JP2004240671A - Processing method and system for distributed computer - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To select an optimum computer resource which can satisfy the request from a customer terminal in which a job process is requested. <P>SOLUTION: This system provides a distributed processing service of the computing in service mediation organizations 19, 20 connected with a customer terminal 10 in which the job process is requested through a network 40 and a plurality of distributed computer resources 30 for executing the job processing through the network 40. The system finds parameters containing a computer resource including the CPU performance, the memory capacity and the disk capacity of the plurality computer resources 30, the accuracy of a job processing result, and the reliability of a processing period. The system provides points to the respective parameters, and calculates the total points of all the parameters. Thereby, each computer resource is ranked by using a point value, and the optimum computer resource for the job processing requested from the customer terminal by the point value ranked is selected. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
According to the present invention, in a distributed processing service of computing sharing computer resources distributed on the Internet, in order to perform job processing efficiently, optimal computer resources are allocated to the job processing and the processing requirements of the customer terminal are satisfied. The present invention relates to a service method and system for outputting a processing result to be performed and providing a reward according to the result to the processing result.
[0002]
[Prior art]
In recent years, with the spread of the Internet, computer networks surrounding the entire world have been improved, Web services have been further enhanced, and a computing environment in which information can be transmitted and collected worldwide is being prepared. Along with this, there is an increasing need to share distributed computer resources, such as efficient use of computer resources that are distributed throughout the world on the Internet for efficient computation (grid computing).
[0003]
With the aim of using computer resources that have traditionally been distributed on the Internet, home and office personal computers that use the computer resources that are not being used through the Internet to perform complex processing such as decryption and medical research. A distributed processing service of a computer, that is, a processing method or apparatus for processing a distributed computer is already known. This distributed processing service of computing is a distributed processing service of computing that requests job processing from a customer terminal (for example, processing of settlement data generated at a financial institution, totalization processing of customer data generated at a store, etc.). Among the computer resources that perform job processing with the intermediary agency of the processing service, the intermediary agency of the distributed processing service of computing determines the selection of the computer resource that performs the job processing based only on the computer resource such as hardware specifications.
[0004]
In other words, the comparison is only based on the design specifications and the theory, and when the processing is actually performed, the complex elements such as the real memory use area (the memory area actually required by the job) and data errors and errors are considered. Therefore, the accuracy of the job processing result requested by the customer terminal may not be satisfied, or the processing time limit may not be strictly observed.
[0005]
In addition, an intermediary agency of a distributed processing service of computing ranks computer resources according to the number of times of execution of job processing, thereby motivating a computer resource owner to perform job processing.
[0006]
Further, as another conventional technique, a client terminal that owns job processing such as printout of a job processing result, a plurality of computer resources for actually executing the job processing, and a request for the job processing on behalf of the customer, In a distributed processing system including a management device that manages computer resources, the management device receives the job weighted for processing time, print quality, and cost from a customer terminal, and recruits computer resources capable of processing the job. Then, an estimate is given to each applied computer resource. Thereafter, the management device compares the estimation results submitted from the respective computer resources based on the weights of parameters such as the processing time, print quality, and cost set by the customer terminal. The estimation result is calculated based on the processing speed between the computer resources and the printer, the capacity of the job, and the unit price per output sheet. Accordingly, provision of a distributed processing system for selecting an optimal computer resource has been proposed (see Patent Document 1).
[0007]
[Patent Document 1]
JP-A-9-198354
[0008]
[Problems to be solved by the invention]
However, in the above-described related art, the ranking is limited to the number of times that the job processing has been performed, regardless of the accuracy of the job processing result and whether or not the job processing time limit is strictly observed. That is, the motivation of the computer resource owner to work largely depends on the reward paid for performing the job processing, regardless of the accuracy of the job processing result and whether or not the job processing deadline is strictly adhered to. As a result, in order to perform job processing frequently, some computer resource owners intentionally report false job processing results and increase the number of times of job processing. .
[0009]
In other words, according to the prior art, the most suitable computer resource for performing the job processing is selected based on hardware specifications such as the computing performance of the computer resource and a criterion for suppressing the processing cost. As a result, when the job processing is actually executed, there is a problem that some jobs do not satisfy the accuracy of the job processing result requested by the customer terminal and the job processing deadline cannot be strictly observed.
[0010]
Further, in the technique of Patent Document 1 described above, computer resources are selected based on only three parameters such as processing time, print quality, and cost, and are merely theoretical comparisons. For this reason, when the processing is actually performed, job characteristics such as the job processing depending on the computation processing performance or memory processing performance of computer resources, or complex elements such as an actual memory use area and data errors and errors are considered. Since no consideration is given, the accuracy of the job processing result requested by the customer terminal is not satisfied, and the processing time may exceed the predicted value of the processing time.
[0011]
In addition, the selection of the optimal computer resource is performed based on the estimation result. However, when a plurality of computer resources return the same estimation result in one of the parameters, the computer resource with the earliest processing start time is selected. select. On the other hand, when a plurality of computer resources return different estimation results, a plurality of computer resource candidates are presented to the customer terminal, and the computer resource is selected based on the judgment of the customer terminal. For this reason, even if one parameter is selected as the optimal computer resource, the other parameter may be a very inappropriate computer resource, and each parameter is comprehensively determined to select the optimal computer resource. The challenge is not to.
[0012]
Therefore, an object of the present invention is to consider the job characteristics such as depending on the arithmetic processing performance or the memory processing performance of the computer resources when selecting the optimum computer resources that can satisfy the request from the customer terminal according to the job characteristics. In this way, past results based on the reliability of job processing deadlines, such as the accuracy of job processing results and the strict adherence to delivery dates, are included in the selection criteria for computer resources, and these are comprehensively determined to select the optimal computer resources. An object of the present invention is to provide a processing method and a system for a distributed computer. Furthermore, based on past performance based on the reliability of job processing deadlines, such as the accuracy of job processing results and whether delivery deadlines were strictly observed, the rewards for each computer resource owner are differentiated, and rewards are given according to the results. To motivate computer resource owners to work.
[0013]
[Means for Solving the Problems]
In order to solve the above problems, the present invention mainly employs the following configuration. A processing method of a distributed computer in a service intermediary institution connected to a customer terminal for requesting job processing and a plurality of distributed computer resources for executing job processing,
CPU parameters of the plurality of computer resources, memory capacity, computer resources including disk capacity, the accuracy of the job processing result, the reliability of the processing deadline, and obtain parameters including,
By assigning points to each of the parameters and calculating the total points of all parameters, each computer resource is rated by a point value,
The computer is configured to select an optimal computer resource for a job process requested from the customer terminal using the rated point value.
[0014]
Further, the unit is configured to include, as the parameter, a unit price of a computer resource that fluctuates according to a job processing result.
[0015]
Further, by associating a request content according to a characteristic of job processing from the customer terminal with a point of the parameter, an optimum computer resource corresponding to the request content from the customer terminal is selected.
[0016]
In addition, the job processing result output from each computer resource is obtained, the next unit price is calculated based on the accuracy of the job processing result and the reliability of the job processing deadline, and a reward according to the job processing result is given to the computer resource. Is sent to the computer resource owner having
[0017]
By employing the above-described configuration, the present invention can select the optimum computer resources and efficiently select the optimum computer resources that can satisfy the request from the customer terminal according to the job characteristics. In this way, it is possible to encourage the computer resource owner to work, and to reduce the number of reports of intended false job processing results.
[0018]
BEST MODE FOR CARRYING OUT THE INVENTION
A processing method and system of a distributed computer according to an embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is an overall configuration diagram of a distributed processing service providing system for computing according to an embodiment of the present invention. FIG. 2 is a processing flow of a distributed processing service providing method of computing according to the present embodiment.
[0019]
In FIG. 1, a customer terminal 10 represents a client actually holding a job. A job is a data processing operation in which some processing result is output using an arithmetic processing function or a database function of a computer, such as processing of settlement data by a financial institution or totalization of customer data by a store.
[0020]
The computer resource 30 is a computer resource that belongs to the intermediary agency 20 of the distributed processing service of computing and actually performs job processing. The intermediary agency 20 of the distributed processing service of computing is an intermediary agency that performs a job processing request from the customer terminal 10 to the computer resource 30. The intermediary organization 20 of the distributed processing service of computing reports the reliability of the service such as the service charge, which is the job processing cost, the accuracy of the job processing result, and whether the job processing deadline was strictly adhered to the customer terminal 10. Specify.
[0021]
In addition, the intermediary agency 20 of the distributed processing service of computing specifies the hourly wage per unit time at the time of processing the job of the computer resource 30 to the computer resource 30. The criterion for the customer terminal to select a specific computing distributed processing service intermediary agency 20 from a plurality of computing distributed processing service intermediary agencies 20 is that the result of the job processing is accurate and the job processing deadline is strictly adhered to. Depends on the reliability of the service. On the other hand, from the computer resources 30, the mediation institution 20 of the distributed processing service of computing is selected based on a criterion such as a high unit price as a criterion of the job processing work.
[0022]
In FIG. 2, the intermediary agency 20 of the distributed computing service distributes, via the Internet 40, a plurality of computer resources 30 that actually perform job processing in the distributed computing service method of the homepage of the computer resource acquisition unit 21. Recruit and enclose in (200). In addition, at the time of enclosing, a penalty for failing to adhere to the unit price and time limit set for each intermediary agency 20 of the distributed processing service of computing, and the fact that the reward is based on the result are clearly indicated. When the computer resources 30 agree, the information such as the identification of the owner of the computer resources 30, the account number, and the hardware specifications of the owned computer resources is notified by a key encryption method or the like, and a registration request is made (300).
[0023]
The client terminal 10 operated by the client makes a job processing request to the intermediary agency 20 of the distributed processing service of computing (100). A contract negotiation is performed based on the job processing deadline and processing cost (201), and in the case of a mutual agreement, a notice of the agreement is received from the intermediary agency 20 of the distributed processing service of computing (202). Thereafter, the customer terminal 10 sends the job for which a processing request is requested to the intermediary agency 20 of the distributed processing service of computing (101).
[0024]
The intermediary agency 20 of the distributed processing service of computing receives the job and selects the optimal computer resource 30 (203). This selection method will be described later in detail with reference to FIGS.
[0025]
The intermediary agency 20 of the distributed processing service of computing negotiates a contract with the owner of the selected plurality of optimal computer resources 30 for a job processing deadline (204). When the owner of the computer resource 30 agrees with the job processing, the job processing is assigned to the computer resource 30 (205). If the owner of the computer resource 30 does not agree, the contract is negotiated with another appropriate owner of the computer resource 30 (204). The intermediary agency 20 of the distributed processing service of computing sends the job and the application for executing the job processing to the assigned plurality of computer resources 30 (206).
[0026]
The plurality of computer resources 30 use the application to perform the same job processing (302), and send the processing result to the intermediary agency 20 of the distributed processing service of computing after the processing is completed (303). The service intermediary agency 20 of the distributed computing process compares the processing results sent from the plurality of computer resources 30 and determines whether the results are accurate. Conventionally, whether a job processing result is accurate is determined by comparing processing results sent from a plurality of computer resources 30.
[0027]
Then, when the processing result of the computer resource 30 is correct, the reward calculating unit 27 calculates a reward for the job processing (207). The method of calculating this will be described in detail in the description of FIGS. 3, 5, and 8 described later. Further, if the processing result of the computer resource 30 is incorrect, a penalty is imposed based on a contract defined for each intermediary agency 20 of the distributed processing service of computing.
[0028]
The intermediary agency 20 of the distributed processing service of computing sends the job processing result and the bill for the processing cost to the customer terminal 10 that has requested the job processing (208). After confirming the output result of the job processing and the bill, the customer terminal 10 pays the intermediary agency 20 of the distributed processing service of computing (102). The intermediary agency 20 of the distributed processing service of computing pays the reward calculated by the reward calculator 27 to the owner of each computer resource 30 (209). Thus, a series of operations in the job processing according to the present embodiment are all completed.
[0029]
Next, a method for selecting the most suitable computer resource 30 for the job processing described above will be described with reference to FIGS. 3, 4, 5, 6, and 7. FIG. FIG. 3 is a diagram showing computer resource information 220 managed by the computer resource information storage unit 22 of the intermediary agency 20 of the distributed processing service of computing. The parameters store information on the computer resources 30. In FIG. 3, a CPU performance 221, a memory capacity 222, and a disk capacity 223 are hardware specifications of computer resources owned by the computer resources 30. Note that the disk capacity indicates a usable free space.
[0030]
The accuracy 224 of the processing result indicates whether the job processing result was accurate by adding or subtracting points. If the accuracy of the job processing result can be satisfied, one point is added, and if not, one point is subtracted. The initial value is 50 points, and addition and subtraction are performed for each job processing. Note that the initial value can be defined for each intermediary institution 20 of the distributed processing service of computing.
[0031]
The job processing time limit reliability 225 is a value calculated from the contracted required days and the actually shortened or delayed days in order to determine whether the computer resource 30 is strictly following the job processing time limit. The calculation formula is (1 ± shortened (or deferred) period / contract period) × charge for each intermediary institution 20 of distributed processing service of computing × unit price 226. For example, if a computer resource with a unit price of 1000 yen submits job processing with a deadline of 10 days in 8 days, the next unit price is (the rate of the intermediary institution is 0.9), (1 + 2/10) × 0. 9 x 1000 = 1080 yen. This ± is plus (+) if earlier than the deadline, and minus (-) if later. That is, the earlier the job processing result is output, the higher the numerical value. Further, the fee for each intermediary agency 20 of the distributed processing service of computing can be defined for each intermediary agency 20 of the distributed processing service of computing. However, if the job processing result is output on time, then 0 or more.
[0032]
The unit price 226 is an amount that fluctuates according to the result of the job processing of the computer resource 30. The calculation formula is (1−unit price 226 / maximum unit price of each computer resource) × 100 + additional points for each intermediary agency 20 of the distributed processing service of computing. For example, in the point calculation formula of a unit price of 1000 yen, if there is an owner whose maximum unit price in each computer resource is 2000 yen (additional points for each intermediary institution are 50 points), (1-1000 / 2000) × 100 + 50 = 100 points. Therefore, a low unit price 226 is a high point. Further, the additional points can be defined for each intermediary agency 20 of the distributed processing service of computing, but in order to obtain consistency with the point values of other parameters, 50 points are added in the example of FIG. 4 described later. (Refer to the unit price column of the point conversion formula 232 in FIG. 4). The initial value of the unit price 226 can be defined for each intermediary agency 20 of the distributed processing service of computing. Further, each parameter can be defined for each intermediary agency 20 of the distributed processing service of computing.
[0033]
FIG. 4 is a diagram showing a point conversion flow in the resource / actual point conversion unit 23 of the intermediary institution 20 according to the present embodiment. The resource / actual point conversion unit 23 is a conversion unit that associates the numerical value of the computer resource information 220 of FIG. 3 with a point value. First, information on the computer resources 220 in FIG. 3 is obtained (230). The intermediary agency 20 of the distributed processing service for computing sets a reference value in order to ensure the consistency of the numerical values of the parameters of the computer resource information 220 in FIG. That is, each computer resource is rated using the hardware specification of the computer resource considered as a standard by the intermediary agency 20 of the distributed processing service of computing as a reference value.
[0034]
In the present embodiment, numerical values as shown in (i) to (n) of the reference numeral 231 are applied, but these numerical values can be defined for each intermediary agency 20 of the distributed processing service of computing. Also, the numerical value of the computer resource information indicated by reference numeral 231 in FIG. 4 is associated with 100 points (h). Next, the obtained numerical value of each parameter is converted into a point value (232). The point conversion formula defines the conversion formula of each parameter from (i) to (l) as indicated by reference numeral 233. As for (m), the same numerical value is used. The conversion formula of the unit price in (n) is defined as indicated by reference numeral 234. Here, x in the conversion formula is a variable that specifies computer resources. x is set in the order of the computer resources A, B, C,... in FIG. Also, the highest unit price is set for y in the parameter of (n) unit price.
[0035]
Reference numeral 232 indicates a conversion example of the computer resource A. Finally, the point value is output to the point table for each parameter in FIG. 5 (235). Thereby, the numerical value of the computer resource information 220 shown in FIG. 3 can be converted into a point value. Each point table 240 is managed by the resource / actual point management unit 24 (see FIG. 1) of the intermediary agency 20 of the distributed processing service of computing.
[0036]
The job characteristic parameter screen 250 shown in FIG. 6 is managed by the job characteristic parameter input unit 25 (see FIG. 1), and the intermediary agency 20 of the distributed processing service of computing responds to a request from a customer terminal according to the job characteristic. 7 is a setting screen for assigning priorities to respective parameters based on the setting screen. For example, in the input example in the case where a payment processing job is requested with a time limit from a financial institution, the processing result in the payment processing is an element that requires the most accuracy. Therefore, the flag 1 is set to the parameter 252 of the accuracy of the processing result. The next important thing is the required output time of the processing result, and the flag 2 is set to the processing capacity 251.
[0037]
Also, the priority is set to each parameter according to the job characteristics of the job processing, such as the CPU-dependent type 254, the memory-dependent type 255, or the disk-dependent type 256. Here, the CPU-dependent job 254 is a job that mainly requires arithmetic processing performance. The memory-dependent job 255 is a job that mainly requires a memory capacity. Further, the disk-dependent job 256 is a job having a high occupation ratio of the disk capacity because the disk capacity of the computer resources may be insufficient when the capacity of the job is large. Finally, the flag 3 is set for the price 253.
[0038]
The relationship between FIG. 5 and FIG. 6 is that FIG. 5 is a point table for each parameter such as the CPU performance 241 and the memory capacity 242, and the computer resource selection unit 26 has the highest priority according to the priority of the parameters set in FIG. Confirm the parameters. Here, since the computer resources 30 are rated by point values for each parameter, it is possible to select the optimum computer resources 30 that can satisfy the request from the customer terminal according to the characteristics of the job processing. Specifically, the highest priority parameter set in FIG. 6 is set as the primary key. Then, for each computer resource 30, the point value of the primary key is doubled and added to the point values of all other parameters.
[0039]
FIG. 7 shows an example of calculating points of the computer resource 30 using the accuracy (accuracy) 244 of the job processing result according to the present embodiment as the top priority key. If the primary keys have the same value, an appropriate priority of the computer resources 30 is determined in consideration of the point values of other parameters. Thereby, the plurality of computer resources 30 can be ranked based on the computer resources such as hardware specifications, the accuracy of the job processing result, and the reliability of the job processing due date.
[0040]
Further, by determining the request of the customer terminal that has requested the job processing, and correlating the plurality of rated computer resources with the customer terminal request based on the job characteristics, the optimum computer that can satisfy the request from the customer terminal according to the job characteristics. Resources 30 can be efficiently selected. As described above, according to the service providing method according to the embodiment of the present invention, by selecting the most suitable computer resource 30 for the job processing, the processing that can satisfy the request from the customer terminal 10 according to the job characteristics Can provide results.
[0041]
Although the point value of the primary key having the highest priority is doubled in this embodiment, the rating of the primary key can be defined for each intermediary agency 20 of the distributed processing service of computing. Also, the intermediary agency 20 of the distributed processing service for computing assigns a priority to each parameter on the job characteristic parameter screen 250 in response to a request from the customer terminal 10 according to the job characteristic. It is also possible to make the customer terminal 10 directly input. In this case, it is possible to efficiently select the optimal computer resources 30 without going through the intermediary agency 20 of the distributed processing service of computing.
[0042]
Next, a reward method according to the result of the job processing according to the embodiment of the present invention will be described below with reference to FIGS. 3, 5, and 8. FIG. FIG. 8 is a diagram illustrating a reward determination flow (charging flow) according to the present embodiment.
[0043]
The intermediary agency 20 of the distributed processing service of computing acquires the unit price 226 information of each computer resource 30 that has executed the current job processing from FIG. 3 (280). In the computer resource 30 that performs job processing for the first time, an initial value defined for each service intermediary agency 20 of the distributed computing is applied. In addition, the computer resource 30 receives and reports from the computer resource 30 the time required for the job processing (280). The service intermediary agency 20 of the distributed computing determines whether the job processing result is accurate (281), and if it is accurate, determines the reward for the current job processing based on the time required for the job processing and the unit price 226. (282).
[0044]
On the other hand, if the job processing result is not accurate, a penalty is imposed based on the contract of each intermediary agency 20 of the distributed processing service of computing, such as no remuneration or paying 50% of the original remuneration (283). . Next, it is necessary to reflect the reliability 225 of the job processing deadline in the reward. When the job processing is output on time or earlier than the deadline, the job processing is calculated by the formula 285. On the other hand, when the job processing result is output later than the time limit, it is calculated by the calculation formula of 286. That is, the earlier the job processing result is output, the higher the unit price 226 becomes.
[0045]
The calculated unit price 226 is used as the next unit price 226. Finally, the unit price 226 calculated this time is reflected in the computer resource information 220 in FIG. 3 (287), and the point table 240 for each parameter in FIG. 5 is updated. Thus, by reflecting the accuracy of the current job processing result and the reliability of the job processing deadline in the reward, it is possible to provide a service that can execute the reward according to the result to the computer resources 30.
[0046]
In the above description, regarding the present invention, the optimal selection among a plurality of computer resources by the intermediary institution has been mainly described, but the optimal selection of the intermediary institution by the customer terminal or the computer resources can also be included in the concept of the present invention. That is, as an intermediary institution, there is a demand to enclose computer resources with high job processing speed and reliability, and as a computer resource, there is a request to belong to an intermediary institution that pays a higher hourly wage. There is a request to request an intermediary agency that has good job processing speed and reliability and accepts job processing at a low price, and a plurality of intermediary institutions (as shown in FIG. 1) price job processing. Competition between each other occurs. For example, if an intermediary institution encloses high-quality computer resources at a low job processing unit price, it causes the computer resources to perform low-cost job processing and receives a large amount of job processing costs from the customer terminal, so The institution can receive the difference as profit. However, if the job processing cost is too high, the customer terminal requests another intermediary agency, and competition between the intermediary agencies occurs.
[0047]
Therefore, as shown in FIG. 1, if a client terminal and a computer resource can access a plurality of intermediaries and acquire information held by the intermediary institution, the customer terminal or the computer resources can be obtained. The most suitable intermediary can be selected.
[0048]
【The invention's effect】
According to the present invention, a plurality of computer resources are classified based on results such as the accuracy of job processing results and the reliability of processing deadlines, and a request from a client terminal that has requested job processing is determined. By associating resources with customer terminal requests based on job characteristics, it is possible to efficiently select optimal computer resources that can satisfy requests from customer terminals according to job characteristics.
[0049]
In addition, computer resources can provide rewards according to their achievements based on their results, such as the accuracy of processing results and the reliability of processing deadlines. , It is possible to reduce the number of reports of job processing results.
[Brief description of the drawings]
FIG. 1 is a diagram showing an overall configuration of a distributed processing service providing system for computing according to an embodiment of the present invention.
FIG. 2 is a diagram showing a processing flow of a distributed processing service providing method of computing according to the present embodiment.
FIG. 3 is a diagram illustrating computer resource information managed by a computer resource information storage unit of an intermediary agency of a distributed processing service of computing according to the embodiment;
FIG. 4 is a diagram showing a point conversion flow in a resource / actual point conversion unit of an intermediary institution according to the present embodiment.
FIG. 5 is a diagram showing a point table for each parameter of computer resource information according to the embodiment.
FIG. 6 is a diagram illustrating a setting screen for assigning priorities to job characteristic parameters according to the embodiment;
FIG. 7 is a diagram illustrating an example of computer resource point calculation using the accuracy of a job processing result as a top priority key according to the embodiment;
FIG. 8 is a diagram showing a reward determination flow (charging flow) according to the present embodiment.
[Explanation of symbols]
10 Customer terminals
20 Intermediaries for distributed computing services
21 Computer Resources Acquisition Department
22 Computer Resource Information Storage Department
23 Resource / Result Point Conversion Unit
24 Resources / Results Point Management Department
25 Job characteristic parameter input section
26 Computer Resource Selection Department
27 Reward calculator
30 Computer Resources
40 Internet and other networks
220 Computer resource information
240 Point table for each parameter
250 Job characteristic parameter screen

Claims (5)

A processing method of a distributed computer in a service intermediary institution connected to a customer terminal for requesting job processing and a plurality of distributed computer resources for executing job processing,
CPU parameters of the plurality of computer resources, memory capacity, computer resources including disk capacity, the accuracy of the job processing result, the reliability of the processing deadline, and obtain parameters including,
By assigning points to each of the parameters and calculating the total points of all parameters, each computer resource is rated by a point value,
A processing method for a distributed computer, comprising: selecting an optimal computer resource for a job process requested by the customer terminal based on the rated point value. In claim 1,
A processing method for a distributed computer, characterized in that a unit price of a computer resource that fluctuates according to a job processing result is included as the parameter. In claim 1 or 2,
A distributed computer that selects the optimum computer resource corresponding to the request content from the customer terminal by associating the request content according to the characteristic of the job processing from the customer terminal with the point of the parameter. Processing method. In claim 2,
Obtain the job processing results output from each computer resource, calculate the next unit price based on the accuracy of the job processing results and the reliability of the job processing deadline, and have the computer resources rewarded according to the job processing results A processing method for a distributed computer, which is sent to a computer resource owner. A processing system of a distributed computer in a service intermediary agency connected to a customer terminal for requesting job processing and a plurality of distributed computer resources for executing job processing,
Means for acquiring parameters including computer resources including CPU performance, memory capacity, and disk capacity of the plurality of computer resources, accuracy of job processing results, and reliability of processing deadlines;
Means for attaching a point to each of the parameters to calculate a total point of all parameters, and rating each computer resource by a point value;
Means for selecting the most suitable computer resource for the job processing requested from the customer terminal based on the rated point value, a processing system for a distributed computer.

Images