JPH06222963A - High-load resource evaluation system - Google Patents

Abstract

PURPOSE:To easily and surely execute performance evaluation work whoever operator may be by automatically deciding/extracting whether any resource generating a performance problem is existent or not when operating a computer system, and outputting the result to an output device. CONSTITUTION:This system is provided with a high-load resource selecting means 5 for comparing each record in a system operating condition recording file 1 with a resource use rate reference value composed of a warning value and a critical value for performance guarantee and extracting the name of the resource, which does not satisfy this value, and recording time, and a high- load resource information record generating means 8 for extracting the information of the other resource name and a resource use rate recorded at the same time as the recording time of the record extracted by this high-load resource selecting means 5 and the information of an executed job name and a resource use rate, integrating these pieces of information while corelating them each other, and outputting it to a high-load resource information file 9 as one high- load resource information record.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a high load resource evaluation system in operation management of a computer system.

[0002]

2. Description of the Related Art FIG. 2 shows the contents of a system operation status recording file. The contents of the system resource use status record that constitutes this system operation status recording file are shown in FIG. 2- (a), and the contents of the job execution status record are shown in FIG. 2- (b).

In the system operation status recording file, a group of programs having a system monitoring function which is a subsystem of the operating system records the usage status of each resource constituting the computer system as a system resource usage status record at regular time intervals. Further, the execution history of each job executed on the system is recorded as a job execution status record at regular time intervals. These records are stored in chronological order in a sequential file format.

For example, in the system resource usage status record of the CPU, information such as record ID, record length, sampling time, CPUID, and CPU usage time is recorded, and the system resource usage status of the main storage device (hereinafter referred to as memory) is recorded. The record records information such as record ID, record length, sampling time, main memory usage of system program and number of page faults, main memory usage of user program and number of page faults, and an external storage device (hereinafter referred to as DISK). The system resource usage record of (1) is recorded with information such as record ID, record length, sampling time, DISK name, device identification name, and input / output count, and is stored in the external storage device control device (hereinafter referred to as MSP). The system resource usage record is record I
Information such as D, record length, sampling time, device identification name, and input / output count is recorded, and the system resource usage record of the file includes record ID, record length, sampling time, file name, and input / output count. The information of is recorded.

Further, the job execution status record includes information such as record ID, record length, sampling time, job identification number, CPU usage time, step number, maximum main memory usage, total page fault count, and input / output count. It is recorded.

As a technique for inputting the data of the system operating condition recording file to analyze the operating condition of the computer system, Japanese Patent Laid-Open No. 63-146138, "Automatic diagnosing apparatus for computer system operating condition", also system operating condition recording file JP-A-63-189960 "System Performance Improvement Method" is a technique for monitoring whether or not a performance problem has occurred by inputting the data of FIG.

[0007]

Problems to be Solved by the Invention JP-A-63-1461
In Japanese Patent No. 38, the operation data processing mechanism inputs the data of the system operation condition recording file to calculate the load condition of each resource, and the system performance analysis / evaluation mechanism performs the system performance evaluation. However, the specific content of this process is not disclosed at all. Also in the drawing, only the data of the system operation status recording file edited and output in time series is presented, and the result of the performance analysis / evaluation, that is, the resource under the high load during the system operation is pointed out. There was a problem that it was not possible to point out the bottleneck in which the resource was under heavy load.

In Japanese Patent Laid-Open No. 63-189960, the system performance data collecting means collects performance data in the system, and the system performance data monitoring means compares the performance data with a performance reference value to cause a performance problem. Although there is a description of the process of determining whether or not it is performed, the contents of the performance reference value, the method of holding the reference value, and the contents of the collation process of the performance data and the performance reference value are not disclosed at all. Also, since the performance analysis / evaluation results are not displayed and output, it is not possible to point out a resource that was under a high load when the system was operating, and it is not possible to point out a bottleneck where that resource was under a high load. there were.

An object of the present invention is to input a system operation status record record and compare it with a preset resource usage rate reference value to check whether a heavy load condition has occurred on each resource constituting the computer system. Even if the user is not familiar with performance evaluation work, it is easy to determine whether or not the bottleneck that caused the resource to be under heavy load is identified and the result is presented to the user. In addition, to ensure that performance evaluation work can be carried out.

[0010]

A high load resource evaluation system according to a first aspect of the present invention records the usage status of each resource constituting a computer system as a system resource usage status record at regular time intervals, and further, the computer system. A system operation status recording file that records the execution history of each job executed above as a job execution status record at regular time intervals, and connection relationship information between the DISK that constitutes the computer system and the MSP that controls this DISK. , The D
A system configuration file that stores the file name information stored in the ISK as system configuration data, a system operation status record record input means for inputting each record in the system operation status record file, and each of the computer systems A resource utilization reference value file that stores a warning value and a limit value for performance assurance of resource utilization as a resource utilization reference value, and the resource of the system resource utilization record input by the system operation record recording input means The resource name and recording time of the system resource usage record that indicates the high-usage usage status that exceeds the limit value of the resource usage reference value in this resource usage reference value file by comparing the usage rate information , And high-load resource selecting means for extracting the resource usage rate, and the system operation status record record inputting means. After the high load resource selecting means is started, the recording time of the system resource use status record extracted by the high load resource selecting means and the job input by the system operation status record record input means recorded at the same time The job name in the execution status record, the resource usage rate information for each resource, the other resource name in the system resource usage status record input by the system operation status record input means, and the resource usage rate information are extracted,
By referring to the system configuration data of the system configuration file, information on each resource and information on a job using the resource, or information on each resource and information on resources constituting the resource are integrated and associated with each other. It comprises a high load resource information record generating means which is one high load resource information record, and a high load resource information file which stores the high load resource information record generated by the high load resource information record generating means.

The high load resource evaluation system of the second invention is
In the high-load resource evaluation system according to the first aspect of the present invention, the resource usage rate reference value in the resource usage rate reference value file is a preset CPU usage rate, memory page fault count,
It is composed of a warning value and a limit value for performance guarantee with respect to the input / output times of DISK and MSP.

The high load resource evaluation system of the third invention is
In the high load resource evaluation system of the first invention, the high load resource information record in the high load resource information file is
It is composed of an outline part and a detail part. The outline part indicates a record ID indicating whether it is CPU information, memory information, DISK information, or MSP information, a sampling time, a position of the detail part, and a size of the detail part. In the detailed portion, when the record ID is CPU, the CPU usage rate and the CPU usage rate for each job, when the record ID is memory, the page fault frequency and the page fault frequency for each job, and the record ID is DISK. , The number of input / output times of this DISK and the number of input / output times for each file in the DISK, and MSP if the record ID is MSP information.
And the number of inputs and outputs of each DISK connected to the MSP.

The high-load resource evaluation system of the fourth invention is
In the high load resource evaluation system of the first invention, the high load resource information record stored in the high load resource information file is input, and a high load resource determination result output means for displaying a graph on an output device is provided. There is.

The high load resource evaluation system of the fifth invention is
In the high load resource evaluation system according to the first aspect of the present invention, the graph displayed by the high load resource determination result output means indicates the usage rate of each resource at each recording time between the warning value and the limit value of the resource usage rate reference value. In addition to those displayed in different characters, the resource usage rate for each job that uses each resource at the recording time, or the input / output count for each DISK for each file is displayed side by side. .

[0015]

An embodiment of the present invention will be described in detail below with reference to the drawings.

Referring to FIG. 1, a high load resource evaluation system according to a first embodiment of the present invention includes a system resource use status record for recording the use status of each resource constituting a computer system at regular time intervals. , A system operation status recording file 1 including a job execution status record for recording an execution history of each job executed on the computer system
And the DISK that composes the computer system and this DISK
Information related to connection with the MSP that controls the system, a system configuration file 7 that stores the file name information stored in the DISK as system configuration data, and a system operation status record that inputs each record in the system operation status record file 1. A record input means 2, a first work area 3 for storing each record in the system operation status record file 1 input by the system operation status record record input means 2, and each resource constituting the computer system, that is, a CP
U utilization, memory page fault count, DISK,
And a resource usage rate reference value file 4 for storing a warning value and a limit value for performance guarantee with respect to the number of input / output times of the MSP as a resource usage rate reference value, and a system resource usage status record stored in the first work area 3. The resource usage rate information is compared with the limit value of the resource usage rate reference value in this resource usage rate reference value file 4, and the resource name and record of the system resource usage status record indicating a high load usage state exceeding this limit value A high load resource selecting means 5 for extracting the time, and a resource name and recording time of the system resource use status record showing the high load use status extracted by the high load resource selecting means 5 and a second for storing the resource usage rate. Of the system resource usage status record stored in the second work area 6 after starting the work area 6, the system operation status record record input means 2, and the high load resource selection means 5. Record time of, was recorded at the same time,
The job name in the job execution status record input by the system operation status record record input means 2, the resource usage rate information for each resource, the resource name in the system resource usage status record input by the system operation status record record input means 2, And resource usage rate information are extracted from the first work area 3, and by referring to the system configuration data in the system configuration file 7, a summary section and a detail section are provided.
PU information, memory information, DISK information, or M
Record ID indicating SP information, sampling time,
The position and size of the detail part are shown. The detail part indicates the CPU usage rate and the CPU usage rate for each job when the record ID is CPU, and the page fault frequency and each when the record ID is memory. Number of page faults for each job, this DIS if the record ID is DISK
K input / output count, input / output count for each file in the DISK, and M when the record ID is MSP information
Number of SP input / output and each DIS connected to the MSP
High-load resource information record generating means 8 for generating one high-load resource information record composed of K input / output times.
And a high load resource information file 9 storing the high load resource information record generated by the high load resource information record generating means 8.

FIG. 3 shows a configuration example of the system configuration file 7. The system configuration data in the system configuration file 7 is information stored in the file by editing the system configuration information held by the computer system in units of DISK before starting the high load resource evaluation system. It includes the number of MSPs to which the DISK is connected and the device identification name, and the number of files and file names set in the DISK.

Next, the operation of the first embodiment of the present invention will be described with reference to FIGS.

First, the user activates the high load resource evaluation system using the "evaluation target section" as a parameter in the format of "evaluation start time-evaluation end time".

The high load resource evaluation system activates the high load resource information record generating means 8 with the "evaluation target section" as a parameter.

The high load resource information record generation means 8 first activates the system operation status record record input means 2 with the "evaluation target section" as a parameter (step 8).
1).

System operation status record record input means 2
First, it judges whether or not there is a record in the system operation status recording file 1 (step 21), and if there is a record, reads one record (step 22). If there is no record in step 21, an error is returned to the high load resource information record generating means 8 (step 23).

Next, the system operation status record record input means 2 determines whether or not the one record read in step 22 is a record recorded in the "evaluation target section" received from the high load resource information record generation means 8. Is determined (step 24), and if the record is recorded in the “evaluation target section”, the record is written in the first work area 3 (step 25).

System operation status record record input means 2
After step 25, it judges whether or not there is a record in the system operation status recording file 1 (step 26), and if there is a record, returns to step 22.

System operation status record record input means 2
If the record is not a record recorded in the “evaluation target section” in step 24, step 26
To

System operation status record record input means 2
If there is no unevaluated record in the system operation status recording file 1 in step 26, it judges whether or not there is a record in the first work area 3 (step 27), and the first work area 3 If there is no record in the system operation status record file 1, the "evaluation target section"
An error that there is no record recorded inside is returned to the high load resource information record generating means 8 (step 28). This error is further returned from the high load resource information record generation means 8 to the high load resource evaluation system, and the user is requested to redesignate the “evaluation target section”.

System operation status record record input means 2
Ends the process if there is a record in the first work area 3 in step 27.

The high load resource information record generation means 8 activates the high load resource selection means 5 after the system operation status record record input means 2 ends (step 82).

FIG. 4 shows an example of resource usage rate reference values in the resource usage rate reference value file 4. The resource usage rate reference value file 4 is set in advance prior to starting the high load resource evaluation system.
Usage rate of each resource in the system resource usage status record, specifically, CPU usage rate, memory page fault count,
It is composed of a warning value and a limit value for performance guarantee with respect to the input / output times of DISK and MSP.

The high load resource selection means 5 first reads one record from the first work area 3 (step 51).

The high load resource selection means 5 determines whether the record read in step 51 is a CPU record, a memory record, a DISK record, or MSP.
It is determined whether or not the record is (step 52).

When the record determined in step 52 is a CPU record, the high load resource selection means 5 divides the CPU usage time of the record by the data collection time interval to calculate the CPU usage rate (unit:%). ) (Step 53), and it is further determined whether or not the CPU usage rate exceeds the limit value of the resource usage rate reference value of the CPU in the resource usage rate reference value file 4 (step 54). If it exceeds the second, the resource name (CPU), sampling time, and CPU usage rate are regarded as one record
Is written in the work area 6 (step 55).

The high load resource selection means 5 determines whether the CPU usage rate does not exceed the limit value in step 54 or whether there is data in the first work area 3 after step 55. Judge (step 56),
If there is a record, the process returns to step 51. If there is no unevaluated record in the first work area 3 in step 56, the process ends.

When the record determined in step 52 is a record in memory, the high load resource selecting means 5 determines the number of page faults of the system program and the number of page faults of the user program of the record, respectively, at data collection time intervals. And then convert to page faults per second (unit is times / sec) (step 57), and the system program page faults per second and user program page faults per It is judged whether or not the limit value of the resource usage rate reference value of the memory in the value file 4 is exceeded (step 58), and if at least one of them exceeds the limit value, it is determined as a resource name (memory). Sampling time and number of system program page faults per second The page fault count over The program is written as one record in the second work area 6 (step 59).

The high load resource selection means 5 determines in step 58 that neither the number of page faults of the system program nor the number of page faults of the user program per second exceeds the limit value, or step 59.
After the end, it is determined whether or not there is data in the first work area 3 (step 5A), and if there is a record, the process returns to step 51. If there is no unevaluated record in the first work area 3 in step 5A, the process ends. When the record determined in step 52 is a DISK record, the high-load resource selection means 5 divides the input / output count of the record by the data collection time interval to obtain the input / output count per second (unit: Number of times /
Seconds) (step 5B), and the number of I / Os per second is DISK in the resource usage rate reference value file 4.
If it exceeds the limit value of the resource usage rate reference value of (step 5C), if it exceeds the limit value, the resource name (DISK), the sampling time, and the number of I / Os per second are set. It is written in the second work area 6 as one record (step 5D).

The high load resource selection means 5 determines whether or not the number of input / output operations per second does not exceed the limit value in step 5C, or whether or not there is data in the first work area 3 after step 5D. Is determined (step 5
E), if there is a record, return to step 51. If there is no unevaluated record in the first work area 3 in step 5E, the process ends.

When the record determined in step 52 is an MSP record, the high load resource selection means 5 divides the input / output count of the record by the data collection time interval to obtain the input / output count per second. (Unit is number /
Seconds) (step 5F), and it is further determined whether or not the number of input / outputs per second exceeds the limit value of the MSP resource utilization reference value in the resource utilization reference value file 4 (step 5G). ), If it exceeds the limit value, the resource name (MSP), the sampling time, and the number of input / output operations per second are written as one record in the second work area 6 (step 5H).

The high-load resource selection means 5 determines whether or not the number of input / output operations per second does not exceed the limit value in step 5G, or whether or not there is data in the first work area 3 after step 5H. Is determined (step 5
I), if there is a record, the process returns to step 51. If there is no unevaluated record in the first work area 3 in step 5I, the process ends.

In the high load resource selecting means 5, the record determined in step 52 is the CPU, the memory, the DIS.
If the record is neither K nor MSP, it is determined whether or not there is data in the first work area 3 (step 5J). If there is a record, step 5 is executed.
Return to 1. If there is no unevaluated record in the first work area 3 in step 5I, the process ends.

FIG. 5 shows the format of the high load resource information record. The high load resource information record has a record I in the overview section.
D, the sampling time, the position of the detail section 1, the size of the detail section 1, the position of the detail section 2, the size of the detail section 2, and the number of the detail sections 2, and the detail section includes the detail section 1 and the detail section 2. Part 2
It is composed of a number of detail units 2. The overview portion has a fixed length, but the detail portion has a variable length, and the high-load resource information record as a whole is a variable length record.

The process of generating the high load resource information record by the high load resource information record generating means 8 will be described below with reference to FIG.

The high load resource information record generating means 8 determines whether or not there is a record in the second work area 6 after the high load resource selecting means 5 is finished (step 83). finish.

If there are records in the second work area 6 in step 83, the high load resource information record generating means 8 sorts these by sampling time (step 84).

Next, the high load resource information record generating means 8
Reads one record from the second work area 6 (step 85) and refers to the resource name of the record to determine whether the record is a CPU, a memory, a DISK, or an MSP (step). 86).

If the record determined in step 86 is a CPU record, the high load resource information record generating means 8 searches the first work area 3 using the sampling time of the record as a key, and the same. All the job execution status records at the sampling time are read (step 87), and for each record, group data is created that is composed of the job identification number and the CPU usage rate obtained by dividing the CPU usage time by the data collection time interval (step). 88).

Next, the high load resource information record generating means 8
Follows the format of the high load resource information record shown in FIG.
U ”, the sampling time in the“ sampling time ”field, the start address of the detailed part 1 in the“ position of detail part 1 ”field, and C in the“ size of detail part 1 ”field.
The size of the CPU usage rate data of the PU record, the start address of the detail section 2 in the "position of detail section 2" field, and the size of the job combination data created in step 88 in the "size of detail section 2" field Then, the number of the set data is stored in the "number of detail parts 2" field (step 89). The start address of the detail section 1 can be obtained as the address next to the address obtained by adding the size of the outline section having a fixed length to the start address of the high load resource information record, and the start address of the detail section 2 is It can be obtained as the next address of the address obtained by adding the size of the detail portion 1 to the start address of.

Highly loaded resource information record generation means 8
Stores the CPU usage rate of the CPU record in the “Details section 1” field of the details section, and stores the set data of the jobs one by one in the descending order of CPU usage rate in the “Details section 2” field (step 8A). The completed high load resource information record is written in the high load resource information file 9 (step 8).
B).

After the completion of step 8B, the high load resource information record generation means 8 determines whether or not there is a record in the second work area 6 (step 8C). If there is a record, the process returns to step 85 and the record is stored. If not, the process ends.

When the record determined in step 86 is a record in the memory, the high load resource information record generation means 8 searches the first work area 3 using the sampling time of the record as a key, and the same. Read all job execution status records at the sampling time (step 8D), and create group data consisting of the job identification number and the page fault count per second obtained by dividing the page fault count by the data collection time interval for each record. (Step 8E).

Next, the high load resource information record generating means 8
According to the format of the high load resource information record shown in FIG. 5, first, "memory" is set in the "record ID" field of the overview section, sampling time is set in the "sampling time" field, and details are set in the "position of detail section 1" field The start address of part 1 is set to the size of detail part 1 field and the size of the sum of the data of the page fault frequency of the system program and the data of the page fault frequency of the user program per second of the memory record is added. , The start address of the detail section 2 in the "position of detail section 2" field, the size of the set data of the job created in step 8E in the "size of detail section 2" field, and the "number of detail section 2" field The number of sets of data is stored in (step 8F). The start address of the detail section 1 can be obtained as an address next to the address obtained by adding the size of the outline section having a fixed length to the start address of the high load resource information record, and the start address of the detail section 2 is
It can be obtained as an address next to the address obtained by adding the size of the detail section 1 to the start address of the detail section 1.

Highly loaded resource information record generation means 8
Stores the data of the number of system program page faults per second and the data of the number of user program page faults per second of the memory record in the "Details 1" field of the details section, and the job in the "Details 2" field. The group data of 1 is stored one by one in descending order of the number of page faults per second (step 8G), and the completed high load resource information record is written in the high load resource information file 9 (step 8H).

After the step 8H, the high load resource information record generating means 8 determines whether or not there is a record in the second work area 6 (step 8I). If there is a record, the process returns to the step 85 and the record is stored. If not, the process ends.

If the record determined in step 86 is a DISK record, the high load resource information record generating means 8 first searches the system configuration data of the system configuration file 7 using the DISK name of the record as a key. Then, all the file names set in the DISK are read (step 8J), and the first work area 3 is searched by using the file name and the sampling time of the DISK record as a key, and the files of the same sampling time are searched. All the system resource usage records are read (step 8K), and a set data is created for each record consisting of the file name and the input / output count per second obtained by dividing the input / output count by the data collection time interval (step 8L).

Next, the high load resource information record generating means 8
Follows the format of the high load resource information record shown in FIG.
SK ”, the sampling time in the“ sampling time ”field, and the detail section 1 in the“ position of detail section 1 ”field.
, The size of the data of the number of input / output times per second of the DISK record in the “size of detail section 1” field, the start address of detail section 2 in the “position of detail section 2” field, and the “detail section” The size of the set data of the file created in step 8L is stored in the "size of 2" field, and the number of the set data is stored in the "number of detail parts 2" field (step 8M). The start address of the detail section 1 can be obtained as the address next to the address obtained by adding the size of the outline section having a fixed length to the start address of the high load resource information record, and the start address of the detail section 2 is It can be obtained as the next address of the address obtained by adding the size of the detail portion 1 to the start address of.

Highly loaded resource information record generation means 8
Stores the data of the input / output count per second of the DISK record in the “detail part 1” field of the detail part, and sets the set data of the file in the “detail part 2” field in descending order of the number of input / output times per second. Each of them is stored (step 8N), and the completed high load resource information record is written in the high load resource information file 9 (step 8O).

After the step 8O, the high load resource information record generation means 8 determines whether or not there is a record in the second work area 6 (step 8P). If there is a record, the process returns to the step 85 to record the record. If not, the process ends.

When the record determined in step 86 is an MSP record, the high load resource information record generating means 8 first searches the system configuration data in the system configuration file 7 using the MSP name of the record as a key. Then, all the DISK names connected to the MSP are read (step 8Q), and the DISK name and M
The first work area 3 is searched using the sampling time of the SP record as a key, and the DIS at the same sampling time is searched.
All the K system resource usage records are read (step 8R), and for each record, group data consisting of the DISK name and the input / output count per second obtained by dividing the input / output count by the data collection time interval is created. (Step 8S).

Next, the high load resource information record generating means 8
Follows the format of the high load resource information record shown in FIG.
P ", the sampling time in the" sampling time "field, the start address of the detailed portion 1 in the" position of detail portion 1 "field, and M in the" size of detail portion 1 "field.
The size of the data of the number of input / output times per second of the SP record, the start address of the detail section 2 in the "position of detail section 2" field, and the set of DISK created in step 8S in the "size of detail section 2" field Data size,
Then, the number of the set data is stored in the "number of detail parts 2" field (step 8T). The start address of the detail section 1 can be obtained as an address next to the address obtained by adding the size of the outline section having a fixed length to the start address of the high load resource information record, and the start address of the detail section 2 is
It can be obtained as an address next to the address obtained by adding the size of the detail section 1 to the start address of the detail section 1.

Highly loaded resource information record generation means 8
Stores the data of the number of input / output times per second of the MSP record in the “detail part 1” field of the detail part, and stores one set of DISK data in the “detail part 2” field in descending order of the number of input / output times. Each of them is stored (step 8U), and the completed high load resource information record is written in the high load resource information file 9 (step 8V).

After the step 8V, the high load resource information record generating means 8 determines whether or not there is a record in the second work area 6 (step 8W). If there is a record, the process returns to the step 85 to record the record. If not, the process ends.

As described above, the processing of the high load resource evaluation system, which is the operation of the first embodiment of the present invention, is completed.

The high load resource evaluation system according to the first embodiment of the present invention inputs a system operation status record and compares it with a preset resource utilization rate reference value,
Performance is evaluated by determining whether each resource that makes up the computer system is under heavy load, and by specifying the job name or resource name that put the resource under heavy load as a bottleneck. It has an effect that even a user who is not familiar with the work can easily and surely perform the performance evaluation work.

Referring to FIG. 9, the high load resource evaluation system according to the second embodiment of the present invention is activated by the high load resource information record generating means 8 in the high load resource evaluation system according to the first embodiment. High-load resource information file 9
Enter the high-load resource information record stored in, and specify the usage rate of each resource at each sampling time using different characters with the warning value and the limit value of the resource usage rate reference value in the resource usage rate reference value file 4 as the boundary. Edited using the third work area 11 so that the resource usage rate for each job that uses each resource at the sampling time, or the input / output count for each file and each DISK is displayed side by side. In addition, the high load resource determination result output means 10 for displaying a graph on the output device 12 is included.

Next, the operation of the second embodiment of the present invention will be described with reference to FIGS. 4, 5 and 9 to 11.

The high load resource information record generation means 8 activates the high load resource determination result output means 10 after the generation of the high load resource information record.

The high load resource determination result output means 10 first sorts the high load resource information records in the high load resource information file 9 by sampling time and record ID (step 101).

Next, the high load resource determination result output means 10
One record with the record ID "CPU" is read (step 102), and the CPU usage rate stored in the detail portion 1 of the record is calculated as the CPU usage rate of the resource usage rate reference value in the resource usage rate reference value file 4. Of the CPU usage rate between 0 and the warning value (%), "S", between the warning value and the limit value (%), "W", and the limit value-100. If it is (%), the character "O" is converted into character data for displaying the CPU usage rate, and the job name stored in the detail part 2 and the CPU usage rate of the job are displayed as "*". The converted data and the like are added and edited into the output format to the output device 12 and written in the third work area 11 (step 10).
4).

After the end of step 104, the high load resource determination result output means 10 determines whether or not the high load resource information record in the high load resource information file 9 still has a record whose record ID is "CPU" (step 105), if there is a record, the process returns to step 102.

Next, the high load resource determination result output means 10
One record with the record ID "memory" is read (step 106), and the number of page faults of the system program and the number of page faults of the user program per second stored in the detail portion 1 of the record are calculated.
Compare the resource usage reference value in the resource usage reference value file 4 with the warning value and limit value of the page fault count (step 1
07), "S" if the page fault frequency is between 0 and the warning value (times / second), "W" if the warning value to the limit value (times / second),
If the number is equal to or more than the limit value (times / second), the character "O" is converted into character data for displaying the page fault frequency, and the job name stored in the detail section 2 and the page fault frequency of the job are converted into "character data". The character data displayed by * "is added to the converted character data, edited into the output format to the output device 12, and written in the third work area 11 (step 1
08).

After the step 108, the high load resource determination result output means 10 determines whether or not there is a record having a record ID "memory" in the high load resource information records in the high load resource information file 9 (step 109), if there is a record, the process returns to step 106.

Next, the high load resource determination result output means 10
One record with the record ID "DISK" is read (step 10A), and the number of input / output times of the DISK stored in the detail portion 1 of the record is calculated based on the resource usage rate reference value in the resource usage rate reference value file 4. Value DISK
Compare the warning value and the limit value of the input / output frequency of
B), "S" if the input / output count is between 0 and the warning value (times / second), "W" if the warning value to the limit value (times / second), and "O" if it is greater than or equal to the limit value (times / second). Converted to character data that displays the input / output count with the character "," and the file name stored in the detail section 2 and the character data that displays the input / output count of the file with "*". Is added to edit the output format to the output device 12 and write it in the third work area 11 (step 10C).

After the completion of step 10C, the high load resource determination result output means 10 determines whether or not the high load resource information record in the high load resource information file 9 still has a record whose record ID is "DISK" (step 10D), if there is a record, return to step 10A.

Next, the high load resource determination result output means 10
One record with the record ID "MSP" is read (step 10E), and the number of MSP input / outputs per second stored in the detail section 1 of the record is calculated based on the resource usage rate in the resource usage rate reference value file 4. Compare the warning value and the limit value of the number of input / output times of MSP (step 10F),
The letter “S” if the input / output count is between 0 and the warning value (times / second), “W” if the warning value to the limit value (times / second), and the letter “O” if it is above the limit value (times / second). Convert the input / output count to character data to be displayed, add the DISK name stored in the detail section 2 and the character data to display the input / output count of the DISK to "*" It is edited in the output format to the output device 12 and written in the third work area 11 (step 10G).

After the step 10G, the high load resource determination result output means 10 determines whether or not the high load resource information record in the high load resource information file 9 still has a record with the record ID "MSP" (step 10H), if there is a record, return to step 10E.

After the end of step 10G, the high load resource determination result output means 10 outputs the character data written in the third work area 11 to the output device 12 (step 10I).

FIG. 11 shows an example of CPU information output. The same can be output for other resources.

With the above, the processing of the high load resource evaluation system, which is the operation of the second embodiment of the present invention, is completed.

The high load resource evaluation system according to the second embodiment of the present invention inputs a system operation status record record and compares it with a preset resource utilization rate reference value,
Determine whether each resource that constitutes the computer system is under heavy load, identify the job name or resource name that put the resource under heavy load as a bottleneck, and enter the result. By presenting the result on the output device to the user, even a user who is not familiar with the performance evaluation work can easily and surely perform the performance evaluation work.

[0079]

As described above, the high load resource evaluation system of the present invention inputs the system operation status record and compares it with a preset resource utilization rate reference value,
Determine whether each resource that constitutes the computer system is under heavy load, identify the job name or resource name that put the resource under heavy load as a bottleneck, and enter the result. By presenting the result on the output device to the user, even a user who is not familiar with the performance evaluation work can easily and surely perform the performance evaluation work.

[Brief description of drawings]

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

FIG. 2 is a diagram showing an example of a configuration of a system operation status recording file 1 according to the first embodiment of this invention.

FIG. 3 is a diagram showing an example of system configuration data of a system configuration file 7 in the first embodiment of the present invention.

FIG. 4 is a diagram showing an example of resource usage rate reference values in a resource usage rate reference value file 4 according to the first embodiment of the present invention.

FIG. 5 is a diagram showing a format of a high load resource information record stored in a high load resource information file 9 according to the first embodiment of this invention.

FIG. 6 is a flowchart showing the processing of the system operation status record record input means 2 in the first embodiment of the present invention.

FIG. 7 is a flow chart showing the processing of the high load resource information record generation means 8 in the first exemplary embodiment of the present invention.

FIG. 8 is a flowchart showing the processing of the high load resource determination means 5 in the first exemplary embodiment of the present invention.

FIG. 9 is a block diagram showing a second embodiment of the present invention.

FIG. 10 is a flowchart showing the processing of the high load resource determination result output means 10 in the second exemplary embodiment of the present invention.

FIG. 11 is a diagram showing an example of a high load resource determination result for the output device 12 by the high load resource determination result output means 10 according to the second exemplary embodiment of the present invention.

[Explanation of symbols]

 1 system operating status record file 2 system operating status record record input means 3 first work area 4 resource utilization reference value file 5 high load resource judgment means 6 second work area 7 system configuration file 8 high load resource information record generation Means 9 High-load resource information file 10 High-load resource determination result output means 11 Third work area 12 Output device

Claims (5)

[Claims] 1. The usage status of each resource constituting the computer system is recorded as a system resource usage status record at regular time intervals, and the execution history of each job executed on the computer system is recorded as a job execution status record. A system operation status recording file that is recorded at fixed time intervals, connection relationship information between an external storage device that constitutes the computer system and an external storage device control device that controls the external storage device, and the connection relationship information that is stored in the external storage device. A system configuration file that stores file name information as system configuration data, a system operation status record input unit that inputs each record in the system operation status recording file, and a usage rate of each resource that constitutes the computer system Storing the warning value and the limit value for the performance guarantee of the server as the resource usage standard value A resource usage rate reference value file, and resource usage rate information of the system resource usage status record input by the system operation status record record input means as a limit value of the resource usage rate reference value in the resource usage rate reference value file. High load resource selecting means for comparing and comparing the resource name and recording time of the system resource usage record indicating a high load usage status exceeding this limit value, and the resource usage rate, and the system operation status record input means After the high load resource selecting means is started, the recording time of the system resource use status record extracted by the high load resource selecting means and the job input by the system operation status record record input means recorded at the same time Job name in execution status record, resource usage rate information for each resource, system operation status record Other resource name and resource usage rate information in the system resource usage record input by the force means are extracted, and the information of each resource and the resource are used by referring to the system configuration data of the system configuration file. Job information, or information on each resource and information on the resources that compose the resource are integrated in a linked form and 1
A high-load resource information record generation unit that is one high-load resource information record, and a high-load resource information file that stores the high-load resource information record generated by the high-load resource information record generation unit. Load resource evaluation system. 2. The resource usage rate reference value in the resource usage rate reference value file is a performance with respect to a preset CPU usage rate, memory page fault frequency, external storage device, and external storage device controller input / output frequency. The high load resource evaluation system according to claim 1, wherein the high load resource evaluation system comprises a warning value and a limit value for guarantee. 3. The high load resource information record in the high load resource information file is composed of an outline part and a detail part, and the outline part is CPU information, memory information, external storage device information, or external storage. The record ID indicating whether the device control device information, the sampling time, the position of the detail part, and the size of the detail part are shown.
If U, CPU usage rate and CPU usage rate for each job; if record ID is memory, page fault frequency and page fault frequency for each job; record ID
Is an external storage device, the number of input / output times of this external storage device and the number of input / output times for each file in the external storage device,
And when the record ID is external storage device control device information, it is composed of the number of input / output times of the external storage device control device and the number of input / output times of each external storage device connected to the external storage device control device. The high-load resource evaluation system according to claim 1. 4. The high load resource determination result output means for inputting a high load resource information record stored in the high load resource information file and displaying the result in a graph on an output device. Load resource evaluation system. 5. The graph displayed by the high load resource judgment result output means displays the usage rate of each resource at each recording time in different characters with the warning value and the limit value of the resource usage rate reference value as a boundary. In addition to the above, the resource usage rate for each job that uses each resource at the recording time, or the input / output count for each file and each external storage device is displayed side by side. The high-load resource evaluation system described.