KR102014246B1 - Mesos process apparatus for unified management of resource and method for the same - Google Patents

Abstract

The resource processing apparatus for integrated resource management may include at least one CPU element, a computing resource unit including at least one GPU element and at least one memory element as resources, and resource scheduling according to a resource allocation request for the computing resource unit. A request for allocation of resources in the computing resource unit, and at least some of the requested resources (hereinafter, allocated resources) have been successfully allocated. When a corresponding task information is received, a request is made to allocate a resource from a Mesos slave unit that performs the task through a docker executor dedicated to the allocated resource and the meso slave unit. To determine the allocation of It includes methoxy source master unit (MESOS Master Unit) that controls the resources shared among the tasks to notify the mail source framework unit.

Description

Mesos processing device and method for integrated resource management {MESOS PROCESS APPARATUS FOR UNIFIED MANAGEMENT OF RESOURCE AND METHOD FOR THE SAME}

The present invention relates to a resource management technology, and more particularly, to a method and an apparatus and method for resource integrated management capable of efficiently sharing various resources including a graphic processing unit (GPU).

Mesos technology manages the sharing and isolation of resources within a data center and can operate based on a mechanism for dynamically allocating and isolating resources in a clustered environment. Mesos technology can support independent execution of each application by allocating existing shared resources by allocating the optimal resources for the application to prevent resource interference between the applications. Mesos technology is being developed in various ways to optimize the execution of various tasks in a dynamic cluster environment.

Korean Patent No. 10-1781063 (2017.09.18) relates to a two-stage resource management method and apparatus for dynamic resource management, and receives a task to be executed in a cluster system, and receives a first task corresponding to the input task. Generating an idle task to generate a first virtual machine corresponding to the generated first idle task, executing the input task in the generated first virtual machine, and executing the input task By monitoring the resource usage of the virtual node of the virtual machine to adjust the virtual node when the virtual node is in an overloaded or lightly loaded state, the resource can be dynamically adjusted according to the resource usage state when the task is executed. It can improve the resource utilization efficiency of the system and prevent the degradation of service due to lack of resources.

Korean Patent Publication No. 10-2017-0088096 (2017.08.01) relates to a virtual cluster resource distribution apparatus and method, and generates a plurality of virtual machines (VIRTUAL MACHINE, VM) corresponding to the number of a plurality of compute nodes It includes a resource allocation unit for generating a virtual cluster by placing and a parallel generation unit for generating the virtual cluster further in consideration of the configuration of the virtual cluster.

Korea Patent Registration No. 10-1781063 (2017.09.18) Korea Patent Publication No. 10-2017-0088096 (2017.08.01)

An embodiment of the present invention is to provide an apparatus and method for processing a resource for integrated resource management that can efficiently share a variety of resources including a graphics processing unit (GPU).

An embodiment of the present invention is to provide an apparatus and method for processing resources for integrated resource management that can improve resource sharing and isolation efficiency through a resource container optimized for allocation resources.

An embodiment of the present invention is to provide a method and method for processing resources integrated resource management that can use the GPU resources and use and manage some or all of the GPU resources of each node.

Among the embodiments, an apparatus for processing resources for integrated resource management may include a computing resource unit including at least one CPU element, at least one GPU element, and at least one memory element as a resource, and a resource allocation request for the computing resource unit. A MESOS Framework Unit (MESOS Framework Unit) for performing task scheduling to generate task information, requesting allocation of resources in the computing resource unit, and at least some of the requested allocation resources (hereinafter, allocating resources). Is successfully allocated and the corresponding resource information is received when a request is made to allocate the resource from a Mesos slave unit (MESOS Slave Unit) and the mesos slave unit performing the task through a docker executor dedicated to the allocated resource. Resource as per policy A Mesos Master Unit (MESOS Master Unit) for determining allocation and notifying the MesoS framework unit to control resource sharing among tasks, wherein the Meso Slave unit is assigned by the Docker performer dedicated to the allocated resource. A lifecycle of the resource container can be managed by instantiating a resource container that can use at least one specific resource among the resources.

The method slave unit may create a docker performer dedicated to the allocated resource as an independent thread in the performer.

delete

The method slave unit may determine whether the resource container is specified by the corresponding task information and all virtual resources associated with the dynamic use of the allocated resource are used in the life cycle of the resource container.

The meso slave unit may request the meso master unit to increase additional resources other than the allocated resources when the virtual resources are used up.

The mesos slave unit may determine an additional resource increase when the usage ratio of the virtual resource exceeds a specific ratio and dynamically adjust the specific ratio based on the following equation.

[Equation]

(Where r represents the specific ratio, r ₀ represents a reference ratio set by a designer or a manager as a prime number greater than 0 and less than 1 with a digit value less than 1 digit, and n _call is Represents the total number of additional resource increase requests generated during the lifecycle process)

If the resource container causes an allocation violation of a virtual resource in the lifecycle of the resource container, the method slave unit may process a system call dedicated to the allocation resource associated with the allocation violation, by a docker performer dedicated to the allocation resource. have.

The meso slave unit may allow the resource container to directly control the resource element without involvement of the meso master unit in a system call processing process dedicated to the allocated resource.

Among the embodiments, the method for processing a method for integrated resource management, the computing resource unit preparing at least one CPU element, at least one GPU element and at least one memory element as a resource, the Meso framework framework unit is Generating task information by performing resource scheduling according to a resource allocation request for a computing resource unit, the method slave unit requesting an allocation of a resource in the computing resource unit, and performing at least some of the requested allocation resources (hereinafter, Performing a task through a docker executor dedicated to the allocated resource when the allocated resource) is successfully allocated and the corresponding task information is received, and when the resource master unit is requested to allocate the resource from the resource slave unit, According to the resource policy And a step of notification by the mail source framework unit determines the allocation of the resource control of the resources shared among the tasks.

Among the embodiments, the recording medium is computer readable and records a program for executing a method of processing a method for integrated resource management on a computer. In the recording medium, the method of processing a method for integrated resource management, the computing resource unit comprises the steps of preparing at least one CPU element, at least one GPU element and at least one memory element as a resource, Meso framework unit Generating task information by performing resource scheduling according to the resource allocation request for the computing resource unit, the method slave unit requesting an allocation of a resource in the computing resource unit, and at least a part of the allocation requested resource Performing a task through a docker executor dedicated to the allocated resource and requesting an allocation of the resource from the meso slave unit Received if given Depending on the policy, it determines the allocation of the resources to notify the mail source framework unit and a step of controlling the resources shared among the tasks.

The disclosed technique can have the following effects. However, since a specific embodiment does not mean to include all of the following effects or only the following effects, it should not be understood that the scope of the disclosed technology is limited by this.

The method and apparatus for processing resources for integrated resource management according to an embodiment of the present invention can efficiently share various resources including a graphic processing unit (GPU).

The method and apparatus for processing resources for integrated resource management according to an embodiment of the present invention can improve resource sharing and isolation efficiency through a resource container optimized for allocation resources.

The method and apparatus for processing resources for integrated resource management according to an embodiment of the present invention may use GPU resources and use and manage some or all GPU resources of each node.

FIG. 1 is a diagram illustrating an apparatus for processing a method for integrated resource management according to an exemplary embodiment of the present invention.
FIG. 2 is a diagram illustrating an example of a method of processing a method for resource integrated management in FIG. 1 in more detail.
FIG. 3 is a diagram for describing a structure and a logical execution structure of a docker performer dedicated to allocation resources in FIG. 2.
FIG. 4 is a flowchart illustrating a process of performing resource integrated management by a method source processing apparatus for resource integrated management in FIG. 1.

Description of the present invention is only an embodiment for structural or functional description, the scope of the present invention should not be construed as limited by the embodiments described in the text. That is, since the embodiments may be variously modified and may have various forms, the scope of the present invention should be understood to include equivalents capable of realizing the technical idea. In addition, the objects or effects presented in the present invention does not mean that a specific embodiment should include all or only such effects, the scope of the present invention should not be understood as being limited thereby.

On the other hand, the meaning of the terms described in the present application should be understood as follows.

Terms such as "first" and "second" are intended to distinguish one component from another component, and the scope of rights should not be limited by these terms. For example, the first component may be named a second component, and similarly, the second component may also be named a first component.

When a component is referred to as being "connected" to another component, it should be understood that there may be other components in between, although it may be directly connected to the other component. On the other hand, when a component is referred to as being "directly connected" to another component, it should be understood that there is no other component in between. On the other hand, other expressions describing the relationship between the components, such as "between" and "immediately between" or "neighboring to" and "directly neighboring", should be interpreted as well.

Singular expressions should be understood to include plural expressions unless the context clearly indicates otherwise, and terms such as "comprise" or "have" refer to a feature, number, step, operation, component, part or implementation thereof. It is to be understood that the combination is intended to be present and does not exclude in advance the possibility of the presence or addition of one or more other features or numbers, steps, operations, components, parts or combinations thereof.

In each step, an identification code (e.g., a, b, c, etc.) is used for convenience of description, and the identification code does not describe the order of the steps, and each step clearly indicates a specific order in context. Unless stated otherwise, they may occur out of the order noted. That is, each step may occur in the same order as specified, may be performed substantially simultaneously, or may be performed in the reverse order.

The present invention can be embodied as computer readable code on a computer readable recording medium, and the computer readable recording medium includes all kinds of recording devices in which data can be read by a computer system. . Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, and the like. The computer readable recording medium can also be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion.

All terms used herein have the same meaning as commonly understood by one of ordinary skill in the art unless otherwise defined. Generally, the terms defined in the dictionary used are to be interpreted to coincide with the meanings in the context of the related art, and should not be interpreted as having ideal or excessively formal meanings unless clearly defined in the present application.

FIG. 1 is a diagram illustrating an apparatus for processing a method for integrated resource management according to an exemplary embodiment of the present invention.

Referring to FIG. 1, the mesos processing apparatus 100 for integrated resource management includes a computing resource unit 110, a mesos framework unit 120, and a mesos slave unit (MESOS Slave Unit) ( 130 and a MESOS Master Unit 140.

Computing resource unit 110 includes at least one CPU element 112, at least one GPU element 114, and at least one memory element 116 as a resource.

Each of the at least one central processing unit (CPU) element 112 is a processor capable of independently processing and performing a task associated with the corresponding meso slave unit 130 when allocated to a particular meso slave unit 130 as a resource. Can be implemented.

Each of the at least one graphic processing unit (GPU) element 114 is a resource, and in one embodiment, when assigned to a specific method slave unit 130, a process of performing a task associated with the method method slave unit 130. It can be implemented as a processor dedicated to graphics operations associated with the task in the.

Each of the at least one memory element 116 may be used to store overall data generated or necessary in the course of performing a task associated with the corresponding method slave unit 130 when allocated to a specific method slave unit 130 as a resource. The memory device may be implemented through at least one of a nonvolatile memory that may function as an auxiliary memory device and a main memory device that may function as a main memory device.

The meso framework framework unit 120 performs resource scheduling according to the resource allocation request for the computing resource unit 110 to generate task information. In one embodiment, the Mesos framework unit 120 is associated with a particular Mesos slave unit 130 from the Mesos master unit 140 and is a computing resource unit 110 executable in the Mesos framework unit 120. ) May receive a resource allocation request including a resource request amount of at least a portion, and, for example, may receive a resource allocation request including a method slave unit identifier and a resource-specific request quota. In one embodiment, the resource allocation request may further include a resource allocation reservation time and a minimum resource allocation request amount designated in a specific time interval.

The meso framework framework 120 may perform resource scheduling to generate at least one task information to be executed and necessary resource information for each task, based on the received resource allocation request. The meso framework framework unit 120 may transmit, to the meso master unit 140, task information and corresponding necessary resource information generated through resource scheduling satisfying at least some conditions of the received resource allocation request.

In one embodiment, the MesoS framework unit 120, upon receipt of a resource allocation request, receives all available resource information (e.g., 20 CPUs, 20 GPUs, and 200GB of memory) and a current resource usage situation (e.g., Analyzing available resource information (for example, 15 current CPUs, 17 GPUs, and 150GB new allocation of memory) by analyzing five CPUs, three GPUs, and 50GB of shared memory, and detected detected resources Resource scheduling for resource allocation corresponding to the resource allocation request may be performed based on the information.

The mesos slave unit 130 may request allocation of resources in the computing resource unit 110. More specifically, the mesos slave unit 130 transmits resource information (for example, five CPUs and 20 GB of memory) to be used by the mesos master unit 140 to determine the resource corresponding to the resource information. You can request an assignment. In one embodiment, the mesos slave unit 130 may request resource allocation from one of the plurality of mesos master units 140 in the standby state.

When at least some of the requested resources (hereinafter, allocated resources) are successfully allocated and the corresponding task information is received, the method slave unit 130 performs the task through a docker executor 212 dedicated to the allocated resources. Perform. This will be described in more detail with reference to FIG. 2.

FIG. 2 is a diagram illustrating an example of a method of processing a method for resource integrated management in FIG. 1 in more detail.

In FIG. 2, the mesos slave unit 130 may receive resource scheduling information generated by the mesos framework unit 120 from the mesos master unit 140, and may include at least one of the resource scheduling information. Tasks can be run through the Mesos agent.

Referring to FIG. 2, the Mesos slave unit 130 includes an executor 210 for executing each Mesos agent, and the executor 210 includes a docker executor 212 and a mesos dedicated to allocated resources. It may include a MESOS executor 214 and a docker executor 216.

In one embodiment, the docker performer 212 dedicated to allocated resources may be performed using Nvidia-docker, for example, using the Mesos execution module by default and building a resource container environment under the Mesos container environment. Can be. In one embodiment, the Mesos executor 214 may be performed using a cgroup, and the docker executor 216 may be performed using a docker.

The docker performer 212 dedicated to the allocated resources may be launched on the Mesos slave unit 130 under the control of the Mesos master unit 140 and sequentially perform tasks in order of the scheduled task information based on the received resource scheduling information. have.

The Mesos slave unit 130 may create a docker performer 212 dedicated to allocated resources as an independent thread in the performer 210, for example, in the performer 210, the mesos performer 214 and the docker performer ( 216 can be created as a thread independent.

In the method slave unit 130, a docker performer 212 dedicated to allocated resources may instantiate a resource container 210 to manage a life cycle of the resource container 310. This will be described in more detail with reference to FIG. 3.

FIG. 3 is a diagram for describing a structure and a logical execution structure of a docker performer dedicated to allocation resources in FIG. 2.

Referring to FIG. 3, the docker performer 212 dedicated to allocated resources may include a method source module, a command line interface (CLI) generation module, and a docker perform module dedicated to allocated resources.

The docker performer 212 dedicated to the allocated resources may create a resource container 310 that can use at least one specific resource (eg, GPU) of the allocated resources, and instantiate the generated resource container 310 to The life cycle of the resource container 310 may be managed. In one embodiment, the docker performer 212 dedicated to the allocated resources instantiates at least one resource container 310 according to the scheduling of the corresponding resource to a specific agent node during the execution of the Mesos agent and through the corresponding resource container 310. An application may be executed, and the resource container 310 may be managed through a lifecycle process including run, kill, and monitoring. For example, the docker performer 212 dedicated to allocation resources may allocate the first application through the first resource container 310a if the first and second tasks and resources according to the task are allocated according to the requested resource allocation request. After the execution, the performance of the first task may be monitored, and if the execution is completed, the first resource container 310a may be killed, and then the life cycle process of running the second application may be controlled through the second resource container 310b. have.

The method slave unit 130 may check whether the resource container 310 is specified by the task information and all virtual resources associated with the dynamic use of the allocated resource are used during the life cycle of the resource container 310. For example, the Mesos slave unit 130 requested resource allocation of 10 CPU elements 112, 10 GPU elements 114, and 100 GB of memory elements 116 for the task, but in fact the order was 7 , 7, and 70 GB of resources, if less than the actual request amount, can be achieved by mapping between the actual allocated resources (7, 7, and 70 GB) and virtual resources (10, 10, and 100 GB). , Three, and 30 GB resources may be mapped to at least some of the allocated resources, and the meso master unit 140 may determine whether all of the corresponding virtual resources are used. In this embodiment, the Mesos framework unit 120 may perform mapping between the virtual resource and the actual allocated resource, and the Mesos master unit 140 may receive and manage the mapping result.

The meso slave unit 130 may request the meso master unit 140 for additional resource increase other than the corresponding allocated resource when all the virtual resources are used. For example, the mesos slave unit 130 may determine that additional resources are needed when all virtual resources mapped with actual allocation resources are used.

In one embodiment, the Mesos slave unit 130 may request the Mesos master unit 140 for additional resource increase if the usage rate of the corresponding virtual resource exceeds a certain rate, and may be specified based on Equation 1 below. The ratio can be adjusted dynamically.

[Equation 1]

(Where r represents a specific ratio, r ₀ is a prime number greater than 0 and less than 1 with a place value less than 1 place, representing a reference rate set by the designer or administrator (eg, 0.9), n _call represents the total number of additional resource augmentation requests that occurred during the lifecycle)

If the resource container 310 causes an allocation violation of a virtual resource during the lifecycle of the resource container 310, the method slave unit 130 allocates the assigned resource associated with the docker performer 212 dedicated to the allocation violation. Dedicated system calls can be handled. For example, the Mesos slave unit 130 is assigned with seven CPU elements 112, seven GPU elements 114, and 70 GB of memory elements 116 as resources, and is not associated with the resource above the reference value. If an attempt is made to access another resource that has not been accessed, it may be determined that an allocation violation of the virtual resource has been caused, and that it is recognized as a system (eg, a new graphical operation) call request dedicated to the allocated resource for the execution of a new process or task. Can be.

The mesos slave unit 130 may allow the resource container 310 to directly control the resource element without the involvement of the mesos master unit 140 in the system call processing dedicated to the allocated resource.

For example, the MesoSlave unit 130 is a resource element (e.g., a CPU) associated with the violation of the allocation based on the temporary direct control authority acquired according to a system call dedicated to the allocation resource. Limited direct control of element 110 may be performed to augment additional resources within a particular predetermined reference range.

For another example, the Mesos slave unit 130 transmits whether the allocation of a virtual resource for a specific resource element is violated to the Mesos master unit 140 during the call of the system dedicated to the allocated resource, and thus the Mesos master unit 140. ) May request direct control by the resource container 310 on the resource element, and the resource container 310 performs non-limiting direct control on the resource element associated with the allocation violation according to the permission. This allows additional resources to be increased within a specific preset range.

When the mesos master unit 140 receives a request for resource allocation from the mesos slave unit 130, it determines the resource allocation according to a given resource policy and notifies the mesoframe framework unit 120 to share resources among tasks. To control. In an embodiment, when the resource allocation request is received, the method of master resource 140 may determine and transmit a request quota for each resource to be requested to the method framework unit 120 based on a previously stored resource policy. According to the resource scheduling and task information received from the MesoS framework unit 120, the node based on the allocated resource based on the node of the MesoSlave unit 130 is executed by the MesoSlave unit 130. You can control the sharing of resources between applications.

In one embodiment, the resource policy is a first resource policy for equally allocating a quota of resources based on the amount of available resources for each MesoS framework unit, a second resource policy for precisely allocating a predetermined reference quota, and at a specific interval. The minimum and maximum quotas are determined based on the third resource policy for determining the minimum and maximum quotas based on the total resource quotas updated and dynamically allocated to the quotas between them, and the minimum and maximum quotas based on the total resource requests updated at specific intervals. And a fourth resource policy for dynamic allocation. In one embodiment, the resource policy may be updated when a new resource policy is received through the policy update module of the Mesos master unit 140 that receives the new resource policy from the outside.

If only some of the requests for allocation of resources requested from the Mesos slave unit 130 through the particular Mesos framework unit 120 are met, the mesos master unit 140 may have at least one request until all of the requests are satisfied. It is possible to repeat the attempt to meet the rest via another MesoS framework unit 120. In one embodiment, the Mesos master unit 140 is the longest of the Mesos framework units 120 in this process or the top N (N is a natural number) associated with the Mesos framework unit 120 the highest. The standby unit of may be detected and an attempt may be made for remaining resource allocation.

The computing resource unit 110, the mesos framework unit 120, the mesos slave unit 130, and the mesos master unit 140 may each be implemented as a physical computing device, at least some of which may be implemented as servers. Can function.

FIG. 4 is a flowchart illustrating a process of performing resource integrated management by a method source processing apparatus for resource integrated management in FIG. 1.

In FIG. 4, the method source processing apparatus 100 for resource integrated management prepares at least one CPU element 112, at least one GPU element 114, and at least one memory element 116 as resources (step 1). S410). The meso framework framework unit 120 performs resource scheduling according to the resource allocation request for the computing resource unit 110 to generate task information (step S420). The method slave unit 130 may request allocation of resources in the computing resource unit 110 (step S430). If the allocated resource is successfully allocated and the corresponding task information is received, the method slave unit 130 performs the corresponding task through the docker performer 212 dedicated to the allocated resource (step S440). When the mesos master unit 140 receives a request for resource allocation from the mesos slave unit 130, it determines the resource allocation according to a given resource policy and notifies the mesoframe framework unit 120 to share resources among tasks. Control (step S450).

Although described above with reference to a preferred embodiment of the present application, those skilled in the art various modifications of the present application without departing from the spirit and scope of the invention described in the claims below And can be changed.

100: Mesos processing unit for integrated resource management
110: computing resource unit
120: Mesos Framework Unit
130: Mesos slave unit
140: Mesos Master Unit

Claims (10)

A computing resource unit including at least one CPU element, at least one GPU element and at least one memory element as a resource;
A MESOS Framework Unit for generating task information by performing resource scheduling according to a resource allocation request for the computing resource unit;
If a request for allocation of resources in the computing resource unit is requested and at least some of the requested resources (hereinafter, allocated resources) are successfully allocated and the corresponding task information is received, the task is performed through a docker performer dedicated to the allocated resources. A mesos slave unit (MESOS Slave Unit) to perform; And
MESOS Master Unit (MESOS Master Unit) to control the sharing of resources among tasks by determining the allocation of the resource according to a given resource policy and notifying the MesoS framework unit when the allocation of the resource is requested from the MesoSlave unit. ),
The method slave unit may manage a lifecycle of the resource container by instantiating a resource container that the docker performer dedicated to the allocated resource can use at least one specific resource among the allocated resources. Mesos processing unit.
The method of claim 1, wherein the Mesos slave unit
And a docker executor dedicated to the allocated resources as an independent thread in the executor.
delete The method of claim 1, wherein the Mesos slave unit
During the lifecycle management of the resource container, the resource container is identified by the corresponding task information and checks whether all of the virtual resources associated with the dynamic use of the allocated resource are used. Processing unit.
The method of claim 4, wherein the Mesos slave unit
And when the virtual resources are used up, requesting additional resource reinforcement other than the allocated resources to the mesos master unit.
The method of claim 4, wherein the Mesos slave unit
And determining an additional resource increase if the usage ratio of the virtual resource exceeds a specific ratio, and dynamically adjusting the specific ratio based on the following equation.
[Equation]

(Where r represents the specific ratio, r ₀ represents a reference ratio set by a designer or a manager as a prime number greater than 0 and less than 1 with a digit value less than 1 digit, and n _call is Represents the total number of additional resource increase requests generated during the lifecycle management process)
The method of claim 1, wherein the Mesos slave unit
When the resource container causes an allocation violation of a virtual resource in the lifecycle management of the resource container, a docker performer dedicated to the allocated resource handles a system call dedicated to the allocated resource associated with the allocation violation. Mesos processing unit for management.
The method of claim 7, wherein the Mesos slave unit
And the resource container allows direct control of the corresponding resource element without involvement of the method master unit in the system call processing process dedicated to the allocated resource.
The computing resource unit preparing at least one CPU element, at least one GPU element and at least one memory element as resources;
Generating, by the method framework unit, resource scheduling according to a resource allocation request for the computing resource unit to generate task information;
A Messock slave unit requests allocation of resources in the computing resource unit, and at least some of the requested resources (hereinafter, allocated resources) have been successfully allocated and the corresponding task information is received, the docker performer dedicated to the allocated resources. Performing the task through; And
If a MesoS master unit is requested to allocate the resource from the MesoSlave unit, determining the allocation of the resource according to a given resource policy and notifying the MesoS framework unit to control resource sharing among tasks. and,
The step of performing the task
And a docker executor dedicated to the allocated resources instantiates a resource container capable of using at least one specific resource among the allocated resources to manage a life cycle of the resource container.
A computer-readable recording medium having recorded thereon a program for executing a method of processing a method for integrated resource management on a computer,
Mesos processing method for the integrated resource management
The computing resource unit preparing at least one CPU element, at least one GPU element and at least one memory element as resources;
Generating, by the method framework unit, resource scheduling according to a resource allocation request for the computing resource unit to generate task information;
If a method slave unit requests an allocation of resources in the computing resource unit, and at least some of the requested allocation resources (hereinafter, allocated resources) are successfully allocated and corresponding task information is received, a docker performer dedicated to the allocated resources Performing the task through; And
If the method of the Meso master unit is requested to allocate the resource from the Meso slave unit, determining the allocation of the resource according to a given resource policy and notifying the Meso framework framework unit to control resource sharing between tasks and,
The step of performing the task
And a docker performer dedicated to the allocated resources manages the life cycle of the resource container by instantiating a resource container capable of using at least one specific resource among the allocated resources.

Images