JP2536304B2 - Job end time prediction system - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a job end time prediction system, and more particularly to a predicting method for predicting a job end time in a data processing apparatus having a processing unit for performing processing on a job basis.

[0002]

2. Description of the Related Art Conventionally, in a data processing apparatus, a job end time is predicted by a user's empirical prediction by obtaining information such as job processing time and CPU (central processing unit) usage time. Alternatively, the user makes a statistical prediction by collecting information such as job processing time and CPU usage time.

In such a conventional data processing apparatus, when the job end time is predicted by the user's empirical prediction, there is a drawback that the prediction result is unreliable.

When the job end time is predicted statistically, information such as job processing time and CPU usage time must be added up in advance. There is a drawback in that it is troublesome and the work of predicting the end time of the job from the processing element of the job is troublesome.

[0005]

SUMMARY OF THE INVENTION The present invention has been made to eliminate the above-mentioned drawbacks of the prior art, and improves the reliability of job end time prediction and facilitates job end time prediction. The purpose is to provide a predictable job end time system.

[0006]

A job end time predicting system according to the present invention is a job end time predicting system for a data processing apparatus having a multi-job type processing means, and the relationship between the usage rate of the processing means and the job processing time. A processing time measuring means for measuring and collecting the usage rate, a usage rate measuring means for measuring and collecting the usage rate of the processing means within a preset predetermined period at fixed time intervals, and a measurement of the usage rate measuring means A calculation unit that calculates the usage rate of the processing unit for each time period from the result, and a usage rate of the processing unit when the job is started by the measurement result of the processing time measurement unit and the calculation result of the calculation unit And a prediction unit for predicting the end time of the job.

[0007]

Next, an embodiment of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of one embodiment of the present invention. In the figure, when a job is input from the input unit 1, the processing time measuring unit 2 measures the CPU load (usage rate) and the processing time of each processing element of the job in the data processing device 7, and the processing element of each job is measured. And stores the statistics of the relationship between the CPU load and the processing time.

The CPU load measuring unit 3 constantly measures the CPU load of the data processing device 7 at preset preset time intervals, and sends the measurement result and the acquisition time thereof to the CPU load condition calculating unit 4.

The CPU load condition calculation unit 4 uses the measurement results obtained by the CPU load measurement unit 3 to determine the CP for each fixed period (for example, daily, weekly, monthly, or yearly).
U statistics of the load situation are taken and stored.

The predicting unit 5 stores the statistical data stored in the processing time measuring unit 2 before the job is submitted from the input unit 1 and the statistical data stored in the CPU load status calculating unit 4 and data. The job end time is predicted based on the CPU load state at the time of job input in the processing device 7 and the job input time from the input unit 1, and the prediction result is displayed on the CRT display 6.

FIG. 2 is a flow chart showing the processing operation of the processing time measuring unit 2 of FIG. 1, and FIG. 3 is a flow chart showing the processing operation of the CPU load measuring unit 3 of FIG.
Is a flowchart showing the processing operation of the CPU load condition calculation unit 4 in FIG. 1, and FIG. 5 is a flowchart showing the processing operation in the prediction unit 5 in FIG. The operation of the embodiment of the present invention will be described with reference to FIGS.

While the data processing device 7 is operating, the CP
The U load measuring unit 3 always operates and measures the CPU load of the data processing device 7 at regular time intervals (step 3 in FIG. 3).
1) The measurement result and the time at the time of acquisition are stored (step 32 in FIG. 3).

The CPU load condition calculation unit 4 inputs the information collected by the CPU load measurement unit 3, that is, the CPU load measured by the CPU load measurement unit 3 and the time at which the CPU load is acquired in a fixed period unit (see FIG. 4). Step 41), the CPU load status statistics for each fixed period are taken from the information (FIG. 4).
In step 42), the totalization result of the statistics is stored in a file (not shown) (hereinafter referred to as a statistical file) or the like (step 43 in FIG. 4).

That is, the CPU load measuring unit 4 calculates the average value of the CPU load with respect to the time obtained by dividing the fixed time stored in the statistical file before performing the statistical processing into N and dividing the fixed time into the CPU load measuring unit 3 The average of the CPU loads for each N-divided time units is calculated by adding the average of the CPU loads for each N-divided time units obtained from the CPU load measured in Step 1 and the time at the time of acquisition, and the average of the CPU loads Write the values to the statistics file and update the statistics file.
The value of N varies depending on the unit of fixed time.

That is, if the fixed time is on a daily basis, how much CPU load is in what time zone of the day is collected every day, and the statistics are stored in the statistics file.

If the fixed time is weekly, the CPU load statistics for each day of the week are taken, if the fixed time is monthly, the CPU load statistics are taken for each day, and if the fixed time is yearly, monthly. The CPU load statistics are taken and stored in the separately created statistical files.

When a job is input from the input unit 1 in the state where the totalized results are stored in the statistical files of the CPU load condition calculation unit 4 by the above processing, the processing time measurement unit 2 causes the processing element of the job to process. 2 is input (step 21 in FIG. 2), and the CPU load and the processing time for each processing element of the job are measured from the information (step 22 in FIG. 2).

Further, the processing time measuring unit 2 obtains statistics of the relationship between the CPU load and the processing time for each processing element of the job from the measurement results, and stores the statistics aggregation result (step 23 in FIG. 2).

At the same time, the predicting unit 5 statistics the relationship between the CPU load and the processing time of each processing element of the job stored in the processing time measuring unit 2 before the job is input from the input unit 1. And the CPU load status calculation unit 4
CP stored in each statistical file
The total result of the average value of the U load is input as the prediction information (step 51 in FIG. 5), and the CPU load state at the time of job input by the input unit 1 is input from the data processing device 7 (step 52 in FIG. 5).

After that, the predicting unit 5 collects the results of the processing time measuring unit 2, the totaling results of the statistical files of the CPU load status calculating unit 4, and the CPU load state at the time of job input from the data processor 7. The job end time is predicted based on the job input time (step 53 in FIG. 5).

That is, the CPU load is calculated statistically by searching the data corresponding to the job start time from the totalized result of each statistical file of the CPU load status calculation unit 4, and the CP
The processing time corresponding to the U load is searched for from the totalized result from the processing time measuring unit 2, and the job end time is determined by the calculated processing time and the job input time.

By the above-described processing, the end time of the job predicted by the predicting unit 5 is notified to the CRT display 6 (step 54 in FIG. 5) and displayed on the CRT display 6.

As described above, the statistical result of the relationship between the CPU load and the processing time of each processing element of the job stored in the processing time measuring section 2 before the job is input from the input section 1 and the CPU The end time of the job is predicted based on the result of totaling the average value of the CPU load for each fixed period stored in each statistical file of the load status calculation unit 4, the CPU load state at the time of job submission, and the job submission time. By doing so, when the job end time is predicted by the user's empirical prediction, the prediction result is not reliable, and when the job end time is predicted by statistical prediction. Eliminating the trouble of aggregation work and predicting the job end time Since the trouble of the work can be eliminated, the reliability of the job end time prediction can be improved, and It is possible to perform the end time predicted easily.

[0025]

As described above, according to the present invention, C
The measurement result of the relationship between the PU usage rate and the job processing time,
By predicting the end time of the job from the CPU usage rate for each time zone and the CPU usage rate when the job is started, the reliability of the prediction of the job end time is improved, and the job end time is improved. There is an effect that the end time can be easily predicted.

[Brief description of drawings]

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

FIG. 2 is a flowchart showing a processing operation of a processing time measuring unit in FIG.

FIG. 3 is a flowchart showing a processing operation of a CPU load measuring unit in FIG.

FIG. 4 is a flowchart showing a processing operation of a CPU load condition calculation unit in FIG.

5 is a flowchart showing a processing operation of a prediction unit in FIG.

[Explanation of symbols]

 2 Processing time measurement unit 3 CPU load measurement unit 4 CPU load status calculation unit 5 Prediction unit 7 Data processing device

Claims (1)

(57) [Claims] 1. A job end time prediction system for a data processing apparatus having multi-job processing means, comprising:
A processing time measuring means for measuring and collecting the relationship between the usage rate of the processing means and the processing time of the job, and measuring and collecting the usage rate of the processing means within a predetermined period set in advance at a constant time interval. Utilization rate measuring means, calculating means for calculating the utilization rate of the processing means for each time zone from the measurement result of the utilization rate measuring means, the measurement result of the processing time measuring means and the calculation result of the calculating means A job end time prediction system, comprising: a prediction means for predicting an end time of the job from a usage rate of the processing means when the job is started.