Management device, system, management method, and management program
The management device optimizes data transfer and job execution by transferring data only when a threshold is met and instructing execution upon completion, addressing integration and throughput issues in data stream and HPC systems.
Patent Information
- Application Number
- JP2022565115
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Priority Date
- 2020-11-27
- Filing Date
- 2021-10-20
- Publication Date
- 2025-01-07
- Estimated Expiration
- 2041-10-20
AI Technical Summary
Existing data stream management systems are difficult to integrate with various application execution infrastructures, and HPC systems struggle with microdata handling, leading to high job management and scheduling costs and decreased throughput.
A management device that transfers data to an execution device when the data amount exceeds a specified value and instructs the execution to execute the job upon completion, reducing redundant data transfers and application executions.
This approach reduces load and improves throughput by minimizing data transfer times and job scheduler loads when handling microdata.
Smart Images

Figure 0007609172000001 
Figure 0007609172000002 
Figure 0007609172000003
Abstract
Description
[Technical field]
[0001] The present invention relates to a management device, a management system, a management method, and a recording medium. [Background technology]
[0002] High Performance Computing (HPC) is mainly used for simulations that input and output very large amounts of data (Patent Document 1). File systems in HPC are good at reading and writing large amounts of data files by multiple clients.
[0003] On the other hand, systems for IoT (Internet of Things) and AI (Artificial Intelligence) handle very small amounts of data (called microdata). In a file system that handles microdata, files are frequently created, so file creation performance is more important than file read / write performance. Systems that handle microdata mainly use data stream management systems (Patent Document 2). [Prior art documents] [Patent documents]
[0004] [Patent Document 1] Special Publication No. 2010-539594 [Patent Document 2] Special Publication No. 2019-537092 Summary of the Invention [Problem to be solved by the invention]
[0005] Since the data stream management system is a dedicated system, it is difficult to use it in conjunction with various application platforms.
[0006] In addition, although HPC systems can be used in conjunction with various application platforms, they are not good at handling microdata, for which file creation performance is important.
[0007] In addition, general job schedulers manage a large number of short-running jobs for IoT and AI one by one, which results in huge job management and scheduling costs and reduced throughput.
[0008] An object of the present invention is to provide a management device, system, management method, and recording medium that make it possible to reduce the load and improve throughput when executing jobs using microdata in a more versatile manner. [Means for solving the problem]
[0009] In one aspect of the present invention, the management device is characterized by having a transfer means for transferring data related to a specified job to an execution device that executes an application when the amount of received data becomes equal to or greater than a specified value, and an instruction means for instructing the execution device to execute the job by the application when the transfer of the data is completed.
[0010] In another aspect of the present invention, the management method is characterized in that, when the amount of received data relating to a specified job becomes equal to or greater than a specified value, the data is transferred to an execution device that executes an application, and, when the transfer of the data is completed, the execution device is instructed to execute the job by the application.
[0011] In another aspect of the present invention, a management program recorded on a computer-readable recording medium causes a computer to execute a transfer function of transferring data related to a specified job to an execution device that executes an application when the amount of received data becomes equal to or greater than a specified value, and an instruction function of instructing the execution device to execute the job by the application when the data transfer is completed. Effect of the Invention
[0012] According to the present invention, it is possible to realize a more versatile method for reducing the load and improving the throughput when executing a job using microdata. [Brief description of the drawings]
[0013] [Figure 1] FIG. 2 illustrates an example of a configuration of a management device according to the first embodiment of the present invention; [Diagram 2] FIG. 4 illustrates an example of the operation of a management device according to the first embodiment of the present invention; [Diagram 3] FIG. 13 is a diagram including a configuration example of a system including a management device according to a second embodiment of the present invention. [Figure 4] FIG. 11 illustrates an example of a configuration of a management device according to a second embodiment of the present invention. [Diagram 5] FIG. 11 illustrates an example of the configuration of an execution device according to a second embodiment of the present invention. [Figure 6] FIG. 11 illustrates an example of the operation of a management device and an execution device according to the second embodiment of the present invention. [Figure 7] FIG. 11 is a diagram illustrating an example of the operation of a management device according to the second embodiment of the present invention. [Figure 8] FIG. 2 is a diagram illustrating an example of a hardware configuration according to each embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0014] [First embodiment] A first embodiment of the present invention will be described.
[0015] 1 shows an example of the configuration of a management device 10 according to this embodiment. The management device 10 according to this embodiment includes a transfer unit 11 and an instruction unit 12.
[0016] When the amount of received data relating to a specific job reaches or exceeds a designated value, the transfer unit 11 transfers the data to an execution device that executes the application. Furthermore, when the data transfer is completed, the instruction unit 12 instructs the execution device to execute the job by the application.
[0017] By configuring the management device 10 in this manner, when the amount of received data relating to a specific job reaches or exceeds a designated value, the management device 10 transfers the data to an execution device that executes an application, and instructs the execution device to execute the job using the application. This allows the management device 10 to reduce the number of times data is transferred to the execution device, compared to when the application is executed each time data is received. As a result, the data transfer time and the load on the file system can be reduced. In addition, since the number of times an application is executed can be reduced, the load on the job scheduler can be reduced. This makes it possible to reduce the load and improve the throughput when executing a job using microdata in a more versatile manner.
[0018] Next, an example of the operation of the management device 10 of this embodiment is shown in FIG.
[0019] When the amount of received data relating to a predetermined job is equal to or greater than a designated value, the transfer unit 11 transfers the data to an execution device that executes the application (step S101). Furthermore, when the data transfer is completed, the instruction unit 12 instructs the execution device to execute the job by the application (step S102).
[0020] By operating in this manner, when the amount of received data relating to a specific job reaches or exceeds a designated value, the management device 10 transfers the data to an execution device that executes an application, and instructs the execution device to execute the job using the application. This allows the management device 10 to reduce the number of times data is transferred to an execution device, compared to when the application is executed each time data is received. As a result, the data transfer time and the load on the file system can be reduced. In addition, since the number of times an application is executed can be reduced, the load on the job scheduler can be reduced. This makes it possible to reduce the load and improve the throughput when executing a job that uses microdata in a more versatile manner.
[0021] As described above, in the first embodiment of the present invention, when the amount of received data relating to a specific job reaches or exceeds a specified value, the management device 10 transfers the data to an execution device that executes an application and instructs the execution device to execute the job by the application. This makes it possible to reduce the load and improve the throughput when executing a job that uses microdata in a more versatile manner.
[0022] [Second embodiment] Next, a second embodiment of the present invention will be described.
[0023] First, FIG. 3 shows an example of the configuration of a system including a management device 20 of this embodiment.
[0024] The management device 20 is a device that manages the execution of applications in the execution devices 40-i. The execution devices 40-i (i is an integer between 1 and m) are devices that execute applications. The management device 20 and the execution devices 40-i are connected to each other.
[0025] A user terminal 60 is connected to the management device 20. The user terminal 60 is a terminal used by a job submitter to cause an application to execute a job.
[0026] Next, Fig. 4 shows an example of the configuration of management device 20 of this embodiment. Management device 20 of this embodiment includes a forwarder 21, a conductor 22, and a scheduler 23. Forwarder 21 corresponds to forwarding unit 11 of management device 10 (Fig. 1) of the first embodiment. Conductor 22 corresponds to instruction unit 12 of management device 10 of the first embodiment.
[0027] The scheduler 23 schedules jobs to be executed by the application. The scheduler 23 receives job information indicating information related to jobs to be executed by the application from the user terminal 60. The job information includes execution application information, maximum number of expansion nodes, ignition condition information, required resource information, and notification means information.
[0028] The execution application information is information indicating the application that will execute the job. The maximum number of expansion nodes is the maximum value for the execution device that will execute the job. The ignition condition information is information regarding the conditions for causing the application to execute the job. In this embodiment, the ignition condition information indicates a condition regarding the amount of data that the management device 20 receives from the user terminal 60. The management device 20 causes the application to execute the job when the amount of data received from the user terminal 60 becomes equal to or greater than the amount of data indicated by the ignition condition information.
[0029] The required resource information is information indicating the amount of resources required for an application to execute the job. The notification means information is information indicating the type of notification means (e.g., file, named pipe, shared memory, etc.) when the receiving unit 41-i of the execution device 40-i notifies the execution unit 42-i of the start of execution of the application. The execution unit 42-i monitors the notification means indicated by the notification means information, and executes the application when the start of execution of the application is notified via the notification means. The notification means is also a write destination interface to which the execution device 40-i writes data received from the management device 20.
[0030] When the scheduler 23 receives the job information, it determines the reception number (job ID (Identification)), authentication token, and port number of the forwarder 21, and transmits the determined information to the user terminal 60. The scheduler 23 also associates the determined information with the job information and stores them in a storage unit (not shown). The scheduler 23 also selects the number of execution devices indicated by the maximum number of expansion nodes. The execution devices selected here are hereinafter referred to as execution device candidates. The scheduler 23 then starts the application on the selected execution device 40-i, and also pauses the application.
[0031] The authentication token is used as authentication information when the user terminal 60 accesses the forwarder 21 of the management device 20.
[0032] The forwarder 21 receives the microdata. The microdata includes extremely small data. The microdata is transmitted from the user terminal 60 to the port number transmitted from the scheduler 23 to the user terminal 60 in response to a user operation. A job ID is assigned to the microdata.
[0033] Furthermore, when the amount of microdata received from the user terminal 60 exceeds a designated value, the forwarder 21 notifies the conductor 22 of this fact. In this embodiment, the designated value is a value indicated by the ignition condition information. The ignition condition information is stored in the storage unit in association with a job ID. The amount of received data is managed for each job ID.
[0034] When the conductor 22 receives a notification that the amount of received microdata has exceeded a specified value, it acquires information on the load status of each execution device 40-i from the scheduler 23. The conductor 22 also determines the execution device 40-i to which the microdata is to be transferred based on the load status information. The conductor 22 determines the execution device 40-i with the smallest load as the destination of the microdata. The conductor 22 selects the execution device 40-i to which the microdata is to be transferred from among the candidate execution devices.
[0035] The conductor 22 determines the destination execution device 40-i of the microdata, for example, in the following manner.
[0036] First, the conductor 22 checks whether there is an execution device 40-i on which a job is not being executed. If there is an execution device 40-i on which a job is not being executed, the conductor 22 determines that execution device 40-i as the transfer destination of the microdata. If there is no execution device 40-i on which a job is not being executed, the conductor 22 determines an execution device 40-i having free resources of the amount indicated by the required resource information as the transfer destination of the microdata. If there is no execution device 40-i having free resources of the amount indicated by the required resource information, the conductor 22 determines an execution device 40-i with the lowest priority of the job being executed among the candidate execution devices as the transfer destination of the microdata.
[0037] The forwarder 21 then transmits the microdata to the destination execution device 40-i determined by the conductor 22.
[0038] Furthermore, when the forwarder 21 is notified by the execution device 40-i that the reception of the microdata has been completed, the forwarder 21 notifies the conductor 22 that the transfer of the data has been completed.
[0039] When the conductor 22 is notified of the completion of the data transfer, it notifies the receiving unit 41-i of the execution device 40-i that the application is ready to be executed.
[0040] Next, the execution device 40-i of this embodiment will be described with reference to Fig. 5. The execution device 40-i includes a receiving unit 41-i and an execution unit 42-i.
[0041] The receiving unit 41-i receives microdata transmitted by the forwarder 21 of the management device 20. The receiving unit 41-i also receives notification means information. Note that the receiving unit 41-i may obtain the notification means information by referring to a storage unit in which the job information is stored, instead of receiving the notification means information.
[0042] When the receiver 41-i receives the microdata, it writes the received microdata into the notification means indicated by the notification means information, and notifies the forwarder 21 of the management device 20 of the completion of reception of the microdata.
[0043] Furthermore, when the receiving unit 41-i is notified by the management device 20 that the application is ready to be executed, the receiving unit 41-i releases the pause of the application. Then, the receiving unit 41-i notifies the execution unit 42-i of the start of execution of the application via the notification means.
[0044] The execution unit 42-i executes the application, and upon completion of the execution of the application, notifies the reception unit 41-i of the completion of the application.
[0045] When the reception unit 41-i is notified of the termination of the application, it puts the application into a suspended state.
[0046] By configuring the management device 20 in this manner, when the amount of received data relating to a specific job reaches or exceeds a designated value, the management device 20 transfers the data to an execution device that executes an application, and instructs the execution device to execute the job using the application. This allows the management device 20 to reduce the number of times data is transferred to an execution device, compared to when the application is executed each time data is received. As a result, the data transfer time and the load on the file system can be reduced. Furthermore, since the number of times an application is executed can be reduced, the load on the job scheduler can be reduced. This makes it possible to reduce the load and improve the throughput when executing a job that uses microdata in a more versatile manner.
[0047] Next, an example of the operation of management device 20 and execution device 40-i of this embodiment will be described with reference to FIG.
[0048] The scheduler 23 of the management device 20 receives job information from the user terminal 60 (step S301). The job information includes execution application information, the maximum number of expandable nodes, ignition condition information, necessary resource information, and notification means information.
[0049] When the scheduler 23 receives the job information, it starts the forwarder 21, the conductor 22, and the receiving unit 41-i for the job (step S302). The scheduler 23 also determines the reception number (job ID), the authentication token, and the port number of the forwarder 21, and transmits the determined information to the user terminal 60 (step S303). The scheduler 23 also stores the determined information in association with the job information in the storage unit. The scheduler 23 also selects the number of execution devices indicated by the maximum number of expanded nodes. The scheduler 23 then starts the application on the selected execution device 40-i, and also pauses the application.
[0050] The forwarder 21 receives the microdata (step S304). The microdata includes a very small data file. The microdata is transmitted from the user terminal 60 to the port number transmitted from the scheduler 23 to the user terminal 60 in response to a user operation. A job ID is assigned to the microdata.
[0051] Furthermore, when the amount of microdata received from the user terminal 60 exceeds a designated value, the forwarder 21 notifies the conductor 22 of this fact (step S305). In this embodiment, the designated value is a value indicated by the ignition condition information.
[0052] When the conductor 22 receives a notification that the amount of received microdata has exceeded a designated value, the conductor 22 acquires information on the load status of each execution device 40-i from the scheduler 23. The conductor 22 also determines the execution device 40-i to which the microdata is to be transferred based on the load status information. The conductor 22 determines the execution device 40-i with the lightest load as the destination of the microdata. The conductor 22 then transmits the information on the transfer destination to the forwarder 21 (step S306).
[0053] The conductor 22 determines the destination execution device 40-i of the microdata, for example, by the method shown in FIG.
[0054] First, the conductor 22 checks whether there is an execution device 40-i on which a job is not being executed (step S401). If there is an execution device 40-i on which a job is not being executed (YES in step S401), the conductor 22 determines the execution device 40-i as the transfer destination of the microdata (step S402). If there is no execution device 40-i on which a job is not being executed (NO in step S401), the conductor 22 checks the free resources of the execution device 40-i. Then, the conductor 22 determines the execution device 40-i with free resources of the amount of resources indicated by the required resource information (YES in step S403) as the transfer destination of the microdata (step S404). If there is no execution device 40-i with free resources of the amount of resources indicated by the required resource information (NO in step S403), the conductor 22 checks the priority of the job being executed by the execution device candidate. Then, the conductor 22 determines, from among the candidate execution devices, the execution device 40-i having the lowest priority of the job being executed, as the transfer destination of the microdata (step S405).
[0055] Then, the forwarder 21 transmits the microdata to the destination execution device 40-i determined by the conductor 22 (step S307 in FIG. 6).
[0056] When the receiver 41-i of the execution unit 40-i receives the microdata, it writes the received microdata into the notification means indicated by the notification means information, and notifies the forwarder 21 of the management unit 20 of the completion of reception of the microdata (step S308).
[0057] Furthermore, when the forwarder 21 is notified by the execution device 40-i that the reception of the microdata has been completed, the forwarder 21 notifies the conductor 22 of the completion of the transfer of the data (step S309).
[0058] When the conductor 22 is notified of the completion of the data transfer, it notifies the receiving unit 41-i of the execution device 40-i that the application is ready to be executed (step S310).
[0059] Furthermore, when the receiving unit 41-i is notified by the management device 20 that the application is ready to be executed, the receiving unit 41-i releases the pause of the application. Then, the receiving unit 41-i notifies the execution unit 42-i of the start of execution of the application via the notification means (step S311).
[0060] The execution unit 42-i executes the application, and upon completion of the execution of the application, notifies the reception unit 41-i of the completion of the application.
[0061] When the reception unit 41-i is notified of the termination of the application, it puts the application into a suspended state.
[0062] By operating in this manner, when the amount of received data relating to a predetermined job reaches or exceeds a designated value, the management device 20 transfers the data to an execution device that executes an application, and instructs the execution device to execute the job using the application. This allows the management device 20 to reduce the number of times data is transferred to an execution device, compared to when the application is executed each time data is received. As a result, the data transfer time and the load on the file system can be reduced. Furthermore, since the number of times an application is executed can be reduced, the load on the job scheduler can be reduced. This makes it possible to reduce the load and improve the throughput when executing a job that uses microdata in a more versatile manner.
[0063] As described above, in the second embodiment of the present invention, when the amount of received data relating to a specific job reaches or exceeds a specified value, the management device 20 transfers the data to an execution device that executes an application and instructs the execution device to execute the job by the application. This makes it possible to reduce the load and improve the throughput when executing a job that uses microdata in a more versatile manner.
[0064] In addition, the execution device 40-i of this embodiment writes the data received from the management device 20 to the notification means indicated by the notification means information. This makes it possible to transfer data using the notification means desired by the user. Furthermore, if a means other than a file is used as the notification means, it is possible to further reduce the load on the file system.
[0065] [Hardware configuration example] An example of the configuration of hardware resources in which the management device (10, 20) and execution device 40-i (hereinafter referred to as the management device, etc.) in each of the above-mentioned embodiments of the present invention are realized using one information processing device (computer) will be described. The management device, etc. may be realized physically or functionally using at least two or more information processing devices. The management device, etc. may be realized as a dedicated device or a general-purpose device. Also, only some of the functions of the management device, etc. may be realized using an information processing device.
[0066] 8 is a diagram illustrating an example of a hardware configuration of an information processing device capable of implementing the management device of each embodiment of the present invention. The information processing device 90 includes a communication interface 91, an input / output interface 92, a computing device 93, a storage device 94, a non-volatile storage device 95, and a drive device 96.
[0067] For example, the transfer unit 11 and the instruction unit 12 in FIG.
[0068] The communication interface 91 is a communication means for the management device or the like of each embodiment to communicate with an external device at least in a wired or wireless manner. When the management device or the like is realized using at least two information processing devices, the devices may be connected to each other via the communication interface 91 so as to be able to communicate with each other.
[0069] The input / output interface 92 is a man-machine interface including a keyboard as an example of an input device and a display as an output device.
[0070] The arithmetic device 93 is realized by, for example, an arithmetic processing device such as a CPU (Central Processing Unit) or a microprocessor, and a plurality of electric circuits. The arithmetic device 93 can read various programs stored in, for example, a non-volatile storage device 95 into the storage device 94 and execute processing according to the read programs.
[0071] The storage device 94 is a memory device such as a RAM (Random Access Memory) that can be referenced by the arithmetic device 93, and stores programs, various data, etc. The storage device 94 may be a volatile memory device.
[0072] The non-volatile storage device 95 is, for example, a non-volatile storage device such as a ROM (Read Only Memory) or a flash memory, and is capable of storing various programs, data, and the like.
[0073] The drive device 96 is, for example, a device that processes reading and writing of data recorded on a recording medium 97, which will be described later.
[0074] The recording medium 97 is any recording medium capable of recording data, such as an optical disk, a magneto-optical disk, or a semiconductor flash memory.
[0075] Each embodiment of the present invention may be realized, for example, by configuring a management device, etc. using an information processing device 90 illustrated in FIG. 8, and supplying a program capable of realizing the functions described in each of the above embodiments to this management device, etc.
[0076] In this case, the embodiment can be realized by having the computing device 93 execute a program supplied to the management device, etc. Also, it is possible to configure some, but not all, of the functions of the management device, etc., in the information processing device 90.
[0077] Furthermore, the above program may be recorded on a recording medium 97, and may be appropriately stored in the non-volatile storage device 95 at the shipping stage or operation stage of the management device, etc. In this case, the method of supplying the above program may employ a method of installing the program in the management device, etc., using an appropriate tool at the manufacturing stage before shipping or at the operation stage, etc. The method of supplying the above program may employ a general procedure such as a method of downloading the program from an external source via a communication line such as the Internet.
[0078] Although the present invention has been described above with reference to the embodiment, the present invention is not limited to the above embodiment. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.
[0079] A part or all of the above-described embodiments can be described as, but is not limited to, the following supplementary notes.
[0080] (Appendix 1) a transfer means for transferring data related to a predetermined job to an execution device that executes the application when the amount of received data related to the job reaches or exceeds a designated value; an instruction means for instructing the execution device to execute the job by the application when the transfer of the data is completed; A management device comprising:
[0081] (Appendix 2) The instruction means determines the execution device to which the data is to be transferred based on a load status of the execution device. 2. The management device according to claim 1 .
[0082] (Appendix 3) The data includes microdata of extremely small size. 3. The management device according to claim 1 or 2.
[0083] (Appendix 4) A management device according to any one of Supplementary Note 1 to Supplementary Note 3; The execution device; A system comprising:
[0084] (Appendix 5) The execution device is a receiving means for receiving notification means information indicating a type of an interface to which the data is to be written, and writing the received data to an interface of the type indicated by the notification means information; an execution means for causing the application to execute the job using the data written to the interface when an execution instruction for the application is received; 5. The system of claim 4, comprising:
[0085] (Appendix 6) The type of the interface is one of a file, a named pipe, and a shared memory. 6. The system of claim 5,
[0086] (Appendix 7) When an amount of received data relating to a predetermined job reaches or exceeds a designated value, the data is transferred to an execution device that executes the application; When the transfer of the data is completed, the execution device is instructed to execute the job by the application. A management method comprising:
[0087] (Appendix 8) The execution device to which the data is to be transferred is determined based on the load status of the execution device. The management method according to claim 7,
[0088] (Appendix 9) The data includes microdata of extremely small size. 9. The management method according to claim 7 or 8,
[0089] (Appendix 10) The execution device is receiving notification means information indicating a type of an interface to which the data is to be written, and writing the received data to an interface of the type indicated by the notification means information; When an instruction to execute the application is received, the application is caused to execute the job using the data written to the interface. The management method according to claim 9,
[0090] (Appendix 11) The type of the interface is one of a file, a named pipe, and a shared memory. 11. The management method according to claim 10.
[0091] (Appendix 12) On the computer, a transfer function for transferring data related to a predetermined job to an execution device that executes the application when the amount of received data related to the job reaches or exceeds a designated value; an instruction function for instructing the execution device to execute the job by the application when the transfer of the data is completed; A computer-readable recording medium having recorded thereon a management program for causing a computer to execute the above-mentioned steps.
[0092] (Appendix 13) The instruction function determines the execution device as a destination of the data transfer based on a load status of the execution device. 13. A computer-readable recording medium having recorded thereon the management program according to claim 12.
[0093] (Appendix 14) The data includes microdata of extremely small size. 14. A computer-readable recording medium having recorded thereon the management program according to claim 12 or 13.
[0094] (Appendix 15) The execution device is receiving notification means information indicating a type of an interface to which the data is to be written, and writing the received data to an interface of the type indicated by the notification means information; When an instruction to execute the application is received, the application is caused to execute the job using the data written to the interface. A computer-readable recording medium having recorded thereon the management program according to claim 14.
[0095] (Appendix 16) The type of the interface is one of a file, a named pipe, and a shared memory. A computer-readable recording medium having recorded thereon the management program according to claim 15.
[0096] This application claims priority to Japanese Patent Application No. 2020-196779, filed on November 27, 2020, the disclosure of which is incorporated herein in its entirety. [Explanation of symbols]
[0097] 10, 20 Management device 11 Transfer section 12 Instruction part 21 Forwarder 22 Conductor 23 Scheduler 40-i Execution unit 41-i Receiving section 42-i Executive Department 60 User terminal 90 Information processing equipment 91 Communication Interface 92 Input / Output Interface 93 Arithmetic unit 94 Storage device 95 Non-volatile storage devices 96 Drive device 97 Recording media
Claims
1. a transfer means for transferring data related to a predetermined job to an execution device that executes the application when the amount of received data related to the job reaches or exceeds a designated value; an instruction means for instructing the execution device to execute the job by the application when the transfer of the data is completed; A management device comprising:
2. The instruction means determines the execution device to which the data is to be transferred based on a load status of the execution device.
2. The management device according to claim 1.
3. A management device according to claim 1 or 2; The execution device; A system comprising:
4. The execution device is a receiving means for receiving notification means information indicating a type of an interface to which the data is to be written, and writing the received data to an interface of the type indicated by the notification means information; an execution means for causing the application to execute the job using the data written to the interface when an execution instruction for the application is received; The system of claim 3, further comprising:
5. The type of the interface is one of a file, a named pipe, and a shared memory.
5. The system of claim 4.
6. A computer comprising: When an amount of received data relating to a predetermined job reaches or exceeds a designated value, the data is transferred to an execution device that executes the application; When the transfer of the data is completed, the execution device is instructed to execute the job by the application. A management method comprising:
7. The computer, The execution device to which the data is to be transferred is determined based on the load status of the execution device. The management method according to claim 6 .
8. On the computer, a transfer function for transferring data related to a predetermined job to an execution device that executes the application when the amount of received data related to the job reaches or exceeds a designated value; an instruction function for instructing the execution device to execute the job by the application when the transfer of the data is completed; A management program for causing the user to execute the above steps.
Citation Information
Patent Citations
A system and method for running computationally intensive database user-defined programs on an accompanying high-performance parallel computer.
JP2010539594A
Processing timestamps and heartbeat events for automatic time progression
JP2019537092A
System for relational-impact based task management
US20190213509A1
Sensor relay device and sensor relay system
WO2017199807A1
Management device, execution environment setting method, stream data processing system
WO2018002976A1