JP2017182221A - Automatic load distribution information processing system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide means for realizing a high efficiency of the entire data processing performed by a plurality of data processors in a data processing system of a distributed processing system eliminating centralization of data processing.SOLUTION: A data processor 420 divides data processing delegated from a data processing delegation source device 401 into parent data processing F0, child data processing F1, and child data processing F2. The data processor 420 determines that one of data processors which can receive a delegation of the child data processing F1 and can use position dependent data D3 used in the child data processing F1 first that can finish the data processing at the minimum cost is a data processor to which the child data processing F1 is to be delegated. The data processor 420 delegates the child data processing F1 to the data processor 421. Recursively conducting similar delegations efficiently allocates the data processing to data processors.SELECTED DRAWING: Figure 4

Description

  The present invention relates to a technology for distributed processing of a large amount of data using a plurality of data processing devices, and more specifically to a technology for easily and efficiently determining the sharing of processing in the distributed processing.

  Recently, communication networks represented by the Internet have become widespread. A number of sensors are connected to the communication network. Many data collected by these sensors are collected and analyzed, and the results are utilized.

  When collecting a large number of data from a large number of sensors, the sensor normally does not process the data generated by measurement or the like, and the data processing device (hereinafter referred to as “central data processing device”) from which data is collected ). The central data processor centrally processes the unprocessed data transmitted from the sensor.

  As described above, the method in which the central data processing apparatus processes data intensively (hereinafter referred to as “centralized processing method”) has an advantage that the configuration of a large number of sensors can be simplified. On the other hand, the centralized processing method has a drawback that the data processing load is concentrated on the central data processing apparatus. In addition, the centralized processing method has a drawback in that when the data generated by the sensor is data having a large amount of data such as data representing an image (still image or moving image), communication traffic is increased and concentrated. .

  As a method for solving the above-described disadvantage of the centralized processing method, a method for distributing data processing to a plurality of data processing devices (hereinafter referred to as “distributed processing method”) has been put into practical use. FIG. 7 is a diagram schematically showing a configuration of a data processing apparatus in a widely adopted distributed processing method. In a widely used distributed processing system, one main data processing device 201 divides data processing and assigns data processing divided into a plurality of slave data processing devices (slave data processing devices 202 to 205 in FIG. 7). I do. In this distributed processing method, data processing is executed by each of a plurality of slave data processing devices (slave data processing devices 202 to 205 in FIG. 7), so that load concentration of data processing does not occur.

  In the distributed processing system having the configuration shown in FIG. 7, the main data processing device 201 centrally manages the allocation of data processing. That is, the main data processing device 201 constantly monitors the contents of data processing, the data to be collected, the position of the sub data processing device in the communication network, the operating status of the sub data processing device (data processing load, etc.), etc. Change the allocation of data processing every time. Hereinafter, the distributed processing method having the configuration shown in FIG. 7 is referred to as a “centrally managed distributed processing method”. The centrally managed distributed processing method has an advantage that it is easy to optimize data processing performed by a plurality of slave data processing devices.

  On the other hand, in the centrally managed distributed processing method, the contents of data processing, the data to be collected, the position of the slave data processing device in the communication network, the operating status of the slave data processing device (data processing load, etc.), etc. frequently change. In this case, there is a disadvantage that the data processing accompanying the data processing allocation becomes complicated, and the main data processing device 201 is required to have high data processing capability. That is, if the number of data locations, the number of subordinate data processing devices, etc. increase, and the data processing load required for data processing allocation management exceeds the data processing capability of the main data processing device 201, the entire system will malfunction. . In addition, when a problem occurs in data communication between the main data processing device 201 and the communication network, the entire system also malfunctions.

  For example, Patent Literature 1 proposes a method for reducing the above-mentioned drawback of the centrally managed distributed processing method, that is, a method for reducing the load of data processing allocation management in the main data processing apparatus 201. Patent Document 1 does not include a main data processing device that manages allocation of data processing, but a data processing request source data processing device broadcasts a data processing request to surrounding data processing devices in a communication network. A method for determining a data processing apparatus that shares data processing has been proposed.

JP-A-10-74146

  According to the method described in Patent Document 1, the sharing of data processing is easily determined. However, in the method described in Patent Document 1, the efficiency of the entire data processing cannot be achieved. That is, in the method described in Patent Document 1, a data processing device with a small cost (time, data traffic, processing load on the data processing device, data communication fee, etc.) required for data processing does not always share the data processing. . In the present application, the cost required for data processing means the time required for data processing or the amount of resources occupied. For example, the communication bandwidth of the communication network and the data processing capability of the data processing device are examples of resources required for data processing, and the data traffic required for data processing, the processing load of the data processing device are examples of costs required for data processing. .

  SUMMARY OF THE INVENTION In view of the above circumstances, an object of the present invention is to provide means for realizing the efficiency of the entire data processing performed by a plurality of data processing devices in a distributed processing type data processing system that eliminates the concentration of data processing. And

  The present invention relates to each of a plurality of data processing devices including its own device connected to a communication network, the type of data processing that can be accepted by the data processing device, the type of data that can be used by the data processing device, Storage means for storing performance data indicating the cost required for the type of data processing for each type of data processing that can be accepted by the data processing apparatus; and processing means capable of executing one or more types of data processing; Receiving means for receiving delegation data indicating delegation of data processing from a data processing apparatus that is a delegation source of data processing; and dividing means for dividing data processing delegated by the delegation data received by the receiving means into a plurality of data processes; For each of the plurality of data processes divided by the dividing unit, the data processing is performed on the basis of the performance data. Deciding means for deciding whether to perform data processing on the data processing device other than its own device and deciding to delegate the data processing to another data processing device And transmission means for transmitting delegation data indicating delegation of the data processing to the data processing apparatus of the delegation destination decided by the decision means with respect to data processing decided to be delegated to a data processing apparatus other than the own apparatus The processing means executes data processing determined to be executed by the determining means in its own device, and the receiving means transmits from the delegation destination data processing apparatus as a response to the delegation data transmitted by the transmitting means. Receiving the result data indicating the result of the received data processing, and the sending means sends a response as a response to the delegation data received by the receiving means. A data processing apparatus that transmits at least one of result data indicating a result of data processing executed by the processing means and result data received by the receiving means from the data processing apparatus that is the delegation destination to the data processing apparatus that is the delegation source. Proposed as an aspect.

  Further, in the first aspect, based on the cost required for the data processing by the processing means, the performance data relating to the own device stored in the storage means is changed, and the transmission means performs data processing of the delegation destination. Based on the cost required by at least one of the communication network and the plurality of data processing devices until the receiving unit receives the result data as a response to the delegation data after transmitting the delegation data to the device, the data is stored in the storage unit Proposed as a second aspect is a configuration comprising changing means for changing the performance data related to the data processing apparatus of the delegation destination.

  Further, in the first or second aspect, the determination unit is not sure whether the certain type of data processing is acceptable based on the performance data regarding the certain type of data processing. If there is a data processing device that determines the presence or absence of a data processing device and is uncertain whether or not the certain type of data processing is acceptable, whether or not the certain type of data processing is acceptable A configuration in which an unknown data processing device is determined as a data processing device to which a certain type of data processing is delegated is proposed as a third aspect.

  Further, in any one of the first to third aspects, the receiving unit receives data from each of the data processing devices capable of data communication via the communication network via the communication network. A configuration in which the performance data relating to each of the data processing apparatuses capable of communication is received and the storage unit stores the performance data received by the reception unit is proposed as a fourth aspect.

  Further, the present invention relates to each of the plurality of data processing devices, each of which includes the own device, each of the plurality of data processing devices connected to the communication network, and the types of data processing that can be accepted by the data processing device, Storing performance data indicating the types of data that can be used by the data processing device and the cost of data processing of that type for each type of data processing that can be accepted by the data processing device; and the plurality of data Any one of the processing devices receives the delegation data indicating the delegation of data processing from the delegation source device, and the data processing device that has received the delegation data uses the delegation data Dividing the delegated data processing into a plurality of data processing, and the data processing apparatus receiving the delegated data, For each of the plurality of data processes divided in the dividing step, the data processing is executed by the own apparatus or delegated to a data processing apparatus other than the own apparatus based on the performance data stored in the storing step. If the data processing is determined to be delegated to another data processing device, the data processing device to which the data processing is delegated is determined, and the data determined to be executed by the own device in the determining step If there is a process, the step of executing the data processing, and the delegation that determines the delegation data indicating the delegation of the data processing if there is a data process determined to be delegated to a data processing apparatus other than the own apparatus in the determining step Transmitting to the previous data processing device and transmitting the delegation data; As a response to the delegation data received in the step of receiving result data indicating the result of the data processing transmitted from the data processing device of the delegation destination as a response to the delegation data transmitted in the step of receiving the delegation data, Transmitting at least one of the result data indicating the result of the data processing executed in the executing step and the result data received from the delegation destination data processing device in the step of receiving the result data to the delegation source device; A data processing method provided is proposed as a fifth aspect.

  The present invention also relates to a computer including a communication unit that performs data communication with another device via a communication network, the device itself and each of a plurality of data processing devices connected to the communication network. Stores performance data indicating the type of data processing that can be accepted by the device, the type of data that can be used by the device, and the cost required for the type of data processing for each type of data processing that the device can accept In processing, processing for receiving delegation data indicating delegation of data processing from a device that is a delegation source of data processing, processing for dividing data processing delegated by the delegation data into a plurality of data processing, and in the division processing For each of the plurality of divided data processes, the data process is executed by the own device based on the performance data. Determining whether to delegate to one of the plurality of data processing devices, and when determining to delegate the data processing to one of the plurality of data processing devices, determines a data processing device to which the data processing is delegated And, if there is a data process that is determined to be executed by the own apparatus in the process to be determined, the process to execute the data process is determined to be delegated to any of the plurality of data processing apparatuses in the process to be determined If there is data processing, as a response to the delegation data transmitted in the process of transmitting the delegation data, the process of transmitting delegation data indicating the delegation of the data process to the data processing apparatus of the delegation destination determined in the determination process A process of receiving result data indicating a result of data processing transmitted from the data processing apparatus of the delegation destination; As a response to the delegation data received in the process of receiving data, the result data indicating the result of the data process executed in the process to be executed and the result data received from the delegation destination data processing apparatus in the process of receiving the result data A program for causing at least one of the processes to be transmitted to the delegation source device to be executed is proposed as a sixth aspect.

  The present invention also relates to a computer including a communication unit that performs data communication with another device via a communication network, the device itself and each of a plurality of data processing devices connected to the communication network. Stores performance data indicating the type of data processing that can be accepted by the device, the type of data that can be used by the device, and the cost required for the type of data processing for each type of data processing that the device can accept In processing, processing for receiving delegation data indicating delegation of data processing from a device that is a delegation source of data processing, processing for dividing data processing delegated by the delegation data into a plurality of data processing, and in the division processing For each of the plurality of divided data processes, the data process is executed by the own device based on the performance data. Determining whether to delegate to one of the plurality of data processing devices, and when determining to delegate the data processing to one of the plurality of data processing devices, determines a data processing device to which the data processing is delegated And, if there is a data process that is determined to be executed by the own apparatus in the process to be determined, the process to execute the data process is determined to be delegated to any of the plurality of data processing apparatuses in the process to be determined If there is data processing, as a response to the delegation data transmitted in the process of transmitting the delegation data, the process of transmitting delegation data indicating the delegation of the data process to the data processing apparatus of the delegation destination determined in the determination process A process of receiving result data indicating a result of data processing transmitted from the data processing apparatus of the delegation destination; As a response to the delegation data received in the process of receiving data, the result data indicating the result of the data process executed in the process to be executed and the result data received from the delegation destination data processing apparatus in the process of receiving the result data A seventh aspect of the present invention proposes a recording medium in which a program for executing at least one of the processes for transmitting to the delegation source apparatus is continuously read by the computer.

  According to the present invention, the delegated data processing is decomposed into a plurality of data processing, and is shared and executed by each of the plurality of data processing devices. At that time, the data processing device delegated to the data processing can use the data required for the delegated data processing, and the data processing is estimated to be able to execute the delegated type of data processing at a lower cost. All or part of the data processing delegated to the device can be delegated. As a result, the data processing is distributed and executed by a plurality of data processing devices without requiring a data processing device for centrally managing the sharing of data processing, and the efficiency of the entire data processing is improved.

It is a figure showing an example of composition of a data processing system concerning one embodiment. It is the figure which showed the function structure of the data processing apparatus which comprises the data processing system concerning one Embodiment. It is the figure which illustrated the composition of the performance data used in the data processor concerning one embodiment. It is a figure for demonstrating the method for determining the data processing apparatus which the data processing apparatus concerning one Embodiment should perform a data process. It is a figure for demonstrating the method for determining the data processing apparatus which the data processing apparatus concerning one Embodiment should perform a data process. It is the figure which illustrated the processing flow for the data processing apparatus concerning one Embodiment to determine the data processing apparatus which should perform a data process. It is the figure which illustrated the processing flow for the data processing apparatus concerning one Embodiment to determine the data processing apparatus which should perform a data process. It is the figure which showed typically the structure of the data processing apparatus in the distributed processing system concerning a prior art.

[Embodiment]
Below, the data processing system 1 which concerns on one Embodiment of this invention is demonstrated. The data processing system 1 includes a plurality of data processing devices connected to a communication network, and any of the plurality of data processing devices is limited in data acquired from a delegation source device (hereinafter, referred to as “data acquisition system”). When the data processing using “position-dependent data” is delegated, the delegated data processing is executed by being shared by a plurality of data processing devices. Each of the plurality of data processing devices included in the data processing system 1 autonomously determines sharing of data processing. Therefore, the data processing system 1 does not have a mechanism for centrally managing the sharing of data processing.

  FIG. 1 is a diagram showing an example of the configuration of the data processing system 1. The data processing system 1 illustrated in FIG. 1 includes a total of six data processing devices 110, 120, 130, 140, 150, and 160. The number of data processing devices constituting the data processing system 1 is not limited. In addition, the data processing devices constituting the data processing system 1 can be detached from the data processing system 1 at an arbitrary timing. In addition, a new data processing apparatus can participate in the data processing system 1 at an arbitrary timing. Therefore, the number of data processing devices constituting the data processing system 1 usually changes in a fluid manner.

  Any of a plurality of data processing devices constituting the data processing system 1 receives data processing delegation using position-dependent data from a data processing delegation source device (for example, a terminal device used by an end user). The data processing apparatus that has received the delegation divides the delegated data processing into a plurality of data processes, and regarding each of the divided plurality of data processes, whether the data processing should be processed by the own apparatus, or other data processing Based on the performance of the data processing apparatus other than the self apparatus and the data processing apparatus, it is determined whether the data should be delegated to the apparatus and which data processing apparatus should be delegated to another data processing apparatus. However, in this application, the performance of the data processing device is the type of data processing that can be delegated, an index that indicates the size of the cost according to each type of data processing that can be delegated, and available Type of position-dependent data.

  The data processing apparatus that has received the data processing delegation executes the data processing determined to be processed by the own apparatus by the own apparatus. In addition, the data processing device that has received the data processing delegation delegates the data processing determined to be delegated to a data processing device other than its own device to the data processing device that has been determined as the delegation destination. When the data processing is delegated, the data processing device that has delegated receives the result data indicating the result of the delegated data processing from the data processing device that has been delegated. The data processing apparatus that has performed the delegation transmits at least one of the result data indicating the result of the data processing performed by the self apparatus and the result data received from the delegation destination data processing apparatus to the delegation source apparatus.

  The above-described processing performed by the data processing device that has been delegated data processing from the delegation source device is similarly executed in the data processing device that has been delegated data processing by the data processing device. When the data processing system 1 receives data processing delegation from a delegation source device or a data processing device other than the own device, the data processing system 1 divides the delegated data processing and executes the divided data processing in each own device or the like. The delegation to the data processing apparatus and the transmission of the result data indicating the result of the delegated data processing to the delegation source apparatus are performed recursively. As a result, the data processing is shared among the plurality of data processing devices so that the data processing is efficiently executed as a whole without requiring a data processing device for managing the sharing of the data processing.

  Data indicating the performance of each data processing device (hereinafter referred to as “performance data”) used when determining a data processing device that performs each of a plurality of data processing divided by the data processing device that has been entrusted with data processing. This is data generated based on the cost required in the data processing executed in the past and the type of position-dependent data used in the data processing executed in the past, and is stored in each data processing device. Each time a new data process is executed or a new data process is delegated and the result is received, each data processing apparatus stores the performance data stored in its own apparatus based on the cost required for the data process. Update. Therefore, the data processing system 1 can flexibly cope with changes in the communication state in the communication network, increase / decrease in the number of data processing devices participating in the data processing system 1, occurrence of failures in the communication network and data processing devices, and the like. Data processing can be performed.

  Note that the processing performed by each of the data processing devices constituting the data processing system 1 mainly includes the division of the delegated data processing and the execution of each of the divided data processing in its own device or other processing as described above. This is a relatively simple process such as delegation to the data processing device, collection of result data indicating the result of the delegated data processing, and transmission to the delegation source device. Thus, one of the features of the data processing system 1 is that the data processing system 1 efficiently performs data processing as a whole by performing relatively simple processing in each of the data processing devices.

  Each of the data processing devices constituting the data processing system 1 is a general computer including a communication unit that performs data communication with other devices via a communication network. That is, each of the data processing devices constituting the data processing system 1 includes at least a memory for storing various data including program data indicating a program, and various data processing according to the program indicated by the program data stored in the memory. And a communication interface which is an interface for performing data communication between the processor and other devices via a communication network.

  Further, at least a part of the data processing device constituting the data processing system 1 includes an input interface that is an interface for receiving position-dependent data generated by the sensor from a sensor that generates position-dependent data. Note that in a data processing apparatus that receives position-dependent data generated by a sensor connected to a communication network, the communication interface also serves as an input interface.

  The operation of the data processing system 1 will be described with reference to FIG. In FIG. 1, arrows indicate the flow of data transmitted and received between data processing apparatuses via a communication network.

  In the case illustrated in FIG. 1, first, the data processing device 110 receives delegation data 114 indicating delegation of data processing from a delegation source device (not shown). As an example, the delegation data 114 indicates delegation of data processing represented by F1 (F2 (F4 (D1), F5 (D2)), F3 (F6 (D3))). The function F # () indicates the type of data processing identified by #, and the data D # indicates the position-dependent data of the type identified by #. The data processing to be delegated may include data processing using data in which the position to be acquired is not limited (for example, time data indicating the current time, etc., hereinafter referred to as “position-independent data”). . Hereinafter, data d # indicates position-independent data of the type identified by #.

  The data processing device 110 divides the data processing indicated by the delegation data 114 into data processing F1 (), data processing F2 (F4 (D1), F5 (D2)), and data processing F3 (F6 (D3)). To do. The data processing F2 (F4 (D1), F5 (D2)) and the data processing F3 (F6 (D3)) are data processing for generating variables of the data processing F1 ().

  Subsequently, the data processing device 110 follows the determination method (described later) based on the performance data stored in the data processing device 110, and the data processing F1 () is data processing to be executed by the own device (data processing device 110). The data processing F2 (F4 (D1), F5 (D2)) is determined to be data processing to be delegated to the data processing device 120, and the data processing F3 (F6 (D3)) is determined to be the data processing device 130. It is determined that the data processing should be delegated. In accordance with this determination, the data processing device 110 transmits the delegation data 116 indicating the delegation of the data processing F2 (F4 (D1), F5 (D2)) to the data processing device 120, and transmits the data to the data processing device 130. Delegation data 118 indicating the delegation of the process F3 (F6 (D3)) is transmitted.

  The data processing device 120 divides the data processing F2 (F4 (D1), F5 (D2)) indicated by the delegation data 116 into the data processing F2 (), the data processing F4 (D1), and the data processing F5 (D2). To do. Subsequently, the data processing device 120 follows the determination method (described later) based on the performance data stored in the data processing device 120, and the data processing F2 () is data processing to be executed by the own device (data processing device 120). The data processing F4 (D1) is determined to be data processing to be delegated to the data processing device 140, and the data processing F5 (D2) is determined to be data processing to be delegated to the data processing device 150. . In accordance with this determination, the data processing device 120 transmits delegation data 126 indicating delegation of the data processing F4 (D1) to the data processing device 140, and delegates the data processing F5 (D2) to the data processing device 150. Delegation data 128 is sent.

  The data processing device 140 determines that the data process F4 (D1) indicated by the delegation data 126 is a process that cannot be further divided. Therefore, the data processing device 140 does not divide the data processing, and performs the data processing F4 (D1) according to a determination method (described later) based on performance data stored in the data processing device 140 (the data processing device). 140), the data processing to be executed is determined. Therefore, the data processing device 140 executes the data processing F4 (D1), and transmits the result data 127 indicating the data processing result F4 (D1) to the data processing device 120 as a response to the delegation data 126.

  The data processing device 150 determines that the data process F5 (D2) indicated by the delegation data 128 is a process that cannot be further divided. Therefore, the data processing device 150 does not divide the data processing, and this data processing F5 (D2) is performed according to a determination method (described later) based on performance data stored in the data processing device 150 (the data processing device). 150), the data processing to be executed is determined. Therefore, the data processing device 150 executes the data processing F5 (D2), and transmits the result data 129 indicating the data processing result F5 (D2) to the data processing device 120 as a response to the delegation data 128.

  The data processing device 120 uses the result F4 (D1) indicated by the result data 127 and the result F5 (D2) indicated by the result data 129 to perform the data processing F2 () (that is, the data processing F2 determined by the own device). (F4 (D1), F5 (D2))) is executed, and the result data 117 indicating the result of data processing F2 (F4 (D1), F5 (D2)) is returned as a response to the delegation data 116. Send to.

  On the other hand, when data processing device 130 receives delegation data 118 from data processing device 110, data processing F3 (F6 (D3)) indicated by delegation data 118 is changed to data processing F3 () and data processing F6 (D3). To divide. Subsequently, the data processing device 130 follows the determination method (described later) based on the performance data stored in the data processing device 130, and the data processing F3 () is data processing to be executed by the own device (data processing device 130). The data processing F6 (D3) is determined to be data processing to be delegated to the data processing device 160. In accordance with this determination, the data processing device 130 transmits delegation data 136 indicating delegation of the data processing F6 (D3) to the data processing device 160.

  The data processing device 160 determines that the data process F6 (D3) indicated by the delegation data 136 is a process that cannot be further divided. Therefore, the data processing device 160 does not divide the data processing, and performs this data processing F6 (D3) according to a determination method (described later) based on performance data stored in the data processing device 160 (the data processing device). 160), the data processing to be executed is determined. Therefore, the data processing device 160 executes the data processing F6 (D3), and transmits the result data 137 indicating the data processing result F6 (D3) to the data processing device 130 as a response to the delegation data 136.

  The data processing device 130 uses the result F6 (D3) indicated by the result data 137 to execute the data processing F3 () determined by the own device to be executed (that is, F3 (F6 (D3))). Result data 119 indicating the result F3 (F6 (D3)) is transmitted to the data processing device 110 as a response to the delegation data 118.

  The data processing apparatus 110 has decided to execute the data processing apparatus 110 using the result F2 (F4 (D1), F5 (D2)) indicated by the result data 117 and the result F3 (F6 (D3)) indicated by the result data 119. Data processing F1 () (that is, F1 (F2 (F4 (D1), F5 (D2)), F3 (F6 (D3)))) is executed, and the result of data processing F1 (F2 (F4 (D1), F5) (D2)) and result data 115 indicating F3 (F6 (D3))) are transmitted to a delegation source device (not shown) as a response to the delegation data 114.

  The above is the description of the operation of the data processing system 1 of the data processing system 1.

  FIG. 2 is a diagram showing a functional configuration of the data processing apparatus that constitutes the data processing system 1. That is, a general computer including a communication unit that performs data communication with another device via the communication network described above performs data processing according to the program according to the present embodiment stored in the memory. 2 functions as an apparatus including the components shown in FIG.

  Each of the data processing devices constituting the data processing system 1 receives storage data 301 for storing various data, delegation data from the delegation source device, and reception of result data from the delegation destination data processing device. Means 302 and dividing means 303 for dividing the data processing indicated by the delegation data into a plurality of data processes.

  Each of the data processing devices constituting the data processing system 1 automatically performs the data processing for each of the plurality of data processes divided by the dividing unit 303 based on performance data (described later) stored in the storage unit 301. Decide whether to execute on the device or to delegate to a data processing device other than its own device, and if it is determined to delegate the data processing to another data processing device, decide to determine the data processing device to which the data processing is delegated Means 304 are provided.

  Each of the data processing devices constituting the data processing system 1 includes processing means 305 that can execute one or more types of data processing and that executes data processing determined by the determining means 304 to be executed by itself. .

  In addition, each of the data processing devices constituting the data processing system 1 determines the delegation data indicating the delegation of the data processing with respect to the data processing determined by the determining unit 304 to be delegated to a data processing device other than the own device. The result data indicating the result of the data processing executed by the processing unit 305 and the data received by the receiving unit 302 as a response to the delegation data received by the receiving unit 302 A transmission unit 306 is provided for transmitting at least one of the result data received from the processing device to the data processing device as a delegation source.

  Further, each of the data processing devices constituting the data processing system 1 changes the performance data (described later) related to the own device stored in the storage unit 301 based on the cost required for the data processing by the processing unit 305, and A plurality of data processing devices constituting the communication network and the data processing system 1 until the receiving unit 302 receives the result data as a response to the delegation data after the transmission unit 306 transmits the delegation data to the delegation destination data processing device And a change unit 307 that changes performance data (described later) related to the data processing apparatus of the delegation destination stored in the storage unit 301 based on the cost required by at least one of the above.

Each of the data processing devices constituting the data processing system 1 includes position-independent data such as position-dependent data generated by a sensor connected to the data processing device and time data generated in the device itself. Acquisition means 308 is provided. Note that the acquisition unit 308 of some data processing apparatuses acquires position-independent data, but position-dependent data may not be acquired.

  FIG. 3 is a diagram illustrating a configuration of performance data stored in the storage unit 301 of each data processing device. The contents of the performance data stored in each of the plurality of data processing devices constituting the data processing system 1 are different from each other.

  The performance data is related to each of a plurality of data processing devices including its own device, the type of data processing that can be accepted by the data processing device, the type of data that can be used by the data processing device, and the data processing device It is data indicating the cost required for the type of data processing for each possible type of data processing.

  The performance data illustrated in FIG. 3 shows that the type of data processing shown in the “processing” column using the position-dependent data shown in the “data” column is past data processing devices shown in the “processing” column. The cost at the time of execution is shown in the “cost” column. However, the data in the “cost” column of the data record in which the “data” column is blank, the data processing of the type indicated in the “processing” column using position-independent data is the data indicated in the “device” column. The cost when executed in the past by the processing device is shown.

  The data shown in the “cost” column of the performance data includes, for example, the time required for the data processing executed by the own device, and the result data from the transmission of the delegated data for the data processing delegated to other data processing devices. Using the time required to receive the data, the processing load required for the data processing (for example, the average value of the processor occupancy), the amount of data transmitted / received during the data processing (ie, the traffic amount), etc. The data is calculated according to Information such as the processing load required for data processing and the amount of data transmitted / received during eta processing is added as metadata to the result data, for example, from the data processing device of the delegation destination to the data of the delegation source The processing device is notified.

  Hereinafter, a method of determining a data processing apparatus to be subjected to data processing based on performance data (FIG. 3) will be described with reference to FIGS. 4 and 5. FIG. 4 shows a state where the data processing apparatus 420 is about to determine each delegation destination of the data processing divided by the dividing unit 303. In the state of FIG. 4, the dividing unit 303 of the data processing device 420 performs the data processing indicated by the delegation data received from the delegation source device 401 as the parent data processing F0 (), the child data processing F1 (), and the child data processing F2. It is divided into ().

  The child data processing means data processing that generates a variable of the parent data processing as a result.

  The parent data process F0 () uses the result of the child data process F1 (), the position-dependent data D1, and the result of the child data process F2 ().

  The parent data processing F1 () uses the result of the child data processing F3 () and the result of the child data processing F4 (). Further, the parent data processing F3 () uses the position dependent data D3. In this case, the determination unit 304 of the data processing device 420 is a data processing device that can accept the data processing F1 () in determining the request destination of the child data processing F1 (), and can use the position-dependent data D3. Among data processing devices, the data processing device with the lowest cost is specified with reference to the performance data.

  The data processing device 420 determines the lowest-cost data processing device 421 identified as described above as the data processing device to which the child data processing F1 () is delegated. In this case, the data processing device 421 cannot directly acquire the position-dependent data D3 from the sensor, but delegates data processing using the position-dependent data D3 to the data processing device 422 or receives the position-dependent data from the data processing device 422. By receiving the data D3, the position-dependent data D3 can be used. Therefore, the data processing device 421 can be a delegation destination for data processing using the position-dependent data D3.

  Parent data processing F2 (), which is another child data processing divided by the dividing unit 303 of the data processing device 420, uses the position-dependent data D2 and the result of the child data processing F5 (). Further, the parent data processing F5 () uses the position dependent data D4. In this case, the determination unit 304 of the data processing device 420 is a data processing device that can accept the data processing F2 () in determining the request destination of the child data processing F2 () and can use the position-dependent data D2. Among data processing devices, the data processing device with the lowest cost is specified with reference to the performance data.

  In this embodiment, when the data processing to be delegated uses a plurality of position-dependent data, the position-dependent data used first among them (in this case, the position-dependent data D2 is preceded by the position-dependent data D4). As such, the position-dependent data D2) is taken into account in determining the data processing device to be delegated. However, for example, both of these position-dependent data may be considered in the determination of the data processing apparatus to be delegated.

  The data processing device 420 determines the lowest-cost data processing device 424 identified as described above as the data processing device to which data processing F2 () is delegated.

  In FIG. 5, in the data processing device 421, the dividing unit 303 divides the data processing delegated from the data processing device 420 into a parent data processing F <b> 1 (), a child data processing F <b> 3 (), and a child data processing F <b> 4 (). The determination means 304 is trying to determine the data processing device to which the child data processing F3 () and child data processing F4 () are delegated, and the data processing device 424 divides the data processing delegated from the data processing device 420. The means 303 is divided into parent data processing F2 () and child data processing F5 (), and the determination means 304 shows a state where the data processing device to which the child data processing F5 () is delegated is to be determined.

  The determination unit 304 of the data processing device 421 is a data processing device that can accept the data processing F3 () as a data processing device to which the child data processing F3 () is delegated, and can use the position-dependent data D3. Among the data processing devices, the data processing device with the lowest cost is identified with reference to the performance data. In this case, the data processing device 422 is determined as the delegation destination of the child data processing F3 ().

  In addition, the determination unit 304 of the data processing device 421 is a data processing device that can accept the data processing F4 () as a data processing device to which the child data processing F4 () is delegated, and has the lowest cost. The device is identified with reference to the performance data. In this case, the data processing device 423 is determined as the delegation destination of the child data processing F4 ().

  The determination unit 304 of the data processing device 424 is a data processing device that can accept the data processing F5 () as the data processing device to which the child data processing F5 () is delegated, and uses the position-dependent data D4. Among the possible data processing devices, the data processing device with the lowest cost is identified with reference to the performance data. In this case, the data processing device 425 is determined as the delegation destination of the child data processing F5 ().

  6A and 6B (hereinafter referred to as “FIG. 6”), a process for determining a data processing apparatus in which the data processing apparatus should execute data processing will be described.

  When the receiving unit 302 receives the delegation data from the delegation source device (step 601), the dividing unit 303 divides the data process delegated by the delegation data into a plurality of data processes (step S602). Subsequently, the determination unit 304 determines whether or not the divided parent data processing can be executed by the processing unit 305 of the own device based on the performance data (step S603). If the parent data processing cannot be executed by the own device (step 603; No), the transmission means 306 transmits information indicating that processing cannot be performed to the delegation source device (step 613), and the data processing device performs a series of processing. The process ends (step S617).

  When the parent data processing can be executed by the own device (step 603; Yes), the determination unit 304 determines whether all the child data processing has been processed (step 604). When all the child data processing has been completed (step S604; Yes), the processing unit 305 executes parent data processing using the result data of those child data processing, and generates result data indicating the result ( Step S614). Subsequently, the transmission unit 306 transmits the result data generated by the processing unit 305 to the requesting device (step 615). Subsequently, the changing unit 307 calculates the cost required for data processing and updates the performance data (step 616). Thereafter, the data processing apparatus ends a series of processes (step S617).

  When there is a child data process that has not been completed (step S604; No), the determination unit 304 selects one of the uncompleted child data processes as a target child data process (step S605), and selects the selected target data. When the top-level data processing of the child data processing and the selected child data processing use position-dependent data, the position-dependent data used first in the selected child data processing is specified from among the position-dependent data (step 606). ).

  Subsequently, the determining unit 304 can delegate the data processing specified in step 606 from the performance data, and if the position-dependent data is specified in step 606, the determining unit 304 can select a data processing apparatus that can use the position-dependent data. Then, the data is extracted as a candidate of the data processing apparatus that executes the target child data processing (step 607).

  When one or more data processing devices are extracted as candidates in step 607 (step S608; Yes), the determination unit 304 processes the child data for the candidate having the smallest cost indicated by the performance data from among these candidates. Is determined as a data processing apparatus to execute (step S609).

  When the data processing device determined in step S609 is the own device (step S610; Yes), the processing unit 305 executes the target child data processing (step 611) and stores the result data indicating the result in the storage unit 301. Remember me temporarily. Thereafter, the data processing apparatus returns the process to step S604.

  On the other hand, when the data processing apparatus determined in step S609 is not the own apparatus (step S610; No), the transmission unit 306 transmits the delegation data to the data processing apparatus determined as the delegation destination, and is transmitted as a response. The receiving means 302 receives the incoming result data (step S612). The receiving unit 302 temporarily stores the result data received from the delegation destination data processing apparatus in the storage unit 301. Thereafter, the data processing apparatus returns the process to step S604.

  If no data processing device candidate is extracted in step S607 (step S608; No), that is, if there is no record of delegating the target child data processing to any data processing device in the past, the determination unit 304 determines whether One non-selected data processing device is selected from data processing devices other than the device according to a predetermined rule (step 618). For example, it is unclear whether the position-dependent data used in the target child data processing can be used as a rule for the determination unit 304 to select the data processing device in step 618. It is unclear whether or not the same kind of data processing as the target child data processing can be accepted, but the rule that the data processing device that can accept the same type of data processing is preferentially selected. Rules that preferentially select data processing devices that can use position-dependent data used by child data processing, rules that preferentially select one with a low average cost, and rules that select randomly And so on.

  The transmitting unit 306 transmits the delegation data to the data processing device selected by the determining unit 304 in step 618 (step S619). When the result data indicating the result of the delegated data processing is transmitted as a response to the delegation data transmitted in step S619 (step 620; Yes), the reception unit 302 receives the result data and stores it in the storage unit 301. Remember me temporarily. Thereafter, the data processing apparatus returns the process to step S604.

  When the result data indicating the result of the delegated data processing has not been transmitted as a response to the delegated data transmitted in step S619 (step 620; No), the determination unit 304 selects all the data processing devices in step 618. It is determined whether or not (step 621). If there is an unselected data processing device in step 618 (step 621; No), the data processing device returns the processing to step 618.

  On the other hand, if all data processing devices have already been selected in step 618 (step 621; Yes), it means that no data processing device capable of executing the target child data processing has been found. Accordingly, the transmission unit 306 transmits a notification indicating that processing is impossible to the delegation source device (step 613), and the data processing device ends a series of processing (step S617).

  With the processing in steps 618 to 620 described above, a data processing device that can be delegated is specified for data processing that has not been delegated in the past, and the result is reflected in the performance data in step 616.

  According to the data processing system 1 described above, data processing is performed in a distributed manner by a plurality of data processing devices without requiring a data processing device that centrally manages the sharing of data processing, and the overall efficiency of data processing is improved. Figured.

[Modification]
The above-described embodiments can be variously modified within the scope of the technical idea of the present invention. Examples of these modifications are shown below.

(1) In the above-described embodiment, the performance data stored in the storage unit 301 by the data processing device indicates the performance related to the data processing device in which the data processing device can perform data communication via the communication network. The performance data may include data indicating the performance related to the data processing device in which the data processing device that stores the performance data cannot directly perform data communication.

  In this modification, the receiving unit 302 receives performance data relating to each data processing device capable of data communication from each data processing device capable of data communication, and the storage unit 301 performs the performance. Store the data.

  According to this modification, the determination unit 304 determines the data processing delegation destination on the basis of the performance of the data processing apparatus in which the own apparatus can directly perform data communication and the performance of the data processing apparatus in which the own apparatus cannot perform data communication directly. A data processing device can be determined. Therefore, the efficiency of the entire data processing can be improved as compared with the case of the above-described embodiment.

(2) In the above-described embodiment, if there is a record of delegating to a data processing device in the past regarding a certain data processing, the data processing device is repeatedly selected as a delegation destination of the data processing. Therefore, even if another data processing device can execute the same data processing at a lower cost, a data processing device having a record of executing the data processing in the past temporarily leaves the data processing system 1. If this situation does not occur, a data processing apparatus capable of executing data processing at a lower cost is not recognized.

  Therefore, if there is a data processing apparatus that has not attempted delegation in the past with respect to data processing to be delegated, the determining unit 304 preferentially selects the data processing apparatus as a data processing apparatus as a delegation destination, and performs data processing. A configuration may be employed in which information regarding the performance of the data processing apparatus is actively collected by attempting delegation.

  In this modification, the determination unit 304 determines the presence / absence of a data processing device that is unknown whether or not the target child data processing is acceptable based on the performance data regarding the target child data processing. When there is a data processing device whose unknown whether or not the target child data processing is acceptable, the processing unit 305 determines the data processing device as a data processing device to which the target child data processing is delegated.

(3) The data, processing flow, etc. used in the description of the above-described embodiments and modifications are examples, and the present invention is not limited to them. For example, the performance data may be configured as data indicating the performance of the data processing device based on the past performance in association with the time or time zone when the data processing is performed. In this case, for example, the determining unit 304 multiplies the data indicated by the performance data by a larger weight to the data at a time closer to the present time and performs a weighted average or the like, thereby obtaining a performance according to the current state of the communication network or the data processing device. The index shown can be obtained. The determining unit 304 extracts and uses, for example, values stored in association with the current time zone among the values indicated by the performance data, so that the communication network and data processing corresponding to the current material time zone are used. An index indicating performance according to the state of the apparatus can be obtained. As a result, a more appropriate delegation destination data processing apparatus is determined, and the data processing as a whole becomes efficient.

(4) A program for causing a computer to implement the data processing apparatus constituting the data processing system according to the above-described embodiment may be downloaded to a computer via a communication line such as the Internet and used. Moreover, this program may be continuously recorded on a recording medium, and read and used by the computer from the recording medium. Further, a recording medium on which the program is recorded may be distributed.

  The present invention is used in a service industry that provides data processing.

DESCRIPTION OF SYMBOLS 1 ... Data processing system 110 ... Data processing device 120 ... Data processing device 130 ... Data processing device 140 ... Data processing device 150 ... Data processing device 160 ... Data processing device 201 ... Main data processing device 202 ... Subordinate data processing apparatus 203 ... Subordinate data processing apparatus 204 ... Subordinate data processing apparatus 205 ... Subordinate data processing apparatus 301 ... Storage means 302 ... Receiving means 303 ... Division means 304 ... Determining means 305 ... Processing Means 306 ... Transmission means 307 ... Change means 308 ... Acquisition means 401 ... Delegation source device 420 ... Data processing device 421 ... Data processing device 422 ... Data processing device 423 ... Data processing device 424 ... Data processing device, 425 ... data processing device

Claims (7)

Regarding each of a plurality of data processing devices including its own device connected to a communication network, the type of data processing that can be accepted by the data processing device, the type of data that can be used by the data processing device, and the data processing device Storage means for storing performance data indicating the cost required for the type of data processing for each type of data processing that can be accepted by
Processing means capable of performing one or more types of data processing;
Receiving means for receiving delegation data indicating delegation of data processing from a data processing device that is a delegation source of data processing;
A dividing unit that divides data processing delegated by the delegation data received by the receiving unit into a plurality of data processes;
For each of the plurality of data processes divided by the dividing means, it is determined whether to execute the data processing on its own device or to be delegated to a data processing device other than its own device based on the performance data. A determination means for determining a data processing device to which the data processing is delegated;
Transmission means for transmitting delegation data indicating delegation of the data processing to the data processing apparatus of the delegation destination determined by the decision means in relation to the data processing decided by the decision means to be delegated to a data processing apparatus other than the own apparatus; Prepared,
The processing means executes data processing determined by the determining means to be executed by the own device,
The receiving means receives result data indicating a result of data processing transmitted from the data processing apparatus of the delegation destination as a response to the delegation data transmitted by the transmission means;
The transmission means, as a response to the delegation data received by the reception means, is at least one of result data indicating a result of data processing executed by the processing means and result data received by the reception means from a data processing apparatus as a delegation destination Is transmitted to the data processing device of the delegation source. Based on the cost required for the data processing by the processing means, the performance data relating to the own device stored in the storage means is changed, and the delegation after the transmission means transmits the delegation data to the data processing apparatus of the delegation destination Based on the cost required by at least one of the communication network and the plurality of data processing devices until the reception unit receives the result data as a response to the data, the data processing device of the delegation destination stored in the storage unit The data processing apparatus according to claim 1, further comprising: a changing unit that changes the performance data relating to the data. The determining means determines whether or not there is a data processing device that is uncertain whether or not the certain type of data processing is acceptable based on the performance data with respect to the certain type of data processing. If there is a data processing device in which it is unknown whether or not the data processing can be accepted, the data processing device in which it is unknown whether or not the certain type of data processing can be accepted The data processing device according to claim 1, wherein the data processing device is determined as a data processing device to be delegated. The receiving means receives, from each of data processing devices capable of data communication via the communication network, the performance data relating to each of the data processing devices capable of communication via the communication network. ,
The data processing device according to any one of claims 1 to 3, wherein the storage unit stores the performance data received by the reception unit. Each of the plurality of data processing devices connected to the communication network can use the type of data processing that can be accepted by the data processing device and the data processing device with respect to each of the plurality of data processing devices including the own device. Storing performance data indicating the type of data and the cost required for the type of data processing for each type of data processing that can be accepted by the data processing device;
Any one of the plurality of data processing devices receives delegation data indicating delegation of data processing from a device delegating data processing; and
The data processing apparatus that has received the delegation data divides the data process delegated by the delegation data into a plurality of data processes;
The data processing device that has received the delegation data executes the data processing on its own device based on the performance data stored in the storing step for each of the plurality of data processing divided in the dividing step. Determining whether or not to delegate to a data processing device other than its own device, and deciding to delegate the data processing to another data processing device, determining a data processing device to which the data processing is delegated; and
If there is data processing determined to be executed by the own device in the step of determining, the step of executing the data processing;
If there is data processing that is determined to be delegated to a data processing device other than its own device in the step of determining, the step of transmitting the delegation data indicating the delegation of the data processing to the delegation destination data processing device;
Receiving result data indicating a result of data processing transmitted from the data processing apparatus of the delegation destination as a response to the delegation data transmitted in the step of transmitting the delegation data;
As a response to the delegation data received in the step of receiving the delegation data, the result data indicating the result of the data processing executed in the step of executing and the result received from the delegation destination data processing device in the step of receiving the result data Transmitting at least one of the data to the delegation source device. In a computer provided with communication means for performing data communication with other devices via a communication network,
For each of a plurality of data processing devices connected to the own device and the communication network, the type of data processing that the device can accept, the type of data that the device can use, and the data processing that the device can accept A process for storing performance data indicating the cost required for the type of data processing for each of the types;
A process of receiving delegation data indicating delegation of data processing from a device that is a delegation of data processing;
A process of dividing data processing delegated by the delegation data into a plurality of data processes;
For each of a plurality of data processes divided in the process to be divided, it is determined based on the performance data whether the data processing is executed by the own apparatus or delegated to any of the plurality of data processing apparatuses. When it is determined to delegate processing to any of the plurality of data processing devices, processing to determine a data processing device to which the data processing is delegated;
If there is data processing determined to be executed by the own device in the determination processing, processing for executing the data processing;
If there is data processing determined to be delegated to any of the plurality of data processing devices in the determining process, the delegation data indicating the delegation of the data processing is transmitted to the data processing device of the delegation destination determined in the determining process Processing to
A process of receiving result data indicating a result of data processing transmitted from the data processing apparatus of the delegation destination as a response to the delegation data transmitted in the process of transmitting the delegation data;
As a response to the delegation data received in the process of receiving the delegation data, the result data indicating the result of the data process executed in the process to be executed and the result received from the delegation destination data processing apparatus in the process of receiving the result data A program for causing at least one of data to be transmitted to the delegation source device. In a computer provided with communication means for performing data communication with other devices via a communication network,
For each of a plurality of data processing devices connected to the own device and the communication network, the type of data processing that the device can accept, the type of data that the device can use, and the data processing that the device can accept A process for storing performance data indicating the cost required for the type of data processing for each of the types;
A process of receiving delegation data indicating delegation of data processing from a device that is a delegation of data processing;
A process of dividing data processing delegated by the delegation data into a plurality of data processes;
For each of a plurality of data processes divided in the process to be divided, it is determined based on the performance data whether the data processing is executed by the own apparatus or delegated to any of the plurality of data processing apparatuses. When it is determined to delegate processing to any of the plurality of data processing devices, processing to determine a data processing device to which the data processing is delegated;
If there is data processing determined to be executed by the own device in the determination processing, processing for executing the data processing;
If there is data processing determined to be delegated to any of the plurality of data processing devices in the determining process, the delegation data indicating the delegation of the data processing is transmitted to the data processing device of the delegation destination determined in the determining process Processing to
A process of receiving result data indicating a result of data processing transmitted from the data processing apparatus of the delegation destination as a response to the delegation data transmitted in the process of transmitting the delegation data;
As a response to the delegation data received in the process of receiving the delegation data, the result data indicating the result of the data process executed in the process to be executed and the result received from the delegation destination data processing apparatus in the process of receiving the result data A recording medium that stores the program for causing the computer to execute the process for transmitting at least one of the data to the delegation source device in a continuous manner.