JP2004246592A - Job execution management device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To efficiently allocate jobs such as the processes of executing a license software to an entire system. <P>SOLUTION: This job execution management device includes a CPU resource managing part 8 which manages the state of CPU resources over a network; a job management database 5 which retains the previous job execution information including average job execution time; a job analyzing part 4 which analyzes the contents of the jobs by referring to the job execution information retained in the job management database 5; and a queue managing part 7 which manages the jobs as a queue when there is no vacancy in the CPU resources capable of executing the jobs. Further, the device includes a job managing part 3 which allocates the best-suited CPU resource capable of executing a job in the shortest period of time according to the result of analysis on the job including the average job execution time corresponding to the job and the state of the CPU resources and which causes the job management database 5 to retain information on the execution of the job. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a job execution management apparatus that is arranged on a network and manages a job requested by a client by allocating the job to an appropriate CPU resource on the network and executing the job. The present invention relates to a job execution management device that efficiently executes the job execution.
[0002]
[Prior art]
Conventionally, there is a system that can use a limited number of licensed software on a network. In this system, if a client requests to execute a job exceeding the number of contracted licenses, Are stored in a queue, and when a license becomes available, this job is allocated to the available CPU resource and executed.
[0003]
[Patent Document 1]
JP 2002-157037 A [Patent Document 2]
JP 2001-222424 A [Patent Document 3]
JP-A-9-62393 [Patent Document 4]
JP-A-5-346851
[Problems to be solved by the invention]
However, in the conventional system, the queue is managed in the order of job requests, and when a license becomes available, the job is simply allocated in this queue order to the CPU that executes the licensed software that has become available. Since the job is executed, there is a problem that the job assignment may not be appropriate, and the entire system may not be efficient. For example, a low-priority job is allocated to a CPU with a high-speed processing capability, and execution of this job is completed in a short time, whereas a high-priority job is allocated to a CPU with a low-speed processing capability, and There is a problem in that the execution of the job is completed, and as a result, the processing of a high-priority job is slower than that of a low-priority job.
[0005]
In addition, there is a problem in that by allocating the inefficient license software, an unnecessary license software is contracted to form a useless network.
[0006]
The present invention has been made in view of the above, and it is possible to efficiently allocate a job such as a process of executing a license software in the entire system and execute a job execution in which a client can easily grasp a job execution status. The purpose is to obtain a management device.
[0007]
[Means for Solving the Problems]
In order to achieve the above object, a job execution management apparatus according to the present invention is arranged on a network, and performs management by allocating a job requested by a client to an appropriate CPU resource on the network and executing the job. A resource management unit that manages a state of CPU resources on the network, a job management database that holds past job execution information including an average job execution time, and job execution information that is held in the job management database. Job analysis means for analyzing the contents of the job, queue management means for managing the job as a queue when there is no free CPU resource capable of executing the job, and the job corresponding to the job. Analysis result of the job including average execution time and C Job management control means for allocating an optimal CPU resource capable of executing the job in the shortest time based on the state of the U resource, executing the job, and holding execution information of the job in the job management database; It is characterized by having.
[0008]
According to the present invention, the resource management unit manages the state of CPU resources on the network, the job management database holds past job execution information including the average job execution time, and when a job is input, Job analyzing means analyzes the contents of the job with reference to the job execution information held in the job management database, and the queue managing means determines if there is no free CPU resource capable of executing the job. A job is managed as a queue, and the job management control unit can execute the job in the shortest time based on the analysis result of the job including the average job execution time corresponding to the job and the state of the CPU resource. An optimum CPU resource is allocated, the job is executed, and the execution information of the job is stored in the job management data. Database, the CPU resources are optimally allocated, the execution time of the entire job including the licensed software can be optimized, the job can be executed efficiently, and the entire system can be executed. Since the execution of the job can be optimized, the number of licenses to be contracted, such as licensed software, can be minimized.
[0009]
The job execution management device according to the next invention is the job execution management device according to the above invention, wherein the job is a job for executing a predetermined number of license software, and a license management unit that manages an execution state of the predetermined number of license software. Further, the job management control means, when a job of the license software is requested, based on a management result of the license management means, when there is no free space for a predetermined number of license software, the job management control means It is characterized by inputting to a queue.
[0010]
The job execution management device according to the next invention is characterized in that, in the above invention, the queue management unit updates the order of the queue based on the analysis result of the job and the state of the CPU resource. I do.
[0011]
The job execution management device according to the next invention is the job execution management device according to the above invention, based on the analysis result of the job including the average job execution time, the state of the CPU resource, and the order of the queue. A job execution predicting unit that predicts at least one of an execution start time, a job execution time, and a job execution end time, wherein the job management control unit outputs the prediction information predicted by the job execution prediction unit to the job; It is characterized by notifying the requesting client.
[0012]
The job execution management device according to the next invention is the above-mentioned invention, wherein the job management control unit is configured to send the job execution prediction unit to the job execution prediction unit when a change occurs in the state of the CPU resource or the queue order. Job prediction information is re-predicted, and the re-predicted prediction information is notified to the client.
[0013]
In the job execution management device according to the next invention, in the above invention, the license management means is a license software assigned to a group to which the client belongs, a shared license software assigned to the entire system, or assigned to another group. Searching for usable license software in the order of shared license software.
[0014]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, an embodiment of a job execution management device according to the present invention will be described with reference to the drawings. It should be noted that the present invention is not limited by the embodiment.
[0015]
FIG. 1 is a block diagram showing a configuration of a job execution management device according to an embodiment of the present invention. FIG. 2 is a block diagram showing the overall configuration of a job execution management system including the job execution management device shown in FIG. 1 and 2, the job execution management system includes a job execution management device 1 that manages job execution, job execution devices 11 to 13 that execute jobs, and client devices 21 and 22 that issue jobs. Each is connected to the network 10. In this embodiment, a job for executing a predetermined number of licensed software will be described.
[0016]
The job execution management device 1 has a communication interface 2 for performing communication processing with a network 10, and a CPU resource management unit 8 and a job management unit 3 are connected to the communication interface 2. The CPU resource management unit 8 always grasps the status of the CPU resources of the job execution devices 11 to 13 as CPU resource information, and notifies the job management unit 3 when there is a change in the CPU resources.
[0017]
The job management unit 3 manages a job input from the client devices 21 and 22 by allocating an optimal CPU resource to the job and executing the job. A job analysis unit 4 is connected to the job management unit 3 and analyzes the input job contents. At this time, the job analysis unit 4 analyzes by referring to the past job execution information stored in the job management database 5. The past job execution information includes, for example, the average job execution time.
[0018]
A license management unit 6 is connected to the job management unit 3. The license management unit 6 determines whether the number of licenses of the license software corresponding to the input job has exceeded a predetermined number of contracts. Are input as a queue managed by the queue management unit 7, and are managed as a queue until a corresponding license software becomes available.
[0019]
A job execution prediction unit 9 is connected to the job management unit 3. The job execution prediction unit 9 also stores job analysis results including the average job execution time of the input job, the state of the CPU resource, and the order of the queue. At the same time, the execution start time of each job, the execution time of the job, the execution end time of the job, and the like are predicted, and the predicted information is notified to the job management unit 3. The job management unit 3 notifies the notified prediction information to the client devices 21 and 22 that issued the corresponding job.
[0020]
The job management unit 3 sequentially sends job execution instructions to the corresponding job execution devices 11 to 13 and, upon receiving the job execution end, causes the job management database 5 to hold the job execution information.
[0021]
Here, a job execution management control process centered on the job management unit 3 will be described with reference to a flowchart shown in FIG. In FIG. 3, first, the job management unit 3 receives a job for executing the license software from the client devices 21 and 22 via the communication interface 2 (step S101), and the job analysis unit 4 analyzes the job. (Step S102). A job includes not only a command for starting an application such as licensed software, but also usually includes designation information such as an option and an input file at the time of executing the licensed software. Specification of priority reservation of CPU resources, and the like can be included. The job analysis processing (step S102) will be described later.
[0022]
Thereafter, a process of acquiring the highest priority job from the queue managed by the license management unit 6 based on the job analysis result is performed (step S103). Details of this processing will also be described later.
[0023]
Thereafter, the job management unit 3 determines an optimal CPU resource corresponding to the highest priority job (Step S104). That is, an optimal CPU resource for executing a job is allocated. Based on the assignment result, the job management unit 3 sends a job execution command to the corresponding CPU resource (step S105).
[0024]
The job management unit 3 causes the job execution prediction unit 9 to execute the job execution prediction processing in response to the execution of the job or the input to the queue of the job (step S106). As described above, the job execution prediction unit 9 performs the execution start time of each job, the execution time of the job, the job execution time, based on the job analysis result including the average job execution time, the state of the CPU resource, and the queue order. To calculate the execution end time of the job, and notifies the job management unit 3 of the result. Upon receiving the notification of the prediction information, the job management unit 3 notifies the client devices 21 and 22 that issued the job of the predicted job execution time. The job execution start time and the job execution end time may be notified in addition to the predicted job execution time.
[0025]
Thereafter, when receiving the job end notification, the job management unit 3 updates the information in the job management database 5 corresponding to the ended job (step S108), and ends this processing.
[0026]
FIG. 4 is a flowchart illustrating a detailed processing procedure of the job analysis processing in step S102. 4, when the job analysis unit 4 receives a job from the job management unit 3 (step S201), the job analysis unit 4 analyzes the job (step S202). That is, as shown in FIG. 5, a job ID indicated by a job is assigned, and an application name, an execution option, an input file size, a required memory size, an executor name, an E-mail address, a priority, and the like are obtained.
[0027]
Thereafter, the job analysis result is compared with the corresponding job in the job management database 5 (step S203), and the average job execution time of this job is calculated with reference to the past average job execution time (step S204). ). As a result, a job analysis result to which the job average execution time data shown in FIG. 5 is added is output to the job management unit 3 (step S205), and the process returns to step S102. Note that the activation place name is determined in step S104.
[0028]
Next, with reference to the flowchart shown in FIG. 6, the detailed processing procedure of the processing of acquiring the highest priority job in step S103 will be described. In FIG. 6, the license management unit 6 determines whether the number of contracts of the license software designated by the input job has not exceeded, that is, whether there is a free license (step S301). If there is a vacant license (step S310, Yes), the license is output to the queue with a license (step S302), the order of the queue is updated, and the process proceeds to step S306. This queue with a license is a queue of jobs that are provided in the job management unit 3 and that are executed prior to the queue in the queue management unit 7.
[0029]
On the other hand, if there is no available license (step S310, No), the job is output to the queue managed by the queue management unit 7 (step S304), and the job order in the queue is updated (step S304). Step S305), and then proceed to step S306.
[0030]
In step S306, if there is a queue with a license, the highest priority job is taken out of this queue. If there is no queue with a license, the top priority job is taken out of the queue managed by the queue management unit 7. , And returns to step S103.
[0031]
FIG. 7 is a diagram illustrating an example of a queue managed by the queue management unit 7. The job ID is assigned a number in the order of input of the job, and indicates that the smaller the numerical value of the job ID, the earlier the job is input to the job execution management device 1. Here, the job IDs “0004” and “0008” are jobs that execute the same application “BB” (license software), and the job ID “0004” is earlier than the job ID “0008”. Since the priority “5” of the job ID “0008” is higher than the priority “1” of the job ID “0004”, the order of the queue is changed. FIG. 7 shows the time when the job ID “0009” is input to the queue. When the job ID “0009” is input, the job ID of the same application “AA” is input as indicated by the arrow. Since the job ID “0009” is compared with the job ID “0005” and the job ID “0009” has a higher priority, the order of the queues of the job ID “0009” and the job ID “0005” is changed thereafter. Will be. Here, the order of the queue is changed for each application. In addition, when the order of the queue changes, the job execution start time and the like are changed. Therefore, the job management unit 3 notifies the issue source of each job of time information accompanying the change. Further, the change of the queue order is also caused by the change of the CPU resource and the like, and also in this case, the job management unit 3 notifies the job issuer of the change of the time information.
[0032]
FIG. 8 is a diagram illustrating a specific example of a specific job execution management system. As shown in FIG. 8, the license is usually configured as a network in which LANs such as offices A, B, and C and a head office H are interconnected. The job execution management device 1 is provided in the head office H. In the above-described embodiment, the license management unit 6 has described that all the license software on the network is centrally managed. However, the license management may be managed in a distributed manner. However, the management contents distributed and managed are notified, for example, to the license server of the head office H.
[0033]
For example, the license software AA, BB, CC is managed by a license server on each LAN. The client manages the license software separately into license software that can be exclusively used by client devices in the local LAN and license software that can be shared and used by client devices on each LAN. Is done. In addition, when executing the license software corresponding to the job, in order of the dedicated license software in the own business office, the shared license software in the own business office, the shared license software in the head office, and the shared license software in the other business office In this case, a search is made to determine whether there is a license available.
[0034]
For example, when the client device A1 in the business establishment A issues a job for executing the license software AA, the processing corresponding to the license management unit 6 is performed in a distributed manner. Is searched, and if there is no free space, the shared license software AA in the office A is searched. If there is no free space, the shared license software AA in the head office H is further searched. If not, the shared license software AA in the other establishments B and C is searched. If there is no free space, it is input to the queue. However, the queue is centrally managed by the job execution management device 1 in the head office H.
[0035]
In this embodiment, licenses on the network are centrally managed, and optimal CPU resources are assigned based on the results of job analysis, so that the execution time of the entire licensed software can be optimized. Further, since the license software can be optimized in the entire system, the number of licenses to be contracted can be minimized. Further, the client can grasp the execution time of the requested job almost in real time, and does not affect other work processes in the client.
[0036]
In the above-described embodiment, the job has been described as a job for licensed software. However, the present invention is not limited to this, and a job in which a normal job other than a job for licensed software is mixed can be processed simultaneously.
[0037]
【The invention's effect】
As described above, according to the present invention, the resource management unit manages the state of the CPU resource on the network, and the job management database holds the past job execution information including the average job execution time, When a job is input, job analysis means analyzes the contents of the job with reference to job execution information held in the job management database, and a queue management means determines a CPU resource capable of executing the job. If there is no free space in the job, the job is managed as a queue, and the job management control means controls the job based on the analysis result of the job including the average job execution time corresponding to the job and the state of the CPU resource. Allocate an optimal CPU resource that can execute a job in the shortest time, execute the job, and execute the job. Information is held in the job management database, so that optimal CPU resources are allocated, the execution time of the entire job including the licensed software can be optimized, and efficient job execution can be performed. In addition, job execution can be optimized in the entire system, so that the number of licenses to be contracted, such as licensed software, can be minimized.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating a configuration of a job execution management device according to an embodiment of the present invention.
FIG. 2 is a block diagram showing a configuration of a job execution management system including the job execution management device shown in FIG.
FIG. 3 is a flowchart illustrating a job management process performed by a job management unit.
FIG. 4 is a flowchart illustrating a detailed processing procedure of a job analysis process in step S102.
FIG. 5 is a diagram showing the contents of a job analysis result.
FIG. 6 is a flowchart illustrating a detailed processing procedure of a top priority job acquisition process in step S103.
FIG. 7 is a diagram illustrating an example of a queue.
FIG. 8 is a block diagram illustrating a specific example of a job execution management system.
[Explanation of symbols]
Reference Signs List 1 job execution management device, 2 communication interface, 3 job management unit, 4 job analysis unit, 5 job management database, 6 license management unit, 7 queue management unit, 8 CPU resource management unit, 9 job execution prediction unit, 10 network , 11 to 13 job execution devices, 21 and 22 client devices.

Claims (6)

A job execution management device that is arranged on a network and manages a job requested by a client by allocating the job to an appropriate CPU resource on the network and executing the job.
Resource management means for managing the state of CPU resources on the network;
A job management database that holds past job execution information including an average job execution time;
Job analysis means for analyzing the contents of the job with reference to job execution information held in the job management database,
Queue management means for managing the job as a queue when there is no available CPU resource capable of executing the job;
Based on the analysis result of the job including the job average execution time corresponding to the job and the state of the CPU resource, an optimal CPU resource capable of executing the job in the shortest time is allocated, and the job is executed. A job management control unit for storing execution information of the job in the job management database;
A job execution management device comprising: The job is a job for executing a predetermined number of licensed software,
A license management unit that manages the execution status of the license software of a predetermined contract number,
The job management control means includes:
When a job of the license software is requested, based on the management result of the license management means, if there is no free space for a predetermined number of license software, the job is input to the queue. The job execution management device according to claim 1. 3. The job execution management device according to claim 1, wherein the queue management unit updates the order of the queue based on an analysis result of the job and a state of the CPU resource. Based on the analysis result of the job including the average job execution time, the state of the CPU resource, and the order of the queue, the execution start time, the job execution time, or the job execution end time of the job A job execution prediction unit for predicting at least one of the job execution prediction units;
4. The job execution management device according to claim 1, wherein the job management control unit notifies the client requesting the job of prediction information predicted by the job execution prediction unit. The job management control unit, when a change occurs in the state of the CPU resource or the queue order, causes the job execution prediction unit to re-predict the prediction information of the job. The job execution management device according to claim 4, wherein information is notified to the client. The license management unit searches for available license software in the order of license software assigned to the group to which the client belongs, shared license software assigned to the entire system, and shared license software assigned to another group. The job execution management device according to any one of claims 2 to 5, wherein

Images