JP7415351B2 - Control device and control program - Google Patents

Description

The present invention relates to a control device and a control program.

Patent Document 1 discloses a management system that manages a plurality of processing environments including at least a plurality of virtual machines capable of processing requests from clients and a load balancing device that transfers requests to the plurality of virtual machines, It has a management means for managing to which virtual machine a request is transferred from the distributed device, and the management means is configured to set the settings for the device that receives requests from the client from the load balancing device of the first processing environment. When the device is changed to a device in a second processing environment different from the first processing environment, a request is transferred from the load balancing device in the first processing environment to the virtual machine in the second processing environment. Disclosed is a management system characterized in that the management system is configured to manage virtual machines in the first processing environment and to prevent requests from being transferred from a load balancer in the first processing environment to a virtual machine in the first processing environment.

Patent Document 2 discloses that for an information processing device group including a plurality of information processing devices that execute processing in response to instructions from one or more information processing terminals corresponding to one or more accounts, the one or more receives a request from an information processing terminal to exclude at least one information processing device from the target of processing execution, and is installed in the information processing device to be excluded from among the plurality of information processing devices in accordance with the received request. The processing execution history is copied from the storage unit that stores the processing execution history to an information processing device different from the information processing device to be excluded, and after the copying of the processing execution history is completed, the information to be excluded is A scale-in processing program is disclosed that is characterized by causing a computer to execute processing that excludes a processing device from a processing execution target.

Japanese Patent Application Publication No. 2018-88114 Japanese Patent Application Publication No. 2018-73099

With the diversification of lifestyle patterns and working hours, information processing systems that accept requests from users day and night and provide the requested services are being introduced.

In such information processing systems, even if the configuration of the information processing system is changed for the purpose of improving functionality, for example, measures are taken to ensure that service provision does not stop during the period of configuration change.

For example, by having the configuration before the change and the configuration after the change coexist in the information processing system, and processing requests received after a predetermined time using the changed configuration, the service suspension period due to the configuration change can be improved. We are trying to prevent this from occurring.

However, when the configuration before the change and the configuration after the change coexist in the information processing system, even after the configuration is changed, the configuration before the change that is no longer used continues to consume resources of the information processing system. Therefore, the cost of maintaining the information processing system may be higher than before the configuration change.

The present invention reduces the cost of the information processing system due to the change in configuration compared to the case where the configuration before the change and the configuration after the change coexist when the configuration of the information processing system that executes requested processing is changed. An object of the present invention is to provide a control device and a control program that can suppress the above.

The control device according to the first aspect includes at least one processing entity that executes the received processing in each of a plurality of processing environments whose configuration is to be changed, which is built in an information processing system that receives the requested processing. is in operation, the information processing system is configured to aggregate the processing entities that execute processes that satisfy predetermined conditions regarding processing time into a specific processing environment, and then allow each of the processing entities to continue processing. It includes a control section for controlling.

In the control device according to the second aspect, in the control device according to the first aspect, the predetermined condition is set such that the processing time from the start to the end of the process exceeds a threshold value.

In the control device according to a third aspect, in the control device according to the second aspect, the control unit refers to history information in which processing times of executed processes executed so far and processing conditions of the executed processes are recorded. Then, the processing time of the requested process is estimated from the history information of the executed process corresponding to the requested process, and control is performed to instruct the information processing system to relocate the processing entity.

In the control device according to the third aspect, when a data size of data received in response to a request for the executed process is recorded as a processing condition in the history information, the control device according to the third aspect The unit performs control to estimate the processing time of the requested task from the processing time of the data size in the executed process corresponding to the requested process.

In the control device according to a fifth aspect, in the control device according to the third aspect or the fourth aspect, the control unit has a control unit that prioritizes a requested process in a processing order over other processes being executed in the processing entity. If the setting is low, control is performed to extend the estimated processing time according to the set priority.

In the control device according to a sixth aspect, in the control device according to any one of the third to fifth aspects, when the processing time estimated for the request before execution exceeds the threshold value, , controlling the information processing system to place the processing entity that executes the pre-execution request in the specific processing environment.

In the control device according to a seventh aspect, in the control device according to any one of the third to sixth aspects, the control unit controls the information processing system while the information processing system is in operation. control to instruct relocation of the processing entity multiple times.

In the control device according to an eighth aspect, in the control device according to any one of the first to seventh aspects, the control unit is configured to control the information processing system that is no longer used due to the termination of processing in the processing entity. The information processing system is controlled to release resources in sequence.

In the control device according to a ninth aspect, in the control device according to the eighth aspect, the control unit deletes the processing environment in order from the processing environment in which the processing entity that executes the process no longer exists, and finally deletes the processing environment. The information processing system is controlled to delete the specific processing environment.

The control program according to the tenth aspect is a program for causing a computer to function as a functional unit of the control device according to any one of the first to ninth aspects.

According to the first aspect and the tenth aspect, with a change in the configuration of an information processing system that executes a requested process, compared to a case where the configuration before the change and the configuration after the change coexist, This has the effect that the cost of the accompanying information processing system can be suppressed.

According to the second aspect, there is an effect that processing entities that execute processing whose processing time exceeds a threshold value can be integrated into a specific processing environment.

According to the third aspect, even if the processing time is unknown, it is possible to estimate in advance whether the processing time will exceed the threshold value.

According to the fourth aspect, there is an effect that the processing time can be estimated more accurately than when estimating the processing time only from the type of accepted processing.

According to the fifth aspect, there is an effect that the processing time can be estimated more accurately than when the processing time is not extended depending on the priority of the requested processing.

According to the sixth aspect, there is an effect that processing entities that execute processing whose processing time exceeds a threshold value can be aggregated in a specific processing environment before the requested processing is executed by the information processing system. .

According to the seventh aspect, there is an effect that processing entities that execute processing whose processing time exceeds a threshold value can be integrated into a specific processing environment according to the actual processing time.

According to the eighth aspect, there is an effect that the cost of the information processing system can be suppressed compared to a case where the resources used by the processing entity are not released even after the processing ends.

According to the ninth aspect, there is an effect that the cost of the information processing system can be suppressed compared to the case where a processing environment that does not include a processing entity exists.

1 is a diagram showing an example of a system configuration of a control system. FIG. 1 is a diagram illustrating a configuration example of an information processing system. FIG. 2 is a diagram showing an example of a functional configuration of a control device according to a first embodiment and a second embodiment. It is a figure showing an example of history information. FIG. 2 is a diagram showing an example of a main part configuration of an electrical system in a control device. 7 is a flowchart illustrating an example of the flow of relocation processing according to the first embodiment. FIG. 6 is a diagram illustrating an example of a change in the configuration of the information processing system when relocation processing is executed. 3 is a flowchart illustrating an example of the flow of resource release processing. 12 is a flowchart illustrating an example of the flow of node placement control processing according to the second embodiment. It is a figure showing an example of functional composition of a control device concerning a 3rd embodiment. It is a figure which shows an example of version information. 12 is a flowchart illustrating an example of the flow of node placement control processing according to the third embodiment.

The present embodiment will be described below with reference to the drawings. Note that components and processes having the same function are given the same reference numerals throughout the drawings, and redundant explanations are omitted.

<First embodiment>
FIG. 1 is a diagram showing an example of a system configuration of a control system 100. The control system 100 includes an input device 10, a control device 20, and an information processing system 30, each of which is connected via a communication line 2.

There are no restrictions on the communication protocol used in the communication line 2, and the communication line 2 may be a wired line or a wireless line, or may be a line in which a wired line and a wireless line are mixed. Furthermore, the communication line 2 may be a dedicated line or a public line that is shared with an unspecified number of users, such as the Internet.

The input device 10 is a device that generates a message for causing the information processing system 30 to execute a requested process, and is used by a user who uses a service provided by the information processing system 30. There are no restrictions on the content of the processing that the user requests from the information processing system 30, and the processing for the service provided by the information processing system 30 is requested. Hereinafter, a message for causing the information processing system 30 to execute a requested process will be referred to as a "request."

The input device 10 transmits the generated request to the control device 20.

The control device 20 is a device that has the role of a relay device that receives requests from the input device 10 and transmits the received requests to the information processing system 30 that executes the process specified by the request.

For convenience of explanation, FIG. 1 shows an example in which one information processing system 30 is connected to the communication line 2, but there is no restriction on the number of information processing systems 30 connected to the communication line 2, and the control device 2 shall relay the received request to the information processing system 30 specified by the request. Further, there is no restriction on the number of input devices 10 connected to the communication line 2, and the control device 20 receives requests from at least one input device 10 connected to the communication line 2.

Furthermore, the control device 20 controls the configuration of the information processing system 30 that executes the processing instructed by the request.

The information processing system 30 is a system that executes processing instructed by a request. There are no restrictions on the form in which the information processing system 30 is realized, and the devices configuring the information processing system 30 may be realized in an on-premises form installed in a building owned by a business operator that provides services using the information processing system 30. However, here, as an example, it is assumed that the information processing system 30 is realized using a cloud service provided by a cloud provider.

FIG. 2 is a diagram illustrating a configuration example when the information processing system 30 executes processing instructed by a request. The configuration of the information processing system 30 refers to an execution environment built in the information processing system 30 to execute the process instructed by the request, and specifically refers to the software configuration that performs the process.

When the information processing system 30 performs processing, first, a virtual server is generated using a cloud service. A virtual server is a processing environment that executes processing, and includes, for example, a CPU (Central Processing Unit) that executes processing, RAM (Random Access Memory), and nonvolatile storage represented by a hard disk. Like a device, the resources necessary to perform the requested processing are allocated.

Note that when executing processing, the cloud provider does not necessarily allocate one server that can be used exclusively. Depending on the contract with the service provider, a cloud provider may divide one server into multiple virtual servers and then provide the virtually divided servers to the service provider. Each of these virtually divided servers is called a "virtual server." Hereinafter, the virtual server will be referred to as "instance 34." Further, the requested processing will be referred to as a "task".

The cloud service usage fee that a business operator pays to a cloud provider is set based on at least one of the number of instances and usage time. That is, as the number of requested instances 34 increases and as the usage time of the instances 34 increases, the usage fee paid to the cloud provider increases.

In the example of FIG. 2, two instances 34, a first instance 34 and a second instance 34, exist in the information processing system 30, and a node 32 is arranged in each instance 34.

The node 32 is a processing entity that uses the resources of the instance 34 to execute a task corresponding to a request received from the input device 10 (hereinafter referred to as a "requested task") according to a program prepared in advance. It is. A node 32 is placed in any instance 34 and uses the resources of the placed instance 34.

Since the program executed on the node 32 differs depending on the type of requested task, for convenience of explanation, the node 32 that executes task A will be referred to as the A node 32, and the node 32 that executes task A will be referred to as task B, which is of a different type than task A. The node 32 that executes this will be referred to as the B node 32. In addition, if there is no need to distinguish between the types of tasks to be executed, it will simply be referred to as "node 32."

Using the above expression, FIG. 2 shows a first instance 34 in which two A nodes 32 and one B node 32 are arranged, and a second instance 34 in which one A node 32 and one B node 32 are arranged. 34 represents the state of an information processing system 30 having 34.

Note that there is no restriction on the number of instances 34 used in the information processing system 30, and there is no restriction on the number of nodes 32 arranged in one instance 34, as long as it is within the scope of the contract with the cloud provider. As described above, the configuration of the information processing system 30 is controlled by the control device 20.

FIG. 3 is a diagram showing an example of the functional configuration of the control device 20. As shown in FIG. As shown in FIG. 3, the control device 20 includes functional units such as a reception section 21, a control section 22, a history management section 23, an output section 24, and a history information DB (Database) 25.

The receiving unit 21 receives requests from the input device 10 and also receives the configuration state of the information processing system 30 that processes the task and the processing status of the task from the information processing system 30, and notifies the control unit 22 of the information processing status.

The configuration state of the information processing system 30 refers to the configuration actually constructed in the information processing system 30, and the information processing system 30 is configured to determine, for example, the number and types of instances 34, and the nodes 32 included in the instances 34. Information indicating the configuration built in the information processing system 30, such as type and number, is notified to the control device 20. The type of node 32 is defined by the type of task executed by the node 32. That is, if the node 32 executes task A, the type of node 32 called A node 32 is included in the configuration state of the information processing system 30. Note that the types of instances 34 will be explained later.

The processing status of a task is the processing status of a task executed by the node 32, and includes, for example, the execution status such as whether the task is being executed or not, and if the task has finished, the execution of the task has started. Includes the processing time required from start to finish. Here, as an example, the processing time of a task will be explained as the time required from when the information processing system 30 starts executing the task until it ends. The processing time may be the time required from the time the requested task is completed until the execution of the requested task is completed.

When the control unit 22 receives a request from the input device 10, it notifies the output unit 24 of the requested task and performs control to request the information processing system 30 to execute the task. Further, the control unit 22 changes the configuration of the information processing system 30 for the information processing system 30 that executes the task according to the configuration state of the information processing system 30 and the processing status of the task received from the reception unit 21. control. Specifically, when there is a task that satisfies a predetermined condition regarding processing time, the control unit 22 performs control such that the nodes 32 that execute the task are aggregated into a specific instance 34 . Note that the predetermined conditions regarding the processing time will be explained later.

Under the control of the control unit 22, the history management unit 23 generates history information 4 in which the content of the request received from the input device 10 and the processing status of the requested task are associated with each other, and manages the history information 4 in the history information DB 25. Further, under the control of the control unit 22, the history management unit 23 acquires the contents of the instructed request and the processing status of the task from the history information DB 25, and notifies the control unit 22 of the information.

FIG. 4 is a diagram showing an example of history information 4 managed by the history information DB 25. The history information 4 includes the reception date and time, request ID (identification), additional data, task ID, and processing time for each request received from the input device 10.

The reception date and time records the reception date and time of the request.

The request ID is an identifier used to uniquely manage requests within the control device 20, and in the case of the history information 4 shown in FIG. 4, a number that does not overlap with each request is assigned.

Additional data is data received from the input device 10 along with the request. For example, if the request is a request to order a product, an order sheet file containing the product name and quantity to be ordered is attached to the request as additional data. Furthermore, if the request is a translation request, a file containing the document to be translated is attached to the request as additional data. Therefore, the file name of the attached file is set as the additional data of the history information 4, for example. There are requests that do not require additional data, but in such cases, "-" is set in the additional data of the history information 4 to indicate that the request does not include additional data.

The task ID is an identifier that uniquely represents the type of task requested by the request, and an identifier representing the type of task is set for each task executed by the information processing system 30.

The processing time is the processing time required from the time when the information processing system 30 starts executing a task until it ends, as described above, and the processing time notified from the information processing system 30 is set. Note that since the information processing system 30 does not notify the processing time of a task whose execution has not been completed, the processing time of the history information 4 corresponding to such a task includes a "-" indicating that the task has not been completed. is set. In other words, a task for which a time is set in the processing time of the history information 4 indicates that the task has been executed, and a task for which "-" is set indicates that the task is currently being executed or is waiting to be executed in the information processing system 30. This indicates that the task is currently being completed. Hereinafter, tasks that have been requested to be executed by the information processing system 30 but have not yet been executed will be collectively referred to as "tasks in progress."

Note that the contents set in the history information 4 are not limited to the items of the history information 4 shown in FIG. For example, the history information 4 may include attributes representing the characteristics of the file attached to the request, such as the file size and the file creator, and instructions for the task, such as the task priority and desired completion deadline.

The additional data, the attributes representing the characteristics of the file, and the contents of instructions for the task are contents that influence the processing of the task in the information processing system 30, and are thus an example of the processing conditions according to the present embodiment.

The output unit 24 outputs the requested task to the designated information processing system 30 together with information that uniquely identifies the task, such as a request ID, under the control of the control unit 22 . Specifically, the output unit 24 outputs the task to the designated information processing system 30 via the communication line 2, but the method of outputting the task to the information processing system 30 is not limited to this. For example, the output unit 24 stores tasks in a storage area predetermined for each information processing system 30 of the control device 20, and the information processing system 30 reads the contents of the storage area allocated to its own system, thereby allowing information processing. Tasks may be exchanged with the system 30.

In the information processing system 30, a task received from the control device 20 is associated with a node 32 that executes the task, for example, through a request ID.

Next, an example of the main part configuration of the electrical system in the control device 20 will be described.

FIG. 5 is a diagram showing an example of the configuration of main parts of the electrical system in the control device 20. As shown in FIG. The control device 20 is configured using a computer 40, for example.

The computer 40 is an example of a processor that has the function of converting, calculating, and processing data based on a procedure, and the computer 40 and the CPU 41 that are responsible for each functional section related to the control device 20 have the functions shown in FIG. 3. A ROM (Read Only Memory) 42 that stores a control program that functions as a part, a RAM 43 that is used as a temporary work area for the CPU 41, a nonvolatile memory 44, and an input/output interface (I/O) 45 are provided.

A CPU 41, a ROM 42, a RAM 43, a nonvolatile memory 44, and an I/O 45 are connected to each other via a bus 46.

The non-volatile memory 44 is an example of a storage device that maintains stored information even if the power supplied to the non-volatile memory 44 is cut off. For example, a semiconductor memory is used, but a hard disk may also be used. The history information DB 25 that manages the history information 4 is constructed in the nonvolatile memory 44, for example.

For example, a communication unit 47, an input unit 48, and a display unit 49 are connected to the I/O 45.

The communication unit 47 is connected to the communication line 2 and implements a communication protocol for performing data communication with the input device 10 and the information processing system 30 connected to the communication line 2, and further with an external device (not shown) connected to the communication line 2. Equipped with

The input unit 48 is a device that receives instructions from an operator operating the control device 20 and notifies the CPU 41, and uses, for example, buttons, a touch panel, a keyboard, a mouse, and the like. When receiving instructions from an operator by voice, a microphone may be used as the input unit 48.

The display unit 49 is an example of a device that outputs information processed by the CPU 41 to an operator, and includes, for example, a liquid crystal display, an organic EL (Electro Luminescence) display, a projector that projects an image onto a screen, and the like.

The units connected to the I/O 45 are merely examples, and it goes without saying that units other than the units shown in FIG. 5 may be connected as necessary, such as an image forming unit that forms an image on a recording medium.

Next, the operation of the control device 20 that controls the configuration of the information processing system 30 that executes the requested task will be described in detail.

FIG. 6 is a flowchart showing an example of the flow of the relocation process executed by the CPU 41 of the control device 20. For example, the relocation process is periodically executed in a situation where a plurality of instances 34 are allocated in the information processing system 30 and at least one node 32 is operating in each instance 34, but the configuration of the information processing system 30 It may be performed periodically regardless of the It is assumed that the control device 20 knows the configuration of the information processing system 30 in operation, for example, based on the configuration status notified from the information processing system 30.

A control program that defines the relocation process is stored in advance in the ROM 42 of the control device 20, for example. The CPU 41 of the control device 20 reads the control program stored in the ROM 42 and executes the relocation process.

In step S<b>10 , the CPU 41 selects any one node 32 from among the nodes 32 that are currently being executed in the information processing system 30 . For convenience of explanation, the node 32 selected in step S10 will be referred to as the "selected node 32."

Since the processing of the currently running node 32 in the information processing system 30 has not yet been completed, "-" is set in the processing time of the history information 4 instead of a time. Therefore, selecting any one node 32 from among the nodes 32 that are currently being executed means that the CPU 41 selects any one from the history information 4 in which "-" is set for the processing time. This corresponds to selecting one piece of history information 4.

In step S20, the CPU 41 stores the executed history information 4 including the same task ID as the task executed by the selected node 32, that is, the task represented by the task ID of the history information 4 selected in step S10, in the history information DB 25. Determine whether it exists. The executed history information 4 is the history information 4 corresponding to a task that has been executed in the information processing system 30, and is the history information 4 in which a time is set as the processing time.

If execution history information 4 including the task ID of the task executed by the selected node 32 exists, the process moves to step S30.

In step S30, the CPU 41 estimates the processing time from the executed history information 4 including the task ID of the task to be executed by the selected node 32. Since tasks of the same type tend to have the same processing time, the processing time of the task executed by the selection node 32 is the same as the processing time of the executed history information 4 that includes the same task ID as the task. It is estimated that they will be the same.

Note that if the history information DB 25 includes a plurality of pieces of executed history information 4 including the task ID of the task executed by the selected node 32, the CPU 41 determines the processing time of any one of the pieces of history information 4. , may be the processing time of the task executed by the selected node 32. However, the method for estimating the processing time of a task when there is a plurality of pieces of executed history information 4 is not limited to this. For example, the processing time statistics in each executed history information 4 including the task ID of the task executed by the selected node 32 may be estimated as the processing time of the task executed by the selected node 32. For example, the average value, minimum value, or maximum value of each processing time is used as the statistical amount of the processing time.

In step S40, the CPU 41 determines whether the processing time estimated in step S30 exceeds a threshold value. The threshold value is a reference value for determining whether a task takes a longer processing time than other tasks executed by the information processing system 30 (hereinafter referred to as a "long-term task"), and This is an example of predetermined conditions. The threshold value is set by the operator of the control device 20 and is stored in advance in the nonvolatile memory 44, for example.

If the task executed by the selection node 32 is a long-term task, the instance 34 including the task must also be kept alive at least until the task is completed. As described above, the usage fee for the cloud service is set based on at least one of the number of instances 34 used in the information processing system 30 and the usage time. Therefore, when an instance 34 includes a long-term task, it is necessary to keep the instance 34 alive only for the long-term task even after the other tasks included in the instance 34 finish, until the long-term task finishes. be. In particular, when a long-term task is distributed over multiple instances 34, each of the multiple instances 34 will last beyond the time represented by the threshold, reducing the cost of maintaining the information processing system 30, i.e. , leading to an increase in usage fees for cloud services.

Therefore, the CPU 41 consolidates long-term tasks into one instance 34 and then rearranges the tasks so that the processing of each task executed in the information processing system 30 continues.

Therefore, if the processing time estimated in step S30 exceeds the threshold, the process moves to step S50, and in step S50, the CPU 41 determines whether the long-term instance 34 exists in the information processing system 30.

The long-term instance 34 indicates the type of instance 34, and is an instance 34 in which long-term tasks are aggregated and arranged. The long-term instance 34 is an example of a specific processing environment according to this embodiment.

Whether or not the long-term instance 34 exists in the information processing system 30 is determined based on the configuration state of the information processing system 30 notified from the information processing system 30. Then, if the long-term instance 34 does not exist, the process moves to step S60.

In step S60, the CPU 41 outputs a generation instruction instructing the information processing system 30 to newly generate a long-term instance 34, and the process proceeds to step S70. As a result, a long-term instance 34 is generated in the information processing system 30.

On the other hand, if it is determined in the determination process of step S50 that the long-term instance 34 exists in the information processing system 30, there is no need to newly generate the long-term instance 34, so the process of step S60 is not executed and the step The process moves to S70.

In step S70, the CPU 41 controls the communication unit 47 so that the selected node 32 that executes the long-term task is placed in the long-term instance 34. A placement instruction that instructs the processing system 30 to change the placement of the selected node 32 is output. This causes the selected node that executes the long-term task to be relocated to the long-term instance 34. After outputting the placement instruction, the process moves to step S80.

On the other hand, in the determination process of step S20, if it is determined that the executed history information 4 including the task ID of the task to be executed by the selected node 32 does not exist, the task to be executed by the selected node 32 is This is the type of task that is executed for the first time. That is, the history information 4 does not record the actual processing time when the same type of task as the task executed by the selected node 32 is executed. Therefore, it is difficult to use the threshold value to determine whether or not the task executed by the selected node 32 is a long-term task, so the process moves to step S80 without executing the processes of steps S30 to S70. Furthermore, if it is determined in the determination process of step S40 that the estimated processing time of the selected node 32 is less than or equal to the threshold, it is determined that the selected node 32 is not a long-term task, so the processes of steps S50 to S70 are executed. The process moves to step S80 without doing so. In these cases, the selected node 32 is not placed in the long-term instance 34 and continues to be placed in the instance 34 that includes the selected node 32 from the beginning.

In step S80, the CPU 41 determines whether there is an unselected node 32 that has not been selected in step S10 among the nodes 32 being executed. If there is an unselected node 32, the process moves to step S10, and any one node 32 is selected from among the unselected nodes 32 in execution.

By repeatedly executing steps S10 to S80 until there are no unselected running nodes 32, the nodes 32 that execute long-term tasks in the information processing system 30 are aggregated into the long-term instance 34.

As will be explained later with reference to FIG. 8, nodes 32 whose tasks have finished executing are sequentially deleted from the information processing system 30, and instances 34 that do not include the nodes 32 are also sequentially deleted from the information processing system 30. Therefore, since the number of instances 34 that survive beyond the threshold is one, the usage fee for the cloud service is reduced compared to the case where nodes 32 that execute long-term tasks are not aggregated into long-term instances 34. become.

On the other hand, if it is determined in the determination process in step S80 that all nodes 32 that are currently being executed have been selected, the relocation process shown in FIG. 6 is ended.

FIG. 7 is a diagram showing an example of a change in the configuration of the information processing system 30 when the control device 20 executes the relocation process shown in FIG.

Similar to the configuration example of the information processing system 30 shown in FIG. 2, FIG. 7 shows that before the control device 20 executes the relocation process shown in FIG. exists in the information processing system 30, two A nodes 32 and one B node 32 are arranged in the first instance, and one A node 32 and one B node 32 are arranged in the second instance 34. It represents the situation.

Assuming that the B node 32 is a node 32 that executes a long-term task, when the control device 20 executes the relocation process shown in FIG. Then, the B nodes 32 are aggregated into the third instance 34.

Note that in the relocation process shown in FIG. 6, the processing time of a task is estimated from the processing time of the executed history information 4, but the method for estimating the processing time of a task executed by the node 32 is not limited to this.

For example, if the types of files attached to requests as additional data are the same, the processing times of tasks corresponding to the requests tend to be similar. Therefore, even if it is determined that there is no executed history information 4 including the task ID of the task to be executed by the selected node 32 in the determination process of step S20 in FIG. The processing time of the task executed by the selected node 32 is estimated from the processing time of the executed history information 4 including additional data in which the same type of file as the file associated with is set. File types can be identified by the file extension.

In this case, the CPU 41 does not directly use the processing time of the executed history information 4 that includes additional data of the same file type as the processing time of the task executed by the selected node 32, but instead uses the processing time according to the data size of the file. The estimated processing time may be corrected. Specifically, it is assumed that the larger the data size of the file, the more data to be processed. The estimated processing time may be corrected to become longer as the proportion of the data size of the file attached to the task executed by the selection node 32 increases.

Correction of the processing time using the data size of such a file is performed when it is determined in the determination process of step S20 in FIG. 6 that there is execution history information 4 including the task ID of the task to be executed by the selected node 32. It may also be applied in cases. For example, if there is a plurality of executed history information 4 including the task ID of the task to be executed by the selected node 32, the data closest to the data size of the file attached to the requested task among the history information 4 The processing time of the history information 4 to which the file of the size is attached may be estimated as the processing time of the requested task.

Further, the priority of the task may be set in the instruction content for the task specified by the request. Task priority defines the order in which tasks are executed when multiple tasks are executed in the information processing system 30, and a task with a higher priority is given priority over other tasks. Resources are allocated and the execution is completed as soon as possible. If there is a task with a higher priority than the priority task (priority task), the task with a lower priority than the priority task will be executed after the priority task has finished or if the priority task is not being executed. Not run in free time. In other words, even if the tasks are of the same type, the lower the priority of the task, the longer the processing time.

Therefore, when there is a node 32 that executes a priority task among the nodes 32 that are being executed, the CPU 41 estimates the priority according to the difference between the priority of the task executed by the selected node 32 and the priority of the priority task. The processing time may be extended. In addition, if the priority of the task executed by the selected node 32 is higher than the priority of the task executed by other nodes 32 that are currently being executed, the CPU 41 determines whether the priority of the task executed by the selected node 32 The estimated processing time may be shortened depending on the difference in priority of tasks executed by other nodes 32 that are currently being executed.

Naturally, correction of the processing time using the priority of the task is performed in the determination process of step S20 in FIG. It may also be applied when a judgment is made. Specifically, the CPU 41, depending on the difference between the priority of the requested task and the priority of the task recorded in the executed history information 4 used to estimate the processing time of the requested task, The estimated processing time may be corrected.

Next, resource release processing for the information processing system 30 will be explained. FIG. 8 is a flowchart illustrating an example of the flow of resource release processing executed by the CPU 41 of the control device 20. The resource release process is executed at the timing when the information processing system 30 is activated, and thereafter, while the information processing system 30 is operating, the resource release process is restarted every time the resource release process is completed.

A control program that defines resource release processing is stored in advance in the ROM 42 of the control device 20, for example. The CPU 41 of the control device 20 reads a control program stored in the ROM 42 and executes resource release processing.

In step S100, the CPU 41 determines whether or not there is a node 32 in the information processing system 30 that has completed execution of a task, that is, a completed node 32.

When the task is completed, the information processing system 30 notifies the control device 20 of the processing time of the task at the completed node 32 along with the request ID corresponding to the task executed at the completed node 32. Therefore, the CPU 41 only needs to determine that there is a completed node 32 when there is a task for which the processing time has been notified.

Note that when the CPU 41 receives the processing time of a task from the information processing system 30, it records the processing time of the received task in the processing time of the history information 4 that includes the request ID that was also received.

If there is no completed node 32, the CPU 41 repeatedly executes the determination process in step S100 and monitors the execution state of the task until a completed node 32 appears. On the other hand, if there is a completed node 32, the process moves to step S110.

In step S110, the CPU 41 controls the communication unit 47 to issue a deletion instruction that instructs the information processing system 30 to delete the completed node 32 along with the request ID associated with the task executed on the completed node 32. Output. As a result, the terminated node 32 is deleted from the information processing system 30, and the resources used by the terminated node 32 are released.

In step S120, the CPU 41 refers to the configuration state of the information processing system 30 notified from the information processing system 30, and determines whether another node 32 exists in the instance 34 where the terminated node 32 was placed. judge. If another node 32 exists, the process moves to step S100, and the execution state of the task is monitored until the next completed node 32 appears.

On the other hand, if no other node 32 exists in the instance 34 where the terminated node 32 was placed, that is, if no node 32 is placed in the instance 34 where the terminated node 32 was placed, , the process moves to step S130. Note that an instance 34 in which no node 32 is arranged will be referred to as an "empty instance 34."

Since no task is being executed in the empty instance 34, deletion will not affect the execution of the task requested by the user. Therefore, in step S130, the CPU 41 controls the communication unit 47 to output a deletion instruction to the information processing system 30 to instruct deletion of the empty instance 34. As a result, the empty instance 34 is deleted from the information processing system 30, and the use of the empty instance 34 ends.

In step S140, the CPU 41 determines whether all instances 34 used in the information processing system 30 have been deleted. If at least one instance 34 exists in the information processing system 30, the process moves to step S100, and the execution state of the task is monitored until the next completed node 32 appears.

That is, by repeatedly executing steps S100 to S140, the terminated nodes 32 whose tasks have been executed and the empty instances 34 are sequentially deleted.

Therefore, if the resource release process is executed for the configuration example of the information processing system 30 after the relocation process shown in FIG. 7 has been executed, and each node 32 executes a task at the same timing, , since the processing time of task B is longer than the processing time of task A, the first instance 34 and the second instance 34 are deleted first, and finally the third instance is deleted.

On the other hand, if the resource release process is executed for the configuration example of the information processing system 30 before the relocation process shown in FIG. 7 is executed, and each node 32 executes the task at the same timing, for example, , B. The first instance 34 and the second instance 34 will continue to exist until the B node 32 finishes executing task B.

That is, in the example of FIG. 7, when the relocation process is executed, the number of instances 34 that survive beyond the threshold is one, whereas when the relocation process is not executed, the number of instances 34 that survive beyond the threshold is one. The number of instances 34 that survive beyond this will be two. Therefore, if the usage fee for the cloud service that builds the information processing system 30 is set by the number of instances 34 and the usage time, it is better to perform the relocation process than to not perform the relocation process. This shows a tendency for usage fees to be suppressed.

On the other hand, if it is determined in the determination process of step S140 that the instance 34 does not exist in the information processing system 30, the resource release process shown in FIG. 8 is ended.

Note that if the information processing system 30 is still in operation even after completing the resource release process, the CPU 41 executes the resource release process shown in FIG. The node 32 and instance 34 that have finished execution are deleted.

As described above, according to the control device 20 according to the present embodiment, the processing time of the task executed by the node 32 is estimated with reference to the history information 4, and if the processing time of the task exceeds the threshold, Nodes 32 that execute tasks are aggregated into long-term instances 34. In addition, while the information processing system 30 is operating, the control device 20 monitors the execution status of tasks in the nodes 32, and each time the existence of a terminated node 32 or an empty instance 34 is detected, the control device 20 deletes the task and terminates the task. The resources used by the completed nodes 32 and empty instances 34 are released.

<Second embodiment>
In the first embodiment, an example has been described in which the control device 20 focuses on the processing time of a task and performs control to change the configuration of the information processing system 30 that has already been constructed. However, the method of controlling the configuration of the information processing system 30 by focusing on the processing time of a task is not limited to this.

In the second embodiment, a control device 20A that controls the configuration of the information processing system 30 before requesting the information processing system 30 to execute a task will be described.

The system configuration example of the control system 100 including the control device 20A is the same as the system configuration example shown in FIG. 1, and the functional configuration example of the control device 20A is the same as the functional configuration example of the control device 20 shown in FIG. It is. Moreover, the main part configuration example of the electrical system in the control device 20A is configured using the computer 40, similar to the main part configuration example of the electrical system in the control device 20 shown in FIG.

FIG. 9 is a flowchart showing an example of the flow of the node 32 placement control process executed by the CPU 41 of the control device 20 when a request is received from the input device 10.

A control program that defines the arrangement control process for the nodes 32 is stored in advance in the ROM 42 of the control device 20A, for example. The CPU 41 of the control device 20A reads the control program stored in the ROM 42 and executes the arrangement control process for the nodes 32.

In step S200, the CPU 41 determines whether executed history information 4 including the same task ID as the task ID of the requested task received from the input device 10 exists in the history information DB 25. If executed history information 4 including the task ID exists, the process moves to step S210.

In this case, the processing time when a task of the same type as the requested task is executed is recorded in the executed history information 4. Therefore, in step S210, the CPU 41 obtains the processing time from the executed history information 4 that includes the same task ID as the task ID of the requested task, and estimates the obtained processing time as the processing time of the requested task. . Note that each of the estimation methods described in step S30 of FIG. 6 in the first embodiment may be applied to the estimation method of the task processing time.

In step S220, the CPU 41 determines whether the processing time estimated in step S210 exceeds a threshold value. If the processing time exceeds the threshold, the requested task is a long-term task, and the process moves to step S230. Then, in step S230, the CPU 41 controls the communication unit 47 to instruct the information processing system 30 to place the task associated with the request ID in the long-term instance 34 along with the request ID associated with the task. Outputs placement instructions.

Then, in step S240, the CPU 41 controls the communication unit 47 to output an execution instruction for instructing the information processing system 30 to execute the task together with the request ID associated with the task. As a result, if the task instructed to be executed is a task instructed to be placed in the long-term instance 34, the information processing system 30 places the node 32 that executes the task in the long-term instance 34, and then executes the task. do.

On the other hand, if it is determined in the determination process of step S200 that there is no executed history information 4 that includes the same task ID as the task ID of the requested task, the processes of steps S210 to S230 are not executed. The process moves to step S240. Further, if it is determined in the determination process of step S220 that the processing time of the task is equal to or less than the threshold value, the process proceeds to step S240 without executing the process of step S230. That is, in these cases, the information processing system 30 does not receive the instruction to execute the task because the placement instruction to place the task in the long-term instance 34 in the process of step S230 is not output to the information processing system 30. The task is placed in an instance 34 different from the long-term instance 34.

With the above steps, the arrangement control process for the nodes 32 shown in FIG. 9 is completed.

The control device 20A may execute the relocation process shown in FIG. 6 while the information processing system 30 is operating. However, the control device 20A executes the arrangement control process for the nodes 32 shown in FIG. 9 before causing the information processing system 30 to execute the task. Therefore, it is determined whether the requested task is a long-term task based on the same history information 4. Therefore, if the relocation process shown in FIG. 6 is simply executed, the configuration of the information processing system 30 will change. Sometimes things don't change.

Therefore, when estimating the processing time of the selected node 32 in step S30 of the relocation process shown in FIG. real time) is acquired from the information processing system 30.

The processing time of the selected node 32 estimated from the history information 4 may differ from the actual processing time of the selected node 32 due to differences in the load on the information processing system 30. Even if a task is presumed not to be a long-term task, it may actually be a long-term task. In this case, the selected nodes 32 whose actual time of executing the task exceeds the threshold will be placed in the long-term instance 34.

As described above, according to the control device 20A according to the present embodiment, each time the control device 20A receives a request from the input device 10, the requested task is determined by referring to the history information 4 managed in the history information DB 25. Determine whether it is a long-term task. Then, if the requested task is estimated to be a long-term task, the control device 20 controls the information processing system 30 so that the node 32 that executes the requested task is placed in the long-term instance 34. Control.

<Third embodiment>
In the first embodiment, a description has been given of the control device 20 that performs control to change the configuration of the information processing system 30 after requesting the information processing system 30 to execute a requested task. Furthermore, in the second embodiment, the control device 20A is described which performs control to specify the configuration of the information processing system 30 before requesting the information processing system 30 to execute a requested task.

In the present embodiment, in a situation where the configuration of the information processing system 30 is changed, for example, when the program corresponding to the type of task used in the node 32 of the information processing system 30 is updated due to a change in the content of the task, the configuration of the information processing system 30 is changed. Compared to the case where the configuration before the change and the configuration after the change coexist in the information processing system 30, the cost of the information processing system 30 due to the change in the configuration can be suppressed without stopping the services provided by the information processing system 30. The control device 20B will be explained.

Note that the system configuration example of the control system 100 including the control device 20B is the same as the system configuration example shown in FIG.

FIG. 10 is a diagram showing an example of the functional configuration of the control device 20B. The functional configuration example of the control device 20B shown in FIG. 10 is different from the functional configuration example of the control devices 20 and 20A shown in FIG. The other points are the same as the functional configuration example shown in FIG.

When the service provided by the information processing system 30 is upgraded, the contents of tasks change accordingly, and the program of the information processing system 30 that is compatible with the old version of the service will execute the task that is compatible with the new version of the service. be unable to do so. On the other hand, when switching services in the information processing system 30, tasks corresponding to the old version of the service accepted before the service switching and tasks corresponding to the new version of the service accepted after the service switching coexist. Even in this state, the information processing system 30 needs to provide each version of the service.

Therefore, the version management unit 26 manages task versions for each type of task. Note that there is a one-to-one correspondence between the task version and the version of the service provided by the information processing system 30, and when the service is upgraded, the corresponding task version is also upgraded.

The correspondence between task types and task versions is defined by version information 6 and managed by version DB 27.

FIG. 11 is a diagram showing an example of version information 6 managed by the version DB 27. As shown in FIG. 11, in the version information 6, a task ID representing the type of task and a task version representing the latest version of the task are managed in association with each other for each type of task. According to the example of version information 6 in FIG. 11, the task version of task A is "1.0" and the task version of task B is "2.0".

The delivery unit 28 queues the task output together with the request ID from the output unit 24 in the queue 29 (see FIG. 10). The delivery unit 28 has a plurality of queues 29 prepared in advance for each type of task and each task version, and the delivery unit 28 sends the request to the queue 29 corresponding to the task ID and task version of the requested task. Queue the task that was created.

The information processing system 30 has queue information that defines which queue 29 corresponds to which combination of task ID and task version, and the information processing system 30 refers to the queue information. Obtain a task from queue 29. That is, the information processing system 30 can obtain the task ID and task version of the obtained task depending on which queue 29 the task was obtained from, without having to determine the task ID and task version of the obtained task by itself. become. Therefore, the information processing system 30 only needs to execute the acquired task at the node 32 corresponding to the type and version of the task.

The main part configuration example of the electrical system in the control device 20B is configured using the computer 40, similar to the main part configuration example of the electrical system in the control device 20 shown in FIG.

Next, the operation of the control device 20B when a program for executing a specific task is upgraded in the information processing system 30 will be described in detail.

FIG. 12 is a flowchart illustrating an example of the flow of the node 32 placement control process executed by the CPU 41 of the control device 20B when a request is received from the input device 10.

A control program that defines the arrangement control process for the nodes 32 is stored in advance in the ROM 42 of the control device 20B, for example. The CPU 41 of the control device 20B reads the control program stored in the ROM 42 and executes the arrangement control process for the nodes 32.

The flowchart shown in FIG. 12 differs from the flowchart of the node 32 placement control process according to the second embodiment shown in FIG. 9 in that step S240 is deleted and steps S250 to S290 are added instead. .

If the requested task is a long-term task, step S250 is executed after outputting a placement instruction to the information processing system 30 to place the requested task in the long-term instance 34.

In step S250, the CPU 41 determines whether the task has been upgraded. Note that whether or not a task has been upgraded may be determined based on, for example, whether or not an upgrade flag has been set. If the version of the task has been updated, the process moves to step S260.

In step S260, the CPU 41 updates version information 6. For example, if the task version of task A has been upgraded from "1.0" to "2.0", the task version of task A in version information 6 shown in FIG. 11 is updated to "2.0". Having confirmed that the task has been upgraded, the CPU 41 resets the version upgrade flag.

In step S270, the CPU 41 creates a new queue 29 corresponding to the upgraded task.

In step S280, the CPU 41 controls the communication unit 47 and instructs the communication unit 47 to update the queue information in which the identification information of the queue 29 from which the upgraded task is obtained is associated with the task ID indicating the type of the upgraded task. is output to the information processing system 30.

As a result, the acquisition destination of the upgraded task, which was associated with the old version of the queue 29 in the queue information, is changed to the newly generated new version of the queue 29.

Therefore, if the task version of task A is upgraded from "1.0" to "2.0", the information processing system 30 that has received the queue information update instruction will , the A node 32 had acquired task A from the queue 29 where task A with the task version "1.0" was queued, but after receiving the instruction to update the queue information, the task version becomes "1.0". 2.0'' task A is acquired from the queue 29 where task A is queued.

On the other hand, if it is determined in the determination process of step S250 that the task has not been upgraded, the process moves to step S290 without executing steps S260 to S280.

In step S290, the CPU 41 refers to the version information 6, obtains the task version from the task ID of the requested task, and queues the requested task in the queue 29 corresponding to the task ID and task version of the requested task. ing.

With the above steps, the arrangement control process for the nodes 32 shown in FIG. 12 is completed.

If the requested task is an upgraded long-term task, the CPU 41 controls the configuration of the information processing system 30 so that the requested task is placed in the long-term instance 34, and then updates the new version. The requested task will be queued in the queue 29 corresponding to.

Note that in the third embodiment as well, the CPU 41 executes the resource release process shown in FIG. The resource release process is restarted each time the process is completed. As a result, the nodes 32 that have finished executing the old version tasks are sequentially deleted, and eventually the instances 34 that included the nodes 32 that were executing the old version tasks are also deleted. In other words, even if the program is upgraded in the information processing system 30, the control device 20B prevents the execution of the requested task from being stopped, and the control device 20B maintains the information processing system that was used before the program was upgraded. Control is performed to delete the configuration of the information processing system 30 and construct the configuration of the information processing system 30 that executes the updated task.

Although the present invention has been described above using each embodiment, the present invention is not limited to the scope described in each embodiment. Various changes or improvements can be made to each embodiment without departing from the gist of the present invention, and forms with such changes or improvements are also included within the technical scope of the present invention. For example, the order of processing may be changed without departing from the gist of the present invention.

In each of the embodiments, an example has been described in which each process is implemented using software. However, processes equivalent to each of the flowcharts shown in FIGS. 6, 8, 9, and 12 can be implemented using, for example, an ASIC (Application Specific Integrated Integrated It may also be implemented in a FPGA (Field Programmable Gate Array), a FPGA (Field Programmable Gate Array), or a PLD (Programmable Logic Device) and processed by hardware. In this case, the processing speed can be increased compared to the case where each process is implemented by software.

In this way, the CPU 41 may be replaced with a dedicated processor specialized for specific processing, such as an ASIC, FPGA, PLD, GPU (Graphics Processing Unit), and FPU (Floating Point Unit).

The operation of the CPU 41 in each embodiment may be realized not only by one CPU 41 but also by a plurality of CPUs 41. Furthermore, the operation of the CPU 41 in each embodiment may be realized by the cooperation of the CPUs 41 in a plurality of computers 40 located at physically separate locations.

Further, in each embodiment, the control devices 20, 20A, 20B and the information processing system 30 are installed separately, but the control devices 20, 20A, 20B and the information processing system 30 may be integrated and operated.

Further, in the above-described embodiment, the control program is installed in the ROM 42, but the present invention is not limited to this. The control program according to the present invention can also be provided in a form recorded on a computer-readable storage medium. For example, the control program according to the present invention may be provided in a form recorded on an optical disc such as a CD (Compact Disc)-ROM or a DVD (Digital Versatile Disc)-ROM. Further, the control program according to the present invention may be provided in a form recorded in a portable semiconductor memory such as a USB (Universal Serial Bus) memory or a memory card.

2 communication line, 4 history information, 6 version information, 10 input device, 20 (20A, 20B) control device, 21 reception section, 22 control section, 23 history management section, 24 output section, 25 history information DB, 26 version management section, 27 version DB, 28 delivery section, 29 queue, 30 information processing system, 32 node, 34 instance, 40 computer, 41 CPU, 42 ROM, 43 RAM, 44 non-volatile memory, 47 communication unit, 48 input unit, 49 Display unit, 100 control system

Claims (8)

If at least one processing entity that executes the received processing is running in each of the multiple processing environments whose configuration is being changed, which is built in the information processing system that accepts the requested processing, the processing time The processing entities that execute processing that satisfy predetermined conditions for are aggregated into a specific processing environment, and the processing of each processing entity is continued , and when the processing in the processing entity is completed, the processing entity exists. controlling the information processing system to sequentially release resources of the information processing system that are no longer being used by deleting the processing environments in order from the processing environment that is no longer being used, and finally deleting the specific processing environment; A control device equipped with a control section. The control device according to claim 1, wherein the predetermined condition is that a processing time from the start to the end of the process exceeds a threshold value. The control unit refers to history information in which processing times of executed processes executed so far and processing conditions of the executed processes are recorded, and determines the history information of the executed processes corresponding to the requested process. 3. The control device according to claim 2, wherein the control device estimates a processing time of a process requested by the information processing system and instructs the information processing system to relocate the processing entity. When the data size of data received in response to a request for the executed process is recorded as a processing condition in the history information,
The control device according to claim 3, wherein the control unit performs control to estimate the processing time of the requested task from the processing time of the data size in the executed process corresponding to the requested process. When the priority regarding the processing order of the requested process is set lower than other processes being executed in the processing entity, the control unit extends the estimated processing time according to the set priority. The control device according to claim 3 or 4, which performs control. The control unit controls the information processing system to place the processing entity that executes the pre-execution request in the specific processing environment when the estimated processing time for the pre-execution request exceeds the threshold. The control device according to any one of claims 3 to 5, which controls. The control unit performs control to instruct the information processing system to relocate the processing entity multiple times during a period when the information processing system is in operation. The control device described in . A control program for causing a computer to function as a control unit of a control device according to any one of claims 1 to 7 .