creation program 3045.

Now, the monitoring job list table 3041
, All the monitoring job modules 402 to 404 of the monitoring program 3044 are registered in advance.

As shown in FIG. 8, in the process of the monitoring job list table setting program 3046, first, the contents of the monitoring job list table 3041 are read, and the job ID of all monitoring jobs registered in the monitoring job list table 3041 is read. Create List L1 (S80
2). Next, from the top of the job ID list L1,
D are taken out one by one and the following S80
4 to S806 are performed (S803, 806).

That is, the job name 502 and the period 503 of the extracted monitoring job are stored in the monitoring job list table 30.
41 via the monitoring content designation unit 2032 of the system administrator terminal 203.
The monitoring job name and cycle are displayed above. At the same time, a character string is displayed so that the system administrator can input a new cycle with the keyboard of the system administrator terminal 203, and waits until the system administrator inputs the new cycle with the keyboard (S804). Next, a new cycle input from the system administrator via the monitoring content specifying unit 2032 of the system administrator terminal 203 is read, and the new cycle value is overwritten in the cycle field 503 corresponding to the job ID in the monitoring job list table 3041. To be stored (S80
5). Here, in the present embodiment, for a monitoring job that is not executed, 0 is accepted as a cycle from the system administrator. It is assumed that the system administrator specifies only a divisor of 60 as the cycle of the monitoring job to be executed.

Then, for all the job IDs, it is determined whether or not the above process has been completed.
It is determined based on whether or not the D list L1 is empty.
The process returns to 803 (S806).

With the above processing, a new cycle for all monitoring jobs is stored in the monitoring job list table 3041. After that, the monitoring job schedule creation program 3045 is activated (S807), and the process ends.

Monitoring job schedule creation program 3
No. 045 reflects the contents of the new monitoring job list table 3041 in the monitoring job regular activation time table 3043.

The processing of the monitoring job schedule creation program 3045 will be described below.

FIG. 9 shows a processing procedure of the monitoring job schedule creation program 3045.

As shown in the figure, the process of the monitoring job schedule creation program 3045 first checks whether or not the monitoring job list table 3041 has been updated since the last time the monitoring job schedule creation program 3045 was started. If not, the process ends (S902).

On the other hand, if it has been updated, each job ID field 601 of the monitoring job execution time table 3042
And the execution time 602, and examines the job ID for which the corresponding value of the execution time 602 is not registered (S90).
3). And the job ID for which the execution time is not registered
If there is, the monitoring job modules 402 to 404 corresponding to the monitoring job are executed, and the time difference between the execution end time and the execution start time is calculated to measure the execution time of the monitoring job (S904). ). Then, the measurement result is added to the monitoring job execution time table 3042 (S905). The execution times of all the monitoring jobs are stored in the table 3042 by the processing of S903 to S905 described above.

Then, next, for each monitoring job described in each job ID field 501 of the monitoring job list table 3041, the corresponding period field 50
The execution time zone is determined so as to match the cycle described in No. 3 and the monitoring job execution time table 3042 is updated (S906). Details of the execution time zone determination processing of each monitoring job in S906 will be described later.

Then, for each execution time zone, the total execution time of the monitoring job to be executed is calculated by the monitoring program 30.
For a monitoring job registered in a time zone longer than the startup cycle of the monitoring job module startup module 401 (one minute in the present embodiment), a process of shifting one set of monitoring job to the preceding or following time zone is performed (S907). ).

For example, when the contents of the monitoring job regular activation time table 2043 are as shown in FIG. 10, the following processing is performed.

In FIG. 10, reference numeral 1101 denotes time 5
8 is a list of monitoring module groups executed in the time zone of 8 (minutes), and indicates that the monitoring jobs 001, 002, and 004 are sequentially activated after the start of the time zone of time 58 (minutes). Similarly, reference numeral 1102 denotes a monitoring module that is executed in the time zone of time 59 (minute).
This indicates that the monitoring job 003 is started after the start of the (minute) time zone.

In such a schedule, as shown in FIG. 11, the total execution time (65 seconds) of the monitoring job registered in the time zone of time 58 (minutes) and the execution time of the monitoring job registered in the time zone of time 59 (minutes) are set. Monitoring job total execution time (20 seconds)
There is a period in which. During this period, the two monitoring modules will operate in parallel. And
In this case, during this time, the load for the two monitoring modules is applied to the computer system 103.

Therefore, in such a case, the monitoring job periodic start time table 2043 is modified as shown in FIG.

That is, since the total execution time of the monitoring job in the time zone of the time 58 (minute) is larger than the activation cycle of the monitoring program 3044 (1 minute in this embodiment), the time 58 (minute) in FIG. As shown in FIG. 12, the monitoring job 001 that is to be started in the time zone is shifted to the previous time, and is started in the time zone of time 57 (minutes).

FIG. 13 shows a monitoring module scheduled to operate at each time when the monitoring module is executed according to the schedule shown in FIG. Time 57
The execution period of the monitoring job in the time zone of (minute) and the time 58
The execution period 1302 of the monitoring job in the time zone of (minute) and the execution period 1303 of the monitoring job in the time zone of 59 (minute)
Are not overlapped, and the computer system 10
3 can be reduced to one monitoring module.

The process of adjusting the monitoring job execution time back and forth in S907 will be described later.

Returning to FIG. 9, when this step S907 is completed, the monitoring job schedule creation program 30
The process at 45 is ended.

The details of the process of determining each monitoring job execution time zone in S906 of FIG. 9 will be described below.

FIG. 14 shows the procedure of this processing.

As shown in the figure, in this processing, first, the time field 7 of the monitoring job activation time table 3043 is set.
For the entire time zone of 01, the corresponding start job ID list 702 is emptied, and the corresponding time job execution time 703 is initialized to 0 (S1402). Next, the content of the monitoring job list table 3041 is read, and the job ID list L2 including all job IDs except for the job ID of the monitoring job for which the cycle 503 is set to 0, that is, for which the execution is not executed, is set. Is made (S1403). Further, the contents of the monitoring job execution time table 3042 are read, and the job ID list L2 is rearranged in descending order of the execution time (S1404). Then, the job ID list L2
From the job IDs one by one (S14)
05), for each of the following S1406 to S14
Step 12 is performed.

That is, the monitoring job list table 30
41 and the monitoring job execution time table 3042, S1
A cycle 503 (C) and an execution time 602 (E) corresponding to the job ID (J) extracted in 405 are read (S1406). Next, a list L3 of integers equal to or larger than 0 and smaller than C is created (S1407). For example, if the cycle is 3, 0, 1, and 2 are the contents of the list L3.

Next, one integer is sequentially extracted from the list L3 (the extracted integer is defined as S), and the following S1409 to S1410 are performed for each S.

That is, N is an integer equal to or greater than 0, and time T
1 = S + N * C <59 For each time zone of T1 that satisfies 59, the time job execution time total corresponding to the time zone of T1 is 7
G (T1) obtained by adding the value F (T1) of 03 to the execution time E
= F (T1) + E is calculated (S1409). In other words, if the cycle C of the monitoring job J being taken out is 15 and S being taken out is 0, T1 is 0, 15, 3
The time zone is 0, 45. Then, when the execution time E of the monitoring job J is 20, if the total time job execution time F (0) in the time zone of 0 is 10, then G (0) in the time zone of 0
Is 30, and if the total time job execution time F (15) in the 15 time zone is 0, G (15) in the 15 time zone is 10.

Then, in each time zone T2 of the time field 701, for each time zone of T2 satisfying the time T2 = S + N * C (N is an integer of 0 or more), H (T2) = G
(T2) = F (T2) + E, and for each other time zone T2, H (T2) = (corresponding time job execution time total F
(T2)). Then, H (T
The maximum value M of 2) is obtained (S1410). Here, H (T
2) is a time field 701 when it is assumed that the monitoring job J is executed in a time zone satisfying S + N * C.
Is the total time job execution time 703 in each time zone T2, M is the maximum value of the total time job execution time in this case,
That is, the value of the load in the time zone where the load is the largest is obtained.

Now, for each S in the list L3, S14
After performing the processing from 09 to S1410, a list L4 of S that minimizes the maximum value M (S) of H (T2) for each S obtained for each S in the list L3 is obtained (S
1411). As a result, the list L4 contains the monitoring job J
Is set in a time zone satisfying S + N * C, a set of S in which the load value is the minimum in a time zone with the largest load. Therefore, when only one S is included in the list L4, that S is extracted, and when a plurality of S are included, one S selected at random is extracted, and time = S + N * C (N is For each time T3 that satisfies an integer of 0 or more, the time job ID list 7 corresponding to the time T3 in the monitoring job regular activation time table 3043
02, the job ID of the monitoring job J is added, and E is added to the total time job execution time 703 (S
1412).

By performing the processing of steps S1406 to S1412 for all the monitoring job IDs to be executed, all the monitoring job IDs to be executed become
043 and registered in one of the time job ID lists 702, and the monitoring job
The total time job execution time 703 of each time zone of No. 3 is calculated and registered.

The monitoring job periodical start time table 3043 created in this way indicates that each monitoring job is started in the cycle specified in the monitoring job list table and is executed in parallel at an arbitrary time. Thus, the maximum number of monitoring jobs to be monitored is kept low. However, at this stage, it is not yet limited so that the monitoring job is not executed in parallel at an arbitrary time. Therefore, as described above, after this, S9 in FIG.
In step 07, an adjustment process is performed before and after the monitoring job execution time.

Hereinafter, as described above, S90 of FIG.
The details of the process of adjusting the monitoring job execution time back and forth of No. 7 will be described.

FIG. 15 shows the procedure of this process.

As shown, in this process, first, the time field 7 of the monitoring job activation time table 3043 is set.
01, and the time job execution time total 70
3 is compared with the execution time threshold B, and a list L5 of the times when the total time job execution time 703 is larger than the execution time threshold B
Is created (S1502). Here, the execution time threshold B is
This is a threshold for determining whether or not to perform an adjustment process such as adjusting the monitoring job forward and backward only at times when the total execution time exceeds B. This is achieved by describing a value in a program for implementing the adjustment procedure S907. In the present embodiment, the execution time threshold B is set to one minute of the start cycle of the monitoring program.

Next, the leading time zones are extracted one by one from the time list L5, and the extracted time zones are set as T4, and the following processes of S1504 to S1509 are performed (S1503, 1510).

That is, the monitoring job start time table 3
For each monitoring job registered in the time job ID list 702 in the time zone of T4 of 043, the execution time 602 of the monitoring job execution time table 3042 is checked, and the monitoring jobs are rearranged in ascending order of the execution time. Until the total time job execution time 703 becomes smaller than the execution time threshold value B (S1509), one monitoring job J is sequentially extracted from the top, and the following monitoring jobs J are performed on each of them.
5 through S1508 (S1504).

That is, the time zone T4-1 when the monitoring job J is shifted to the time zone T4-1 one minute before the time zone T4.
The time job execution time total S1 and the time job execution time total S2 at time T4 + 1 in the case of shifting to the time zone T4 + 1 one minute after the time zone T4 are calculated (S1505). Then, it is determined whether or not S1 or S2 is equal to or less than the execution time threshold B. If both are not small, the movement of the monitoring job forward or backward is determined by adding the total time job execution time in the preceding and following time zones to the execution time. If it is determined that it cannot be performed because the value exceeds the threshold value B, and there is a time zone next to the time list L5 (step 1510), the process returns to S1503 to perform processing for the time zone next to the time list L5. . On the other hand, if the next time zone of the time list L5 does not exist, the process ends (S1510).

Now, in S1506, S1 or S
If it is determined that 2 is smaller than B, S1 and S2 are compared, and the monitoring job J is moved to the smaller time zone T5 (if one is the same), it corresponds to the time zone T5. The job ID of the monitoring job J is added to the job ID list 702 to be executed, and the job ID of the monitoring job J is deleted from the time job ID list 702 in the time zone T4 (S1507). In addition, in order to update the total time job execution time 703 for each time zone of times T5 and T4, the time job execution time 602 is added to the job IDs registered in the time job ID list 702 to execute the time job execution. The value of the total time 703 is calculated, and the value is written in the total time of job execution time 703 (S1508).

Then, it is determined whether or not the total time job execution time 703 in the time zone T4 has become equal to or less than the execution time threshold value B (S1509). Returning to step S1504, the process for the next monitoring job is performed. On the other hand, if it is below, or if there is no next monitoring job, S150
3, the process for the next time zone of the time list L5 is performed.
On the other hand, if the next time zone of the time list L5 does not exist, the process ends (S1510).

By performing the above processing for all time zones T4 in the list L5, the adjustment before and after the monitoring job is completed for all time zones T4.

The monitoring job periodical start time table 3043 adjusted in this manner indicates that the maximum number of monitoring jobs executed in parallel at any time can be changed by moving the monitoring job to the preceding and following time zones. It is limited to 1 in the range. In addition, the execution start cycle of each monitoring job does not completely match the cycle registered in the monitoring job list table 3041 for those that have moved to the preceding and following time zones. , The shift of the cycle from the cycle registered in the monitoring job list table due to this movement is slightly suppressed. Note that, depending on the range in which this cycle deviation can be tolerated, the time zone that can be moved to the time zone before and after the monitoring job is expanded not only to the time zone immediately before and immediately after, but also to a further time zone. You may.

The details of the processing of the monitoring job schedule creation program 3045 have been described above.

The processing of the monitoring job module activation module 401 of the monitoring program 3044 will be described below.

FIG. 16 shows the procedure of this processing.

As shown, when the monitoring job module activation module 401 is activated by the OS every minute, it first checks the current time (S1602). Next, the time zone indicated by the minute value T6 of the current time is set in the time field 701 of the monitoring job regular activation time file 3043.
The job ID in the time job ID list 702 corresponding to the time zone of T6 is listed in the list L6 as the monitoring job to be started in this time zone (S160).
3).

Next, one job ID is extracted from the list L6, and the monitoring job function name field 504 of the monitoring job list table 3041 corresponding to the job ID is obtained.
In addition to executing the function designated by (1), that is, the monitoring job module, the time required for execution is measured, and the monitoring job execution time table 30 is calculated using the measured time or the average value of the time measured for the monitoring job module.
The execution time 602 of 42 is updated (S1604).

By continuing the above processing until the list L6 becomes empty, all monitoring jobs to be processed in the current time zone are started.

The embodiment of the present invention has been described above.

The above monitoring program 3044,
The individual functions of the monitoring schedule creation program 3045 and the monitoring job list table setting program 3046 or the functions constituting the parts thereof are arranged in an arbitrary arrangement on the service computer 1031, the management computer 202, and the system administrator terminal 203, respectively. May do it.

In this embodiment, a case has been described in which the execution start times of a plurality of monitoring job modules that are periodically and repeatedly executed are scheduled. It is widely applicable when scheduling the execution start time of a task or the like.

[0083]

As described above, according to the present invention, a plurality of jobs (tasks) that are repeatedly executed periodically can be scheduled so as to reduce the load on the system. In particular, it is possible to schedule the periodic monitoring of each monitoring item of the information system so that the influence on the service provided by the information system is reduced.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an overall configuration of a monitoring system to which an embodiment of the present invention has been applied.

FIG. 2 is a block diagram showing a configuration for monitoring the computer system shown in FIG.

FIG. 3 is a block diagram illustrating a configuration of a state monitoring processing unit illustrated in FIG. 2;

FIG. 4 is a diagram showing a configuration of a monitoring program shown in FIG. 3;

FIG. 5 is a diagram showing a configuration of a monitoring job list table shown in FIG.

FIG. 6 is a diagram showing a configuration of a monitoring job execution time table shown in FIG.

FIG. 7 is a diagram showing a configuration of a monitoring job regular activation time table shown in FIG. 3;

FIG. 8 is a flowchart showing a processing procedure of a monitoring job list table setting program shown in FIG. 3;

FIG. 9 is a flowchart showing a processing procedure of a schedule creation program shown in FIG. 3;

10 is an example of a regular startup time table of a monitoring job in which a load on the computer system shown in FIG. 1 increases.

11 is a diagram illustrating execution timing of a monitoring module that increases a load on the computer system illustrated in FIG. 1;

12 is an example of a regular startup time table of a monitoring job in which the load on the computer system shown in FIG. 1 is distributed.

FIG. 13 is a diagram showing execution timing of a monitoring module in which the load on the computer system shown in FIG. 1 is distributed.

FIG. 14 is a flowchart illustrating a procedure of a process of determining a monitoring job execution time in FIG. 9;

FIG. 15 is a flowchart illustrating a procedure of a process of adjusting a monitoring job execution time to a time zone before and after the monitoring job in FIG. 9;

FIG. 16 is a flowchart showing a processing procedure of a monitoring job module activation module shown in FIG. 4;

[Explanation of symbols]

101 ... user terminal, 102 ... network, 103 ...
Computer system, 202: management computer, 203: system administrator terminal, 301: computer on which the monitoring processing unit operates, 302: memory, 303: bus, 304: fixed disk, 305: processor, 306: I / O control unit , 3
07: communication control unit, 401: monitoring job module activation module, 402 to 404: monitoring job module
1031 to 1032: service computer, 1033: service system network, 2011: monitoring processing unit, 2
012: service providing unit, 2021: result report processing unit,
2022: monitoring processing unit, 2031: monitoring result display processing unit, 2032: monitoring content specifying unit, 3021: operating system, 3041: monitoring job list table, 3042: monitoring job execution time table, 3043
... Monitoring job regular start time table, 3044 ... Monitoring program, 3045 ... Monitoring job schedule creation program, 3046 ... Monitoring job list table setting program

 ──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Masatoshi Terada 1099 Ozenji Temple, Aso-ku, Kawasaki City, Kanagawa Prefecture Inside System Development Laboratory, Hitachi, Ltd. 5B098 GC01 GC10 GC16 GD02 JJ08 F term in the Software Division of Hitachi, Ltd. (Reference)

Claims (6)

[Claims] 1. A monitoring method for monitoring a monitoring target using an electronic computer, comprising: a monitoring method for each of a plurality of monitoring processing programs prepared for each of a plurality of items and monitoring the monitoring target for the item; A first step of determining a load required to execute the processing program; a second step of determining a cycle at which the monitoring processing program is to be executed for each of the plurality of monitoring processing programs prepared in advance for each of the plurality of items; Based on the load required to execute each monitoring processing program obtained in the first step and the cycle to execute each monitoring processing program determined in the second step, it is prepared in advance for each of the plurality of items. The load caused by the execution of each of the plurality of monitoring processing programs is temporally distributed, and each of the plurality of monitoring processing programs is in the second state. A third step of creating an execution start schedule for each of the plurality of monitoring processing programs so that the plurality of monitoring processing programs start at the corresponding cycle determined in the step, and the plurality of monitoring processing programs created in the third step. A fourth step of starting execution of each of the plurality of monitoring processing programs according to each execution start schedule. 2. A monitoring device for monitoring a monitoring target, comprising: monitoring required time information describing time required to execute each of a plurality of item-specific monitoring processes for monitoring the monitoring target for a predetermined item; A storage unit for storing monitoring cycle information describing an execution cycle of each of the plurality of item-by-item monitoring processes, and monitoring schedule information that is a schedule for starting execution of each of the plurality of item-by-item monitoring processes; The plurality of item-based monitoring processes are each performed based on the time required to execute each item-based monitoring process described in (1) and the execution period of each item-based monitoring process described in the monitoring period information. The execution is started in the corresponding execution cycle described in the above, and the plurality of items are monitored such that the maximum value of the total execution time of the item-by-item monitoring process started in an arbitrary time zone is minimized. Itemized monitoring process creates a respective start of execution of the schedule, the monitoring apparatus characterized by comprising: a scheduling means, the storing in the storage means the schedule as the monitoring scheduling information. 3. The monitoring device according to claim 2, wherein the scheduler executes each of the plurality of item-based monitoring processes according to the created execution start schedule of the plurality of item-based monitoring processes. If there is a time zone in which the total execution time of the item-by-item monitoring process that is started to be executed in that time zone does not fit within the time zone, at least one item-by-item monitoring that is to be started in that time zone By shifting the execution start time of the process to a time zone before and after the time zone, the total execution time of the item-by-item monitoring process to be executed in the time zone falls within the time zone.
A monitoring apparatus, wherein an execution start schedule of each of the plurality of item-by-item monitoring processes is adjusted. 4. The monitoring device according to claim 2, wherein the scheduling means executes the at least two item-by-item monitoring processes in series during a time period when at least two item-by-item monitoring processes are started. Thus, a monitoring apparatus for creating an execution start schedule for each of the plurality of item-by-item monitoring processes. 5. A storage medium storing a program for monitoring a monitoring target, wherein the program is read from the storage medium by an electronic computer and executed to execute monitoring of a monitoring target for each predetermined item. Monitoring required time information describing the time required to execute each of the plurality of item-based monitoring processes, monitoring cycle information describing the execution cycle of each of the plurality of item-based monitoring processes, and the plurality of item-based monitoring processes Storage means for storing monitoring schedule information which is a schedule for starting each execution; time required for execution of each of the plurality of item-specific monitoring processes described in the monitoring required time information; and description in the monitoring cycle information Based on the execution cycle of each of the plurality of item-by-item monitoring processes, the corresponding one of the plurality of item-by-item monitoring processes described in the monitoring cycle information. To start running in the row cycle,
In addition, a schedule for starting execution of each of the plurality of item-by-item monitoring processes is created so that the maximum value of the total execution time of the item-by-item monitoring processes to be started in an arbitrary time zone is minimized, and the schedule is defined as And a scheduling means for storing the monitoring schedule information in the storage means on the computer. 6. A method for scheduling a plurality of jobs or tasks to be executed at a predetermined cycle on a computer, comprising: a first step of determining a time required for executing each of the plurality of jobs or tasks; Based on the time required for execution of each of the plurality of jobs or tasks determined in the first step and the execution cycle of each of the plurality of jobs or tasks, the execution cycle corresponding to each of the plurality of jobs or tasks is A second step of scheduling the execution start time of each of the plurality of jobs or tasks such that the maximum value of the total execution time of the jobs or tasks to be started and started to be executed in an arbitrary time zone is minimized And a scheduling method comprising: