JPH09288648A - Distributed processing method in network and its system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an efficient distributed processing system coping with the fluctuation of load conditions in a network. SOLUTION: A client calculates its own effective processing speed (procedure 1) and requests the processing of a task provided in a server to the server along with the information (procedure 2). The server judges which device is to process the task so as to shorten turn-around time relating to the task based on the received processing speed of the client and its own processing speed (procedure 3). By the judgement, the server himself processes the task (procedures 4a and 5a) or the client processes it (procedure 5b) after moving the task to the client (procedure 4b).

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a distributed processing method including a client and a server and a system thereof, and more particularly to a distributed processing method for reducing turnaround time of task processing.

[0002]

2. Description of the Related Art In recent years, with the development of network communication technology typified by the Internet and the sophistication of terminal devices typified by workstations, etc., a distributed processing system of a client / server type has become popular. As a first conventional method for realizing a client-server type distributed processing system, there is a method by remote procedure call (hereinafter referred to as "RPC method"). For example, "Concurrent Systems", Addiso
n-Wesley, 1992. Here, the client refers to a terminal on the network that requests processing of a specific task (hereinafter referred to as “specific task”),
The server refers to a terminal on the network that has received the request, but here it is assumed that the server does not always execute the processing of the specific task.

FIG. 9 (a) is an explanatory diagram showing communication exchanges between a client and a server in the RPC system.
FIG. 9B is a sequence diagram thereof. This method mainly consists of three procedures. That is, (1) the client sends a processing request message to the server in order to request processing of a specific task. (2) The server executes the requested specific task. (3) Upon completion of execution, the server returns a reply message to the client to notify the result.

For example, a client may
This corresponds to the case where the server has a database requesting a search process and a process of returning the result as a graph image, and the client receives the generated image information.
The advantage of this method is that a relatively low-priced terminal and communication means perform sophisticated processing as a whole system.

As a second conventional method for realizing a client / server type distributed processing system, there is a method by task migration. For example, SUN micr
JAVA ("Hooked on Java", Add advocated by osystems
This is the case as described in ison-Wesley, 1995) and those proposed by Netscape (hereinafter, this method is referred to as "JAVA method").

FIG. 10 (a) is an explanatory diagram showing communication exchanges between a client and a server in the JAVA system, and FIG. 10 (b) is a sequence diagram thereof. This method also consists of three major steps. That is, (1) the client sends a processing request message to the server in order to request processing of a specific task. (2) The server moves a specific task that packages a processing procedure and necessary data. That is, the specific task body is constructed as a message,
Reply to the client as a reply message.
(3) The client receiving the specific task executes the specific task by itself.

The advantage of this method is that by causing the client to perform task processing, it is possible to reduce the load on the server, which tends to concentrate the processing, and thus to achieve load distribution in the entire system.

[0008]

However, in the above-described conventional method, the turnaround time, that is, the processing result can be used after the processing request is generated, due to the fluctuation of the load on the client or the server. There is a problem that the time to reach the state is not always the minimum.

FIG. 11 is a model diagram showing a system configuration for explaining the problem. As shown in this figure, the turnaround time in each distributed processing method is evaluated as follows by modeling the processing performance, data transfer rate, and load status of the system. First, for each terminal device, when one task is newly added to the currently processed task, the speed at which each task is processed, that is, the effective processing speed is defined.

The maximum value of the processing speed of the client is Rcmax
(Work unit / second) and the number of tasks concurrently processed by the client is Nc (pieces), the effective processing speed Rc (work unit / second / task) of the client is Rc = Rcmax / (Nc + 1) ... Assuming that the maximum value of the processing speed of the server is Rsmax (work unit / second) and the number of tasks performing parallel processing in the server is Ns (pieces), The processing speed Rs (work unit / second · task) is represented by Rs = Rsmax / (Ns + 1) (Equation 2).

Here, the maximum processing speeds Rcmax and Rsmax are units showing the processing capability peculiar to each terminal device, and correspond to the SPECmark of the CPU of each device, and the work unit is the processing size of the task. This is a unit to represent, and corresponds to a normalized amount of calculation that can be processed in one second by an apparatus having an effective processing speed of 1 work unit / second. Furthermore, the work amount of a specific task is Mtask (work unit), and the communication speed of the communication path is Rt.
(Communication unit / second), communication volume of request message Mcall
(Communication unit), reply message communication volume Mrep
(Communication unit), the communication amount of a specific task accompanying task movement is defined as Mmig (communication unit).

Then, the turnaround time Trpc (second) in the RPC system is the sum of the times corresponding to the above-mentioned three procedures, and therefore Trpc = Mcall / Rt + Mtask / Rs + Mrep / Rt (Expression 3) On the other hand, the turnaround time Tjava (seconds) in the JAVA method is the sum of the times corresponding to the above-mentioned three procedures, and is therefore expressed as Tjava = Mcall / Rt + Mmig / Rt + Mtask / Rc (Equation 4). It

The turnaround time in both equations is evaluated using concrete numerical values as follows. Now, as the system performance, the maximum value Rcmax of the processing speed of the client is 40, and the maximum value Rsmax of the processing speed of the server is 10.
0, the communication speed Rt of the communication path is 5, and as each communication amount / work amount, the work amount Mtask of the specific task is 200, the communication amount Mcall of the request message is 5, the communication amount Mrep of the reply message is 5, and the task movement is It is assumed that the communication amount Mmig of the specific task associated with is 100.

As a first load condition, consider a case where the number of tasks Nc concurrently processed by the client is 1 and the number of tasks Ns concurrently processed by the server is 19. From Equations 1 and 2, the effective processing speeds Rc and Rs of the client and server are Rc = 40 / (1 + 1) = 20 Rs = 100 / (19 + 1) = 5. Therefore, from Equations 3 and 4, the RPC method is used. And JAV
Turnaround time Trpc and Tjava in method A
Is Trpc = 5/5 + 200/5 + 5/5 = 42 (seconds) Tjava = 5/5 + 100/5 + 200/20 = 31
(Seconds). FIG. 12 is a time chart showing this result.

As described above, under the first load condition, the turnaround time in the JAVA method is shorter than that in the RPC method. Therefore, it can be said that the JAVA method is a more preferable distributed processing method. Next, consider a second load situation in which the number of tasks Ns concurrently processed by the server has decreased from 19 up to that of 9 to 9.

When calculated in the same way as in the first load situation, the effective processing speed Rc of the client remains unchanged at 20, but the effective processing speed Rs of the server increases to Rs = 100 / (9 + 1) = 10. Therefore, the turnaround time Trpc in the RPC system is reduced to Trpc = 5/5 + 200/10 + 5/5 = 22 (seconds). On the other hand, the turnaround time Tjava in the JAVA method remains unchanged at 31 seconds. FIG.
This result is shown in a time chart.

As described above, in the second load situation, unlike the first load situation, the RPC method has a shorter turnaround time than the JAVA method. Therefore, the RPC method is more preferable. It can be said to be a distributed processing method. As described above, the magnitude relationship of the turnaround time in both systems is reversed as the load condition of the system fluctuates, but any conventional system is a fixed system that does not depend on the fluctuation of the load. Therefore, it cannot be said that a turnaround time smaller than that of the other scheme is necessarily provided. That is, depending on the load situation,
There is a problem that its performance drops sharply.

Therefore, the present invention has been made in order to solve the above-mentioned problems, and an efficient distributed processing method that can obtain a minimum turnaround time even when the load condition of the system changes, and The purpose is to provide the system.

[0019]

In order to achieve the above object, according to the distributed processing method and system of the present invention,
When the client issues a request to use the processing result of the specific task provided in the server, first, in order to reduce the turnaround time for the specific task, either the server or the client should process the specific task. If it is determined that the server should process, then the specific task is processed by the server and the result is reported to the client, while the client is determined to process. If the specified task is moved from the server to the client, the specified task is processed by the client.

As a result, every time a processing request is issued from the client, a processing method for reducing the turnaround time required for the processing is adopted.
The turnaround time is shortened as compared with the conventional method in which processing is fixedly performed by either the server or the client, and an efficient distributed processing system that responds to changes in the system load situation is realized.

[0021]

Embodiments of the present invention will be described below in detail with reference to the drawings. (Embodiment 1) FIG. 1 is a block diagram showing a hardware configuration of a distributed processing system according to Embodiment 1,
Only two related terminal devices 1 and 11 in the entire network system are shown.

This system comprises a client 1 and a server 11 connected via a communication path 20. The client 1 and the server 11 are the processor 2,
12, memory 3, 13 and communication units 4, 14. The processors 2 and 12 are CPUs and the like, and control the entire apparatus or process a plurality of tasks in parallel.

The memories 3 and 13 are composed of a ROM, a RAM, a hard disk, etc., and hold in advance control programs, system performance parameters, tasks, etc., which will be described later.
It is also used as a work area for the task being processed. The communication units 4 and 14 are Ethernet interface cards and the like, and relay the communication path 20 and the devices 1 and 11.

The client 1 and the server 11 are
The hardware configuration is the same as that of a terminal device such as a general-purpose workstation, but differs from the general-purpose terminal device in that the memories 3 and 13 hold a unique control program for implementing the present embodiment. The operation of the distributed processing system configured as above will be described. Now consider the same situation as described in the prior art. That is, the client 1 wants to obtain the processing result of the specific task held in the memory 13 of the server 11. The distributed processing method in this case will be described.

FIG. 2 is a sequence diagram showing an exchange of communication between the client 1 and the server 11 in the distributed processing method of this embodiment, which is shown in FIG. 9 (b) and FIG.
This corresponds to (b). This sequence diagram simultaneously shows the flow in two different cases. That is, the case consisting of the following steps 1-3, 4a, and 5a and the steps 1-3, 4
There are two cases consisting of b and 5b. (Procedure 1) First, the client 1 calculates its own effective processing speed prior to requesting processing of a specific task.

Specifically, the processor 2 detects its own processing load, that is, the number Nc of tasks loaded in the memory 3 and processed in parallel, and further, the maximum processing speed Rcmax stored in the memory 3 is detected. And read according to Equation 1,
The effective processing speed Rc is calculated. (Procedure 2) Subsequently, the client 1 transmits to the server 11 one message M'call indicating the calculated effective processing speed Rc and the fact that the processing result of the specific task held by the server 11 is to be used.

Specifically, the processor 2 transmits the transmission source, that is, the network address of the client 1 and information for identifying a specific task of the server to the server 11 via the communication unit 4. (Procedure 3) Server 11 that has received message M'call
Determines whether to process the specific task by itself or the client 1 in order to reduce the turnaround time for the specific task.

Specifically, the processor 12 includes the communication unit 1
By referring to the memory 13 that holds the specific task specified by the message M'call received via the message M'call, the work amount Mtask of this specific task, the communication amount Mmig accompanying the task movement, and the communication amount Mrep of the reply message Mrep And the effective processing speed Rc of the client 1 included in the message M'call is extracted.

Subsequently, the processor 12 detects the current average communication speed Rt of the communication path 20 based on the periodic observation of the communication path 20 by the communication unit 14. And
The processor 12 calculates its own processing load, that is, the effective processing speed Rs, as in the case of the client 1. Based on the above values, the processor 12 calculates and compares the following two types of times T1 and T2 consisting of the second and third terms of the equations 3 and 4, respectively.

T1 = Mtask / Rs + Mrep / Rt (Equation 5) T2 = Mmig / Rt + Mtask / Rc (Equation 6) In addition, T1 represented by Equation 5 and T2 represented by Equation 6 The magnitude relationship is the same as the magnitude relationship between the turnaround time Trpc by the RPC method expressed by Expression 3 and the turnaround time Tjava by the JAVA method expressed by Expression 4.
This is because the first terms of equations 3 and 4 are common.

As a result of the comparison, when T1 <T2, the server 11 judges that the server itself should process this specific task, and executes the following steps 4a and 5a. (Procedure 4a) The server 11 processes a specific task. Specifically, the processor 12 moves the specific task requested by the client 1 from the holding unit of the memory 13 to the work area, and executes it by time sharing in addition to the task already executed. (Procedure 5a) When the processing of the specific task is completed, the server 11
Returns the result obtained by the processing to the client 1 as a reply message.

Specifically, the processor 12 transmits the data obtained by executing the specific task to the client 1 via the communication unit 14 as a response to the request from the client 1. Steps 1 to 3 above, 4a, 5a
As a result, the specific task requested by the client 1 is processed by the server 11, and as a result, the conventional RP
This means that the distributed processing method similar to the C method has been performed.

On the other hand, in the comparison in the procedure 3, T1 ≧ T2
In this case, the server 11 determines that the client 1 should process the specific task, not the server 11 itself, and executes the following steps 4b and 5b instead of the above steps 4a and 5a. (Procedure 4b) The server 11 moves the specific task to the client 1.

Specifically, the processor 12 reads the specific task requested by the client 1 from the holding unit of the memory 13 and transfers it to the client 1 via the communication unit 14. Note that the information transferred here includes not only the specific task itself, that is, the processing procedure and necessary related data, but also a message indicating that the task has been moved in response to the request from the client 1 according to step 2. included. (Procedure 5b) Client 1 that received the specific task
Handles the specific task by itself.

Specifically, the processor 2 executes the specific task loaded in the work area of the memory 3 via the communication unit 4,
Execute by time sharing in addition to the already executed task. By the above steps 1-3, 4b, 5b,
The specific task requested by the client 1 is processed by the client 1 itself, and as a result, the conventional JA
This means that a distributed processing method similar to the VA method has been performed.

As described above, according to the present method, T1 <T2, that is, Trpc <Tjava, by considering the processing load situation in the system, that is, the number of tasks processed in parallel in the client 1 and the server 11. If it is determined that the task processing by the conventional RPC method is performed, on the other hand, if it is determined that T1 ≧ T2, that is, Trpc ≧ Tjava, the task processing by the conventional JAVA method is performed.

As a result, one of the RPC method and the JAVA method that can obtain a smaller turnaround time is dynamically determined according to the processing load of the system. FIG. 3 is a flowchart showing the operation of only the server 11 in the above sequence.

After receiving the processing request message M'call sent from the client 1 (step S31), the server 11 causes the communication unit 14 to communicate with the communication speed Rt of the communication path 20.
Is detected (step S32). Then, the content of the specific task to be executed is determined from the received message M'call, and the effective processing speed Rc of the client 1 is extracted (step S33).

Further, the effective processing speed Rs of the server 11 is obtained (step S34), the communication amount Mmig of the task accompanying the task movement, the number of operations Mtask required for processing, and the message required for reply at the end of the task processing. The communication amount Mrep is obtained (step S35). Using the above information, T1 represented by Equation 5 is compared with T2 represented by Equation 6 (step S36). If the former is smaller, the server 1
After performing the task processing in step 1 (step S37), the result is replied to the client 1 (step S3).
8) If not, the task body is moved to the client 1 so that the specific task is processed in the client 1 (step S39).

FIG. 4 is a flow chart showing the operation of only the client 1 in the sequence shown in FIG. First, the client 1 obtains the effective processing speed Rc (step S41) and uses it for the processing request message M'cal.
It is added to l and transmitted to the server 11 (step S4)
2).

Then, when the reply from the server 11 is received (step S43) and if it is the specific task body (step S44), the specific task is executed in the client 1 (step S45), otherwise. For example, the reply is used as the processing result of the specific task. The flow of the above sequence will be described using specific numerical values similar to those in the conventional art.

That is, as the system performance, Rcmax = 4
0, Rsmax = 100, Rt = 5, and as communication / work volume, Mtask = 200, M'call = 5, Mrep = 5, M
mig = 100, and the first load status is Nc = 1,
Consider the case of Ns = 19. Then, Rc = 40 / (1 + 1) = 20 Rs = 100 / (19 + 1) = 5, so that T1 = 200/5 + 5/5 = 41 T2 = 100/5 + 200/20 = 30. Therefore, in the above procedure 3, Since it is determined that T1 ≧ T2 and the client 1 should process the specific task, the above steps 4b and 5b, that is, the conventional JAVA method is adopted thereafter.

FIG. 5 is a diagram showing a time chart of this system under the first load condition in comparison with the conventional fixed RPC system. From Equations 3 and 4, the turnaround time (31 seconds) obtained by this method is shorter than that of the conventional method (42 seconds). Then, Ns =
Consider a second load situation that has decreased to 9.

Since Rs = 100 / (9 + 1) = 10 increases, T1 decreases to 200/10 + 5/5 = 21, but T2 remains (30). Therefore, in the above procedure 3, T1 <T2 Since it is determined that the client 1 should process the specific task, the procedures 4a and 5a described above, that is, the conventional RPC method are adopted.

FIG. 6 is a diagram showing a time chart of the present system in the second load situation in comparison with the conventional fixed JAVA system. From equations 3 and 4, the turnaround time (22 seconds) obtained by this method is
It is smaller than that of the conventional method (31 seconds). As described above, according to the present method, the situation of the processing load in the system at the time when the request for the task processing is generated is taken into consideration, so that a distributed turnaround time that is smaller than the RPC method or the JAVA method can be obtained. The processing method is dynamically determined. (Second Embodiment) Next, a distributed processing system according to a second embodiment of the present invention will be described.

The present embodiment is different from the first embodiment in the procedure from the generation of the processing request from the client to the determination of the apparatus to perform the processing, and is otherwise the same as the first embodiment. Therefore, only the points different from the first embodiment will be described here. FIG. 7 is a sequence diagram showing an exchange of communication between the client 1 and the server 11 in the distributed processing method of this embodiment. As can be seen by comparing with FIG. 2 in the first embodiment, in the present embodiment, the first embodiment
In place of the procedures 1 and 2, different procedures 1 to 4 are adopted, and the subsequent procedures are the same. (Procedure 1) First, the client 1 transmits to the server 11 only a message indicating that it wants to use the processing result of the specific task held by the server 11. Here, unlike the first embodiment, this message does not include information on the effective processing speed Rc of the client 1. (Procedure 2) The server 11, which has received the message, inquires the client 1 of the effective processing speed Rc, for example, when the processing load of the server 11 is lightened. (Procedure 3) The client 1 that has received the inquiry calculates its own effective processing speed Rc at the present time by the same method as in the first embodiment. (Procedure 4) Then, the client 1 transmits the effective processing speed Rc to the server 11. (Procedures 5, 6a, 7a, 6b, 7b) The operations in the client 1 and the server 11 are the same as those in the first embodiment.
Steps 3, 4a, 5a, 4b, and 5b are the same.

FIG. 8 is a flowchart showing the operation of only the server 11 in the above sequence and corresponds to FIG. 3 in the first embodiment. As described above, the present embodiment is different from the first embodiment only in the information preparation method (procedures 1 to 4) required for the determination in the server 11 in the procedure 5, and in the procedure 5. It is needless to say that the optimum distributed processing method considering the status of the processing load in the system is dynamically performed, as in the first embodiment, because the determination contents and the subsequent flow depending on the result are the same.

Further, the client 1 does not calculate the effective processing speed Rc every time a processing request message is issued, but only when inquired by the server.
The load on the client 1 is reduced. Furthermore, server 1
Since 1 can acquire the effective processing speed Rc of the client 1 at a timing convenient for itself, it is possible to make the determination in step 5 at a time close to the time when the specific task is actually processed in step 6a or 7b. Therefore, a more accurate processing method is determined even in a load condition that changes every moment.

Although the distributed processing system according to the present invention has been described above based on the embodiments, it goes without saying that the present invention is not limited to these embodiments. That is, (1) In the above embodiment, the effective processing speed of each terminal device is expressed by Expression 1 and Expression 2, assuming that the processing capacity of the processor is evenly distributed to each task to be processed. However, the effective processing speed is not limited to those expressed by these equations. For example, in the case of a terminal device in which weighted scheduling with priority is performed, the formula takes into account the weighting. (2) The determination of the processing device in the present distributed processing method is performed by the server 11, but is not limited to this device, and for example, the system is assumed to be performed by the client 1 or the third terminal device. It is also possible.

[0050]

As is apparent from the above description, according to the distributed processing method and system of the present invention, when a client makes a request to use the processing result of a specific task provided in the server, first, When it is determined that the server or the client should process the specific task in order to reduce the turnaround time of the specific task (first step), and then the server determines that the specific task should be processed. The result is reported to the client after the specific task is processed by the server, while the specific task is moved from the server to the client after the client is determined to process the specific task. The task is processed by the client (second step).

As a result, each time a processing request is issued from the client, a processing method for reducing the turnaround time required for the processing is performed, so that the processing is fixedly performed by either the server or the client. As compared with the conventional method, there is an effect that an efficient distributed processing method and system that can cope with changes in the system load situation are realized.

Here, as a method of deciding which of the server or the client should process, first, the effective processing speeds of the server and the client at that time are detected, and then the server determines based on the effective processing speed. Determine a first time required to process a specific task and report the result to the client, and a second time required to move the specific task from the server to the client and the client processes the specific task. It can be performed by comparing.

The effective processing speeds Rs and Rc of the server and the client detect the respective numbers Ns and Nc of tasks being processed in parallel, and the maximum processing speeds Rsmax and Rsmax of the respective tasks are detected.
Rcmax is used to determine by the equations Rs = Rsmax / (Ns + 1) and Rc = Rcmax / (Nc + 1), and the first time T1 and the second time T2 are Mtask, the work amount associated with the processing of a specific task, and Assuming that the communication speed of the road is Rt, the communication volume associated with the report is Mrep, and the communication volume associated with the task movement is Mmig, the equation T1 = Mtask / Rs + Mrep /
It can also be calculated by Rt and T2 = Mmig / Rt + Mtask / Rc.

As a result, a flexible and highly accurate distributed processing method that takes into consideration the processing capabilities of both devices, which changes moment by moment, becomes possible. Further, in the first step, the client may calculate its own effective processing speed at the present time, notify the server of the information indicating the effective processing speed at the same time as the request, and the server may make the determination.

As a result, since the processing capability of the client is notified to the server at the same time as the processing request from the client, the increase in communication traffic for this distributed processing can be minimized. Further, in the first step, the server that has received the request may inquire of the client, the client may calculate its effective processing speed at the present time and reply, and the server may make the determination.

As a result, the server can know the processing capacity of the client at a timing convenient for itself, so that there is an effect that the apparatus to be processed can be more accurately determined.

[Brief description of drawings]

FIG. 1 is a block diagram showing a hardware configuration of a distributed processing system according to first and second embodiments of the present invention.

FIG. 2 is a sequence diagram showing an exchange of communication between a client 1 and a server 11 in the distributed processing method according to the first exemplary embodiment.

FIG. 3 is a flowchart showing an operation of only the server 11 in the sequence of FIG.

FIG. 4 is a flowchart showing an operation of only the client 1 in the sequence of FIG.

FIG. 5 is a diagram showing a time chart in the first load situation of the first embodiment in comparison with a conventional fixed RPC method.

FIG. 6 is a diagram showing a time chart in the second load situation of the first embodiment in comparison with a conventional fixed JAVA system.

FIG. 7 is a sequence diagram showing an exchange of communication between a client 1 and a server 11 in the distributed processing method of the second exemplary embodiment.

8 is a flowchart showing the operation of only the server 11 in the sequence of FIG.

FIG. 9 (a) is an explanatory diagram showing a communication exchange between a client and a server in a conventional RPC system, and FIG. 9 (b) is a sequence diagram thereof.

FIG. 10 (a) is an explanatory diagram showing a communication exchange between a client and a server in a conventional JAVA system, and FIG. 10 (b) is a sequence diagram thereof.

FIG. 11 is a model diagram of a system configuration for evaluating turnaround time in each distributed processing method.

FIG. 12 is a time chart of a conventional example in which the turnaround time in the JAVA method is shorter than that in the RPC method.

FIG. 13 is a time chart of a conventional example in which the turnaround time is shorter in the RPC method than in the JAVA method.

[Explanation of symbols]

 1 client 2 processor 3 memory 4 communication unit 11 server 12 processor 13 memory 14 communication unit 20 communication path

Claims (7)

[Claims] 1. A specific task in which one terminal device (client) is provided in another one terminal device (server) in a network in which a plurality of terminal devices capable of concurrently processing a plurality of tasks are connected to a communication path. In the distributed processing method in the case of wanting to use the processing result, when the client issues a request to use the result, either the server or the client is required to reduce the turnaround time related to the specific task. The first step of determining whether to process the particular task, and if the server determines to process, the server processes the particular task and reports the result to the client, while the client Server decides that it should handle it, the server moves the specific task to the client and Distributed processing method bets is characterized by comprising a second step of processing the specific tasks. 2. The first step is performed by the server, the determining step of determining an effective processing speed of the server and the client at the present time, and the server performing the specific task based on the determined effective processing speed. And a second time required when the specific task is moved from the server to the client and the client processes the specific task. The distributed processing method according to claim 1, further comprising a determination step of making the determination by obtaining and comparing. 3. In the determining step, the effective processing speeds Rs, Rc of the server and the client detect the respective numbers Ns, Nc of tasks being processed in parallel, and the respective maximum processing speeds Rsmax, Rcmax are determined. make use of,
It is determined by the formulas Rs = Rsmax / (Ns + 1) and Rc = Rcmax / (Nc + 1), and in the determining step, the first time T1 and the second time T2 are the workloads involved in the processing of the specific task. Mtask, the communication speed of the communication path is Rt, the communication volume associated with the report is Mrep, and the communication volume associated with the task movement is Mmig.
Then, the distributed processing method according to claim 2, wherein the calculation is performed by the formulas T1 = Mtask / Rs + Mrep / Rt and T2 = Mmig / Rt + Mtask / Rc. 4. In the determination of the effective processing speed of the client in the determining step, the client calculates its own effective processing speed at the present time and notifies the server of the information indicating the effective processing speed at the same time as the request. The distributed processing method according to claim 2, wherein the distributed processing method is performed by the server that has received the notification. 5. The determination of the effective processing speed of the client in the detecting step is performed by the server that receives the request, inquires of the client, the client calculates and returns its own effective processing speed at the present time, and The distributed processing method according to claim 2, wherein the distributed processing method is performed by the server that receives a reply. 6. A distributed processing system in which a server having a specific task and a client that uses the processing result of the specific task are connected via a communication path, and the client wants to use the result to the server. Means for making a request to that effect, first calculating means for calculating its own effective processing speed at the present time and notifying the server, and processing the specific task if the specific task is moved from the server The server itself, based on the effective processing speeds calculated by the client and the server when the request is received, the server itself The first time it takes for the specific task to process the specific task and report the result to the client and the specific task for the client. A comparing means for moving to the ant and comparing with a second time required when the client processes the specific task; and when the first time is short, the server itself processes the specific task. A means for reporting the result to the client, and moving the specific task to the client when the second time is short, the distributed processing system. 7. The first calculation means, first storage means for storing in advance a maximum processing speed Rcmax of the client, first detection means for detecting the number Nc of tasks currently being processed in parallel, No. 1 first speed calculating means for calculating the effective processing speed Rc from the maximum processing speed Rcmax read from the storage means and the number of tasks Nc detected by the first detecting means by the formula Rc = Rcmax / (Nc + 1) The second calculation means includes a second storage means that stores in advance a maximum processing speed Rsmax of the server, a second detection means that detects the number Ns of tasks currently being processed in parallel, and the storage A second speed for calculating the effective processing speed Rs from the maximum processing speed Rsmax read from the means and the number of tasks Nc detected by the detecting means by the equation Rs = Rsmax / (Ns + 1). The comparison means stores in advance the work amount Mtask associated with the processing of the specific task, the communication rate Rt of the communication path, the communication amount Mrep associated with the report, and the communication amount Mmig associated with the task movement. And the effective processing speeds Rc and Rs calculated by the first speed calculation means and the second speed calculation means when the request is received. Then, the first time T1 and the second time T2 are expressed by the formula T1 = Mtask
/ Rs + Mrep / Rt and T2 = Mmig / Rt + Mtask / Rc, and a means for comparing the first time T1 and the second time T2 calculated by the time calculating means. The distributed processing system according to claim 6.