JP2008305065A - Cpu resource management method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a CPU resource management method that is better than before. <P>SOLUTION: A processor 1 reads X pieces of CPU available time allocated to X process groups from a storage part 210 as X pieces of read CPU available time, and reads management time, i.e., a total value of X pieces of CPU available time, from the storage part 212 as read management time, and executes a first process for subtracting actual time from the read management time. If an (i)-th piece of read CPU usable time ä1≤ i (X-1)} assumes a value greater than zero, one process among an (i)-th process group is executed, and a second process for subtracting its execution time from the (i)-th piece of read CPU usable time is executed. If the (i)-th piece of read CPU available time shows zero, the second process is executed on the (i+1)-th process group. If the read management time shows zero, the first and second processes are executed for processes yet to be executed among the X process groups. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a CPU resource management method for controlling CPU resources (CPU usage time).

  In an information processing apparatus, a method for controlling CPU resources has been developed. For example, a conventional CPU resource management system is described in Japanese Patent No. 2508687 (Patent Document 1). In this CPU resource management system, CPU usage time is allocated (distributed) for each task. Next, when an interrupt (exception) is generated from a task that has consumed CPU resources, the CPU usage time of the related task is integrated and the remaining CPU usage time is redistributed. That is, if it is not necessary to end the task, the remaining CPU usage time is assigned (redistributed) as the CPU usage time of the task.

  However, this conventional CPU resource management system has the following problems.

  First, in the conventional CPU resource management system, CPU resources are controlled by an interrupt from a task that first consumed CPU usage time. For this reason, as a first problem, CPU resources cannot be controlled unless there is a task that consumes CPU usage time.

  Second, in the conventional CPU resource management method, after the remaining CPU usage time is redistributed, the operation is not always performed according to the reset value. In other words, even when there is a task that consumes CPU usage time, it is necessary to add up the CPU usage time of multiple tasks and determine the necessity of redistribution. Overuse and shortage will occur. For this reason, as a second problem, the task turnaround time increases.

  Third, in the conventional CPU resource management method, an increase in cache access or exclusive control is required to accumulate the CPU usage time of other tasks when the CPU usage time of the task is consumed. For this reason, as a third problem, the system performance of the multiprocessor configuration deteriorates due to data competition on the cache block by control and overhead by exclusive control.

  Thus, there is a need for a CPU resource management method and management method that can solve the above-described problems and that is superior to conventional ones, and a computer program that implements the same.

  Introduces technologies related to CPU resources.

  Japanese Patent No. 2508687 (Patent Document 1) describes a “CPU usage time management system for jobs”. This CPU usage time management method includes a means for processing a plurality of tasks constituting a job, a timer counter provided for each task, the value of which decreases according to the time that the task uses the CPU of the computer, And a means having a function of generating an exception when the value of the timer counter reaches 0, the CPU usable time designated by the user for a specific job at an arbitrary ratio. A means for allocating to tasks and setting each timer counter, a means for calculating the remaining CPU usable time by integrating the CPU time used by each task up to that point, and the remaining CPU usable time Means for determining whether or not to redistribute the remaining CPU usable time, and when the determining means determines to redistribute, Means for reallocating the available time to each task and resetting it to each timer counter; and when the decision means decides not to redistribute, all the remaining CPU available time is not redistributed. Means for resetting the timer counter, and means for terminating the job when one of the reset counter values becomes 0, and the exception generating means calculates the remaining time when an exception occurs. It is composed of a plurality of tasks characterized by operating means.

  Japanese Patent Laid-Open No. 2003-296289 (Patent Document 2) describes a “load distribution method”. This load balancing method includes an average CPU usage time management table for storing an average CPU usage time required for executing a transaction processing message for each transaction processing message identifier for identifying the transaction processing message, and a plurality of transaction processings. The average CPU usage time required to execute each of the plurality of transaction processing messages is received by searching the average CPU usage time management table for each transaction using the transaction processing message identifier of each of the plurality of transaction processing messages received as a key. And calculating a predicted value of CPU usage time required to execute the plurality of transaction processing messages from the acquired average CPU usage time. And determining the distribution distribution of the plurality of transaction processing messages to the plurality of servers based on the predicted value and the current CPU usage rate of each of the plurality of servers acquired at regular time intervals. And a transaction distribution unit that inputs the plurality of transaction processing messages to each of the plurality of servers.

  Japanese Laid-Open Patent Publication No. 62-11941 (Patent Document 3) describes a “multiprocessing control method”. This multi-processing control method is a method in which a plurality of processing units are assigned to a central processing unit by a time slice method in an information processing system, and each processing unit provided corresponding to each processing unit is time sliced. A storage area for storing time slice time intervals for execution in the storage area, and a storage area for each processing unit that is provided for each processing unit and that accumulates the time for which each processing unit uses the central processing unit and stores this integration time And an allocation ratio table that divides the processing unit into a plurality of groups and presets and stores the usage allocation ratio of the central processing unit for each group, and the usage time ratio of the central processing unit for each group every predetermined time And a corresponding group in the ratio table of the use time ratio for each group. And changing means for changing the time slice time interval of each processing unit in the group according to the error, and integrating each processing unit with the use time of the central processing unit. However, it is characterized in that it is executed at the time slice time interval.

  Japanese Patent No. 3150114 (Patent Document 4) describes a “dispatching device”. This dispatch apparatus is an apparatus for controlling the queuing control by managing the allocation state of a plurality of CPUs and a plurality of queues in a multiprocessor computer. Execution time that accumulates, calculates a ratio of the total value of execution time for each queue at a predetermined timing, and initializes the total value of job execution time for each queue every time the CPU is assigned An allocation control unit that allocates the CPU to the queue according to a ratio of a total value of job execution times for each queue calculated by the execution time management unit at a predetermined timing; Each time the predetermined set time elapses, the ratio of the total value of job execution times for each queue by the execution time management means And output processing, and is characterized by comprising a timing generation means for notifying the timing of assignment processing of the CPU for the queue by the assignment control unit to the execution time management means and said allocation control means.

  International Publication No. WO2002 / 069174 (Patent Document 5) describes “Parallel Process Execution Method”. This parallel process execution method is a method for processing a plurality of parallel processes generated from one parallel program and other processes by a plurality of processors in a time-sharing manner. Setting a time allocation rate to be allocated to processing, allocating each parallel process of the parallel program to one of the plurality of processors, and simultaneously starting processing of the allocated parallel processes in the plurality of processors; When the elapsed time from the start of execution of the parallel process reaches a time corresponding to the time allocation rate set in the parallel program within the predetermined period, the time is allocated in the plurality of processors. It is characterized by ending execution of parallel processes at the same time.

Japanese Patent No. 2508687 JP 2003-296289 A JP 62-111941 A Japanese Patent No. 3150114 International Publication No. WO2002 / 069174 Pamphlet

An object of the present invention is to provide a CPU resource management method capable of reliably controlling CPU resources.
Another object of the present invention is to provide a CPU resource management system capable of shortening the turnaround time as compared with the prior art.
Still another object of the present invention is to provide a CPU resource management system that can eliminate overhead.

The CPU resource management system of the present invention includes a processor that operates on a CPU (Central Processing Unit) and a storage device.
The storage device includes a process information storage unit storing X process groups from the first to the Xth (X is an integer equal to or greater than 2), and the use of X CPUs allocated to each of the X process groups. The CPU resource initial information storage unit storing the available time and the management time storage unit storing the management time that is the total value of the X CPU usable times are provided.
The processor reads the X CPU usable time from the CPU resource initial information storage unit as X read CPU usable time, reads the management time from the management time storage unit as the read management time, and calculates the real time from the read management time. A first process of subtraction is performed.
The processor executes one process of the i-th process group when the i-th read CPU available time {i is an integer satisfying 1 ≦ i (X−1)} is greater than 0, and A second process of subtracting the execution time from the i-th read CPU usable time is performed.
When the i-th read CPU usable time represents 0, the processor performs the second process on the (i + 1) -th process group.
When the read management time represents 0, the processor performs the first process and the second process on an unexecuted process that has not yet been executed among the X process groups.

  In the CPU resource management system of the present invention, as a first effect, CPU resources of X process groups (X CPU usable time) are controlled in units of processors according to the management time. That is, the CPU resource is controlled without depending on the task that consumed the CPU usage time. Thereby, in the CPU resource management system of the present invention, the CPU resource can be reliably controlled.

  In the CPU resource management system of the present invention, the CPU resource can be controlled without going through software control as the second effect by the first effect. Thereby, in the CPU resource management system of the present invention, the process turnaround time can be greatly shortened.

  Further, in the CPU resource management system of the present invention, as a third effect, the CPU resource is controlled in units of processors without software control due to the first and second effects. As a result, the CPU resource management system of the present invention is capable of completely eliminating the data competition on the cache block and the overhead due to the exclusive control, which cause the system performance to deteriorate in the multiprocessor configuration. For this reason, the system can be operated stably.

  The CPU resource management system of the present invention will be described below in detail with reference to the accompanying drawings.

[Constitution]
FIG. 1 is a block diagram showing the configuration of the CPU resource management system of the present invention. The CPU resource management system of the present invention includes N (an integer greater than or equal to 1) processors 1 and N storage devices 2. Each of the N processors 1 is assigned to N storage devices 2, and in the present invention, CPU resource management is controlled in units of processors. For example, when N is 4, the processor numbers for identifying the four processors 1 are represented as processors “00”, “01”, “02”, and “03”, respectively.

  The processor 1 includes a control unit 10 that operates on a CPU (Central Processing Unit) and a management unit 11 having a timer function.

  The storage device 2 includes a process information storage unit 20 and a processor control information storage unit 21. The processor control information storage unit 21 includes a CPU resource initial information storage unit 210, a CPU resource information storage unit 211, and a management time storage unit 212.

  The process information storage unit 20 includes X process groups waiting for execution from the first to the Xth (X is an integer of 2 or more), X process group identification information for identifying the X process groups, and Is stored.

  The CPU resource initial information storage unit 210 stores X process group identification information. The CPU resource initial information storage unit 210 further stores X CPU usable times assigned to the X process group identification information as CPU resource initial information.

  The management time storage unit 212 stores a management time that is a total value of X CPU usable times.

  The processor 1 performs the first process and the second process on the X process groups waiting to be executed.

  First, the processor 1 performs a first process.

  The control unit 10 of the processor 1 reads the CPU resource initial information (X CPU usable time) from the CPU resource initial information storage unit 210 as CPU resource information (X read CPU usable time), and stores the CPU resource information. Stored in the unit 211. Further, the control unit 10 reads the management time from the management time storage unit 212 as the read management time, and sets the read management time in the management unit 11 of the processor 1. The management unit 11 counts the real time and subtracts the real time from the read management time.

  Next, the processor 1 performs a second process.

  When the i-th read CPU usable time {i is an integer satisfying 1 ≦ i (X−1)} represents a value larger than 0, the control unit 10 of the processor 1 selects one process in the i-th process group. And the execution time is subtracted from the i-th read CPU usable time. Here, when the i-th read CPU usable time represents 0, the control unit 10 performs the second process on the (i + 1) -th process group.

  When the read management time represents 0, the processor 1 performs the first process and the second process described above for an unexecuted process that has not yet been executed in the X process group.

[Operation]
Next, the operation of the CPU resource management system of the present invention will be specifically described with reference to FIGS. 2 and 3A to 3E. The storage device 2 stores (installs) a computer program (not shown) for the processor 1 (computer) to execute the operation (flow) shown in FIG. In the present invention, since CPU resource management is controlled in units of processors, the operation of the processor 1 whose processor number is “00” will be described.

  The process information storage unit 20 includes three process groups waiting to be executed from the first to the third (X is 3) and three process group identification information for identifying the three process groups. Assume that it is stored. Here, as shown in FIG. 3A, the three process groups are referred to as process groups “A”, “B”, and “C”, respectively. The processes waiting to be executed are the processes “1”, “2”, “3”, “4”, “5”, and the processes “1”, “2” are included in the process group “A”, and the process “3” ”And“ 4 ”are included in the process group“ B ”, and the process“ 5 ”is included in the process group“ C ”. Further, the time required from the start of execution of the processes “1”, “2”, “3”, “4”, “5” to the end thereof is “40 ms”, “40 ms”, “20 ms”, respectively. ”,“ 20 ms ”, and“ 50 ms ”.

  The CPU resource initial information storage unit 210 stores three process group identification information and three CPU usable times assigned to each of the three process group identification information. Here, as shown in FIG. 3A, the three CPU usable times are defined as CPU usable times “40 ms”, “30 ms”, and “30 ms”, respectively.

  As shown in FIG. 3A, it is assumed that the management time storage unit 212 stores a management time “100 ms” that is a total value of the CPU usable time “40 ms”, “30 ms”, and “30 ms”.

  The processor 1 “00” performs the first process and the second process on the process group “A”, “B”, “C” waiting for execution.

  First, the processor 1 “00” performs the first process.

  As shown in FIG. 3A, the control unit 10 of the processor 1 “00” sets the CPU usable time “40 ms”, “30 ms”, “30 ms” from the CPU resource initial information storage unit 210 to the first, second, second 3 Reads as the CPU available time, and stores it in the CPU resource information storage unit 211. Further, the control unit 10 reads the management time “100 ms” from the management time storage unit 212 as the read management time, and sets the read management time “100 ms” in the management unit 11 of the processor 1 “00”. The management unit 11 counts the real time and subtracts the real time from the read management time “100 ms”.

  Next, the processor 1 “00” performs the second process on the process group “A” (step S1-YES).

  As shown in FIG. 3A, the usable time of the first read CPU is “40 ms”. That is, it represents a value greater than 0 (step S2-YES). At this time, the control unit 10 of the processor 1 “00” executes the process “1” of the process group “A” for “40 ms”, and uses the execution time “40 ms” as the first read CPU usable time “40 ms”. Is subtracted from "" (step S3).

  As shown in FIG. 3B, after the process “1” of the process group “A” is executed, the first read CPU usable time is “0 ms”. That is, it represents 0. The read management time represents “60 ms”. At this time, since the control unit 10 of the processor 1 “00” cannot execute the process “2” of the process group “A” for “40 ms”, the execution of the process “2” is suspended (step S1-YES, S2-NO).

  Next, the processor 1 “00” executes the second process on the process group “B” (step S4-YES).

  As shown in FIG. 3B, the second read CPU usable time is “30 ms”. That is, it represents a value greater than 0 (step S2-YES). At this time, the control unit 10 of the processor 1 “00” executes the process “3” of the process group “B” for “20 ms”, and uses the execution time “20 ms” as the second read CPU usable time “30 ms”. Is subtracted from "" (step S3).

  As shown in FIG. 3C, in this case, the second read CPU usable time is “10 ms”. That is, it represents a value larger than 0. At this time, the control unit 10 of the processor 1 “00” executes the process “4” of the process group “B” for “10 ms” of “20 ms”, and secondly reads the execution time “10 ms”. The CPU usable time is subtracted from “10 ms” (steps S1-YES, S2-YES, S3).

  As shown in FIG. 3D, during the execution of the process “4” of the process group “B”, the second read CPU usable time is “0 ms”. That is, it represents 0. The read management time represents “30 ms”. At this time, the control unit 10 of the processor 1 “00” cannot execute the process “4” of the process group “B” for the remaining “10 ms”, and therefore suspends the execution of the process “4” (steps S1-YES, S2). -NO).

  Next, the processor 1 “00” executes the second process on the process group “C” (step S4-YES).

  As shown in FIG. 3D, the third read CPU usable time is “30 ms”. That is, it represents a value greater than 0 (step S2-YES). At this time, the control unit 10 of the processor 1 “00” executes the process “5” in the process group “C” for “30 ms” in “50 ms”, and reads the execution time “30 ms” in the third reading. The CPU usable time is subtracted from “30 ms” (step S3).

  As shown in FIG. 3E, the third read CPU usable time is “0 ms” during the execution of the process “5” of the process group “C”. That is, it represents 0. The read management time represents “0 ms”. At this time, since the control unit 10 of the processor 1 “00” cannot execute the process “5” of the process group “C” for the remaining “20 ms”, the execution of the process “5” is suspended (steps S1-YES, S2). -NO, S4-NO).

  As shown in FIG. 3E, the read management time represents “0 ms”. For this reason, the processor 1 “00” performs the above-described process “2”, “4”, “5” among the process groups “A”, “B”, “C” with respect to the processes “2”, “4”, “5”. The first process and the second process are performed (steps S1 to S4).

  In this way, in the CPU resource management system of the present invention, by controlling the CPU resource of the process group to which the process belongs according to the management time for each processor, data competition and exclusive control associated with software operation time and software control are performed. Thus, CPU resource management can be performed without being affected by a factor of system performance degradation due to the above, and the system can be operated stably.

[effect]
The effect of the CPU resource management system of the present invention will be described.

  First, in the conventional CPU resource management method, as shown in FIG. 4A, as a first problem, CPU usage time is allocated (distributed) for each task, and an interrupt (exception) from a task that consumes CPU resources. When this occurs, the CPU usage time of the related task is integrated and the remaining CPU usage time is redistributed. That is, if it is not necessary to end the task, the remaining CPU usage time is assigned (redistributed) as the CPU usage time of the task. For this reason, in the conventional CPU resource management system, CPU resources cannot be controlled unless there is a task that consumes CPU usage time.

  Further, in the conventional CPU resource management method, as a second problem, even when there is a task that consumes the CPU usage time, the CPU usage time of a plurality of tasks is accumulated and the necessity of redistribution is determined. Processing is required, and task turnaround time increases.

  Further, in the conventional CPU resource management method, as a third problem, in a multiprocessor configuration, system performance degradation factors such as software operation time, data competition due to software control, and exclusive control increase. The excess of use time cannot be suppressed completely.

  On the other hand, in the CPU resource management system of the present invention, as shown in FIG. 4B, the first effect is that the CPU resources (CPUs) of the process groups “A”, “B”, and “C” correspond to the management time “100 ms”. (Usable time) “40 ms”, “30 ms”, “30 ms” are controlled in units of processor “00”, that is, the CPU resources are controlled without depending on the task that consumed the CPU usage time. Thus, in the CPU resource management system of the present invention, the CPU resource can be reliably controlled.

  Further, in the CPU resource management system of the present invention, as a second effect, the CPU resources “40 ms”, “30 ms”, and “30 ms” can be controlled without going through software control. Thereby, in the CPU resource management system of the present invention, the process turnaround time can be greatly shortened.

  In the CPU resource management system according to the present invention, the first and second effects provide a third effect. As a third effect, the CPU resources “40 ms”, “30 ms”, “ 30 ms "control is performed. As a result, the CPU resource management system of the present invention is capable of completely eliminating the data competition on the cache block and the overhead due to the exclusive control, which cause the system performance to deteriorate in the multiprocessor configuration. For this reason, the system can be operated stably.

[application]
In the present invention, the difference (change) between the CPU resource initial information (X CPU usable time) and the CPU resource information (X read CPU usable time) is stored in the storage unit 20 described above of the storage device 2. By recording in a storage area other than 21, it is possible to collect the operating state independently of the CPU resource management.

  In the present invention, when switching to a spare processor in the event of a processor failure or when adding a processor, the CPU resource information is set only for the target processor or the system is in operation by preparing in advance. It is possible to easily cope with a change in the configuration.

  In the present invention, the management time can be defined flexibly by defining the management time as a single value in the process group unit or system, or by defining a set of arbitrary processors as a processor group and defining in a processor group unit. It becomes possible to do.

  In the present invention, it is possible to perform dynamic resource management by making it possible to set CPU resource initial information, CPU resource information, and management time setting values by software.

  INDUSTRIAL APPLICABILITY According to the present invention, it can be applied to uses such as stable operation of system work in cases such as outsourcing and integration of multiple systems.

FIG. 1 is a block diagram showing the configuration of the CPU resource management system of the present invention. FIG. 2 is a flowchart showing the operation of the CPU resource management system of the present invention. FIG. 3A is a diagram for explaining the operation of the CPU resource management system of the present invention. FIG. 3B is a diagram for explaining the operation of the CPU resource management system of the present invention. FIG. 3C is a diagram for explaining the operation of the CPU resource management system of the present invention. FIG. 3D is a diagram for explaining the operation of the CPU resource management system of the present invention. FIG. 3E is a diagram for explaining the operation of the CPU resource management system of the present invention. FIG. 4A is a diagram for explaining a problem of the conventional CPU resource management method. FIG. 4B is a diagram for explaining the effect of the CPU resource management system of the present invention.

Explanation of symbols

1 processor,
2 storage devices,
10 control unit,
11 Management Department,
20 process information storage unit,
21 processor control information storage unit,
210 CPU resource initial information storage unit,
211 CPU resource information storage unit,
212 management time storage unit,

Claims (7)

A processor that operates on a CPU (Central Processing Unit);
A storage device,
The storage device
A process information storage unit storing X process groups from the first to the Xth (X is an integer of 2 or more);
CPU resource initial information storage unit storing X CPU usable time allocated to each of the X process groups;
A management time storage unit storing a management time that is a total value of the X CPU usable times,
The processor is
The X CPU usable time is read from the CPU resource initial information storage unit as X read CPU usable time, the management time is read from the management time storage unit as a read management time, and the actual time is read from the read management time. Perform the first process to subtract time,
When the i-th read CPU usable time {i is an integer satisfying 1 ≦ i (X−1)} represents a value larger than 0, one process of the i-th process group is executed, and the execution time is set as the execution time. Performing a second process of subtracting from the i-th read CPU usable time,
When the i-th read CPU usable time represents 0, the second process is performed on the (i + 1) -th process group,
A CPU resource management system that performs the first process and the second process on an unexecuted process that has not yet been executed among the X process groups when the read management time represents 0. The storage device
A CPU resource information storage unit;
The processor reads the X CPU usable time from the CPU resource initial information storage unit as X read CPU usable time in the first process, and stores it in the CPU resource information storage unit. CPU resource management method described in 1. The processor is
A management unit,
In the first process, the processor reads the management time from the management time storage unit as a read management time, sets the read management time in the management unit,
The CPU resource management system according to claim 1, wherein the management unit counts the real time and subtracts the real time from the read management time. The processor is
In the second process, when the i-th read CPU usable time represents a value larger than 0, the first process of the i-th process group is executed, and the execution time can be used for the i-th read CPU. Subtract from time,
After the execution of the first process of the i-th process group, when the i-th read CPU usable time represents 0, the execution of the second process of the i-th process group is suspended, and the (i + 1) -th process is suspended. Performing the second process on the process group;
The said 1st process and said 2nd process are performed with respect to the said unexecuted process including the said 2nd process of the said i-th process group when the said read management time represents 0. The described CPU resource management method. The processor is
In the second process, when the i-th read CPU usable time represents a value larger than 0, the first process of the i-th process group is executed, and the execution time can be used for the i-th read CPU. Subtract from time,
In the second process, when the i-th read CPU usable time represents a value greater than 0 after the execution of the first process of the i-th process group, the second process of the i-th process group is And subtracting the execution time from the i th read CPU usable time,
During the execution of the second process of the i-th process group, when the i-th read CPU usable time represents 0, the execution of the second process of the i-th process group is suspended, and the (i + 1) th Performing the second process on the process group;
The said 1st process and said 2nd process are performed with respect to the said unexecuted process including the said 2nd process of the said i-th process group when the said read management time represents 0. The described CPU resource management method. A method for managing CPU resources using a processor that operates on a CPU (Central Processing Unit),
When executing X process groups from the 1st to the Xth (X is an integer of 2 or more),
X CPU usable time allocated to each of the X process groups is read as X read CPU usable time from the CPU resource initial information storage unit, and is a total value of the X CPU usable times. Performing a first process of reading the management time from the management time storage unit as a read management time, and subtracting the real time from the read management time;
When the i-th read CPU usable time {i is an integer satisfying 1 ≦ i (X−1)} represents a value larger than 0, one process of the i-th process group is executed, and the execution time is set as the execution time. Performing a second process of subtracting from the i-th read CPU usable time;
Performing the second process on the (i + 1) -th process group when the i-th read CPU usable time represents 0;
A CPU resource comprising: a step of performing the first process and the second process on an unexecuted process that has not yet been executed in the X process group when the read management time represents 0 Management method. A computer program installed on a computer that operates on a CPU (Central Processing Unit),
When executing X process groups from the 1st to the Xth (X is an integer of 2 or more),
X CPU usable time allocated to each of the X process groups is read as X read CPU usable time from the CPU resource initial information storage unit, and is a total value of the X CPU usable times. Performing a first process of reading the management time from the management time storage unit as a read management time, and subtracting the real time from the read management time;
When the i-th read CPU usable time {i is an integer satisfying 1 ≦ i (X−1)} represents a value larger than 0, one process of the i-th process group is executed, and the execution time is set as the execution time. Performing a second process of subtracting from the i-th read CPU usable time;
Performing the second process on the (i + 1) -th process group when the i-th read CPU usable time represents 0;
When the read management time represents 0, the computer executes the step of performing the first process and the second process on an unexecuted process that has not yet been executed in the X process group. Computer program to let you.

Images