KR102688689B1 - Management apparatus for virtual resource, and control method thereof - Google Patents

Abstract

The present invention unifies the form of virtual infrastructure deployed in the under cloud with the over cloud when implementing a managed Kubernetes service based on Mobile-Edge Computing (MEC). It relates to a virtual resource operation device and a method of operating the virtual resource operation device for management.

Description

Virtual resource management device and operation method of the virtual resource management device {MANAGEMENT APPARATUS FOR VIRTUAL RESOURCE, AND CONTROL METHOD THEREOF}

The present invention relates the form of virtual infrastructure deployed in the under cloud to the over cloud and over cloud when implementing a managed Kubernetes service based on Mobile-Edge Computing (MEC). It is about a plan for unified management.

Serverless computing is a form of service in which all elements, excluding applications, are provided by a cloud provider. Except for applications, physical infrastructure, virtual machines, containers, and integrated management areas are provided by the cloud provider, and users can manage the applications.

Among these, containers are a much lighter virtualization technology than virtual machines. They support configuring all related elements (related libraries, configuration files, etc.) for application execution into one package, shortening the creation and deployment time of containers. It not only has the advantage of being able to restart quickly.

Meanwhile, most services provided in the cloud environment are composed of a service architecture based on Kubernetes, an open source system that automatically deploys, scales, and manages containerized applications.

Therefore, support for Kubernetes is also essential in the Mobile-Edge Computing (MEC) environment, which is a concept that seeks to provide various IT services and technologies provided in the existing cloud environment at the edge of the mobile network.

In this regard, the existing managed Kubernetes service creates a virtual machine (VM) through a hypervisor installed in bare metal (BM), and Kubernetes It is provided by installing the cluster on the corresponding virtual machine.

However, in the existing managed Kubernetes service, under cloud is composed of bare metal for building virtual infrastructure, whereas in over cloud, Kubernetes cluster is composed of containers. , it is difficult to monitor and control various types of resources from one management panel.

The present invention was created in consideration of the above circumstances, and the purpose to be achieved by the present invention is to implement under cloud (Under Cloud) service when implementing managed Kubernetes service based on Mobile-Edge Computing (MEC). The goal is to manage the form of virtual infrastructure deployed in the cloud by unifying it with the over cloud.

A virtual resource operation device according to an embodiment of the present invention for achieving the above object includes a virtual infrastructure provision unit that creates a virtual machine for the workload Kubernetes using a containerized hypervisor; and a cluster provider that builds a cluster of the workload Kubernetes in a form including a master node and worker nodes matching the master node within the virtual machine.

Specifically, the virtual resource management device may further include a bootstrap provider that connects a worker node to the master node and activates the workload Kubernetes in a serviceable state.

Specifically, the virtual infrastructure provider may configure the containerized hypervisor in the form of a pod (POD) and create the virtual machine within the pod using the hypervisor.

Specifically, the cluster includes two or more master nodes according to predefined specifications for the workload Kubernetes, and the cluster provider, a load balancer for the two or more master nodes, and the workload Kubernetes Additional firewalls can be created and placed in front of the cluster.

Specifically, the virtual resource management device may further include a cluster status checker that periodically checks whether the workload is available for Kubernetes using a heart beat message transmitted to the load balancer. there is.

A method of operating a virtual resource management device according to an embodiment of the present invention for achieving the above object includes a virtual infrastructure provision step of creating a virtual machine for a workload Kubernetes using a containerized hypervisor; And a cluster provision step of constructing a cluster of the workload Kubernetes in a form including a master node and a worker node matching the master node within the virtual machine.

Specifically, the method may further include a bootstrap providing step of activating the workload Kubernetes to a serviceable state by joining the worker node to the master node.

Specifically, in the virtual infrastructure provision step, the containerized hypervisor may be configured in the form of a pod (POD), and the virtual machine may be created within the pod using the hypervisor.

Specifically, the cluster includes two or more master nodes according to predefined specifications for the workload Kubernetes, and the cluster provision step includes a load balancer for two or more master nodes and the workload Kubernetes. You can create an additional firewall for TS and place it in front of the cluster.

Specifically, the method may further include a cluster status check step of periodically checking whether the workload Kubernetes is available using a heart beat message transmitted to the load balancer.

Accordingly, according to the virtual resource operation device and the operation method of the virtual resource operation device of the present invention, when implementing a managed Kubernetes service based on Mobile-Edge Computing (MEC), under cloud )'s virtual infrastructure is unified and managed in the form of a container, similar to the over cloud workload Kubernetes, enabling E2E (End To End) management using Kubernetes' closed loop control. Let's do it.

Figure 1 is an example diagram illustrating a managed Kubernetes service environment based on mobile edge computing according to an embodiment of the present invention.
Figure 2 is a schematic configuration diagram of a virtual resource management device according to an embodiment of the present invention.
Figure 3 is an example diagram for explaining the arrangement of a virtual machine according to an embodiment of the present invention.
Figure 4 is a schematic configuration diagram of a virtual infrastructure provision unit according to an embodiment of the present invention.
Figure 5 is a flowchart illustrating a method of operating a virtual resource management device according to an embodiment of the present invention.

Hereinafter, preferred embodiments of the present invention will be described with reference to the attached drawings.

In one embodiment of the present invention, virtualization technology based on Mobile-Edge Computing (MEC) is addressed.

Most services provided in the cloud environment are composed of a service architecture based on Kubernetes, an open source system that automatically deploys, scales, and manages containerized applications.

Therefore, support for Kubernetes is also essential in the mobile edge computing environment, which is a concept that seeks to provide various IT services and technologies provided in the existing cloud environment at the edge of the mobile network.

In this regard, the existing managed Kubernetes service creates a virtual machine (VM) through a hypervisor installed in bare metal (BM), and Kubernetes It is provided by installing the cluster on the corresponding virtual machine.

However, in the existing managed Kubernetes service, under cloud is composed of bare metal for building virtual infrastructure, whereas in over cloud, Kubernetes cluster is composed of containers. , it is difficult to monitor and control various types of resources from one management panel.

Accordingly, in one embodiment of the present invention, when implementing a managed Kubernetes service based on mobile edge computing, we propose a method of unifying and managing the virtual infrastructure of the under cloud in the same form as the over cloud.

In this regard, Figure 1 shows a managed Kubernetes service environment based on mobile edge computing according to an embodiment of the present invention.

As shown in Figure 1, in the mobile edge computing-based managed Kubernetes service environment according to an embodiment of the present invention, the workload Kubernetes (Workload kubernetes) from the mobile edge computing orchestrator (10, MEO) ) According to the creation request, the Mobile Edge Platform Manager (30, MEPM) is called, and the Mobile Edge Platform Manager (30) again calls the Virtualization Infrastructure Manager (40, VIM).

Accordingly, the virtualization infrastructure manager 40 calls the virtual resource operation device 50 through the MM6 interface, and the virtual resource operation device 50 responds to the creation, deletion, and deletion of workload Kubernetes. It manages the entire life cycle of Kubernetes workloads, including updates and updates.

Here, workload Kubernetes is managed Kubernetes assigned to a specific tenant and can be generally understood as an application built using a virtual machine.

In this regard, the virtual resource management device 50 can be understood as a component that builds the Kubernetes cloud in the over-cloud and operates the under-cloud, which consists of core components necessary to set up and manage the over-cloud.

Meanwhile, the mobile edge computing-based managed Kubernetes service environment according to an embodiment of the present invention includes, in addition to the above-described configuration, essential functions for running the mobile edge computing application in a virtualized infrastructure, which is called the mobile edge computing service. It may further include a mobile edge computing platform (20, MEP) provided to (MEC service).

As above, in the managed Kubernetes service environment based on mobile edge computing according to an embodiment of the present invention, the virtual infrastructure of the under cloud can be unified and managed in the same form as the over cloud based on the above-described configuration. In this section, the virtual resource operation device 50 for realizing this will be described in more detail.

Figure 2 schematically shows the configuration of a virtual resource operation device (50, Cluster Operator) according to an embodiment of the present invention.

As shown in FIG. 2, the virtual resource operation device 50 according to an embodiment of the present invention includes a virtual infrastructure provider (51) that creates virtual resources belonging to the virtual infrastructure category, and a workload cue. It may have a configuration that includes a cluster provider (52) that creates a Vernetis cluster.

In addition, the virtual resource operation device 50 according to an embodiment of the present invention, in addition to the above-mentioned configuration, includes a bootstrap provider 53 (Bootstrap Provider) that processes clustering for each node in the workload Kubernetes cluster; It may have a configuration that further includes a cluster health checker (54) that verifies normal operation of the workload Kubernetes.

The entire configuration or at least a portion of the virtual resource management device 50 may be implemented in the form of a hardware module or a software module, or may be implemented in a combination of a hardware module and a software module.

Here, the software module can be understood as, for example, a component for building a Kubernetes cloud on an overcloud and operating an undercloud composed of core components necessary for setting up and managing the overcloud.

The virtual resource operation device 50 according to an embodiment of the present invention can manage the virtual infrastructure of the under cloud in the same manner as the over cloud through the above-described configuration. Hereinafter, the virtual resource operation to realize this will be described. We will continue with a more detailed description of each component within the device 50.

The virtual infrastructure provision unit 51 performs the function of creating virtual resources for the workload Kubernetes.

More specifically, the virtual infrastructure provider 51 creates a virtual machine for Workload Kubernetes using a containerized hypervisor.

At this time, the virtual infrastructure provision unit 51 uploads the containerized hypervisor in the form of a pod (POD) using System Kubernetes, for example, as shown in Figure 3, and installs the workload within the pod. You can recreate a virtual machine to run Kubernetes.

As such, the method of creating a virtual machine using a containerized hypervisor in one embodiment of the present invention is significantly different from the existing method of creating a virtual machine by directly calling the hypervisor installed in bare metal (BM). You can confirm that they are separated.

However, in order for Workload Kubernetes to operate normally, a virtual machine is essential, creating a virtual network, a virtual disk to contain Workload Kubernetes, a key pair for accessing Workload Kubernetes, and workload Kubernetes requires pre-built virtual image resources.

Accordingly, the virtual infrastructure provision department 51 is a virtual infrastructure to support the normal operation of workload Kubernetes, including not only virtual machines for workload Kubernetes, but also virtual networks, virtual disks, key pairs, and virtual images. Create additional resources.

Meanwhile, in order to build such a virtual infrastructure, the virtual infrastructure provision unit 51, for example, provides a virtual machine management module (Virtual Machine Manager) for creating a virtual machine as shown in FIG. 4, and virtual network management for creating a virtual network. module (Virtual Network Manager), volume management module (Volume Manager) for creating virtual disks, key pair management module (Keypair Manager) for creating key pairs, and image management for creating virtual image resources. May include a module (Image Manager).

The cluster provider 52 performs the function of building a workload Kubernetes cluster.

More specifically, when a virtual machine is created for workload Kubernetes, the cluster provider 52 builds a cluster of workload Kubernetes that follows predefined specifications for workload Kubernetes within the virtual machine. do.

Here, the workload Kubernetes cluster built within a virtual machine includes a master node that controls the entire workload Kubernetes cluster and several worker nodes that receive commands from this master node and generate and service the actual workload. May be included.

In particular, in the case of master nodes within a workload Kubernetes cluster, two or more can be deployed for load balancing purposes.

In other words, in the case of a worker-loaded Kubernetes cluster according to an embodiment of the present invention, it can be constructed to include two or more master nodes and several worker nodes joining each of these two or more master nodes.

In this regard, the cluster provider 52 adds a load balancer for two or more master nodes and a firewall for the workload Kubernetes on a cluster basis to support HA (High Availability) of the workload Kubernetes. You can create it and place it in front of the cluster.

The bootstrap provider 53 performs the function of activating the workload Kubernetes into a serviceable state.

More specifically, when a cluster of workload Kubernetes is established within a virtual machine, the bootstrap provider 53 joins a worker node to the master node of the cluster to enable service of workload Kubernetes. It is activated with .

At this time, the bootstrap provider 53 connects to the master node using the authentication key (key pair) generated by the key pair management module (Keypair Manager) under the assumption that the master node is available and then connects to the Kubernetes within the master node. Executes all Teeth components.

Here, Kubernetes components within the master node may include a kublet, container runtime, container network interface (CNI), controller, and scheduler.

Furthermore, when all Kubernetes components in the master node are started and can receive external joins, the bootstrap provider 53 connects to each worker node and authenticates the worker node with an authentication key (key pair). ), the workload Kubernetes can be activated in a serviceable state by connecting to the master node.

The cluster status check unit 54 performs the function of checking the availability status of the workload Kubernetes.

More specifically, the cluster status check unit 54 periodically checks whether the workload Kubernetes is available using a heart beat message.

At this time, the cluster status check unit 54 periodically transmits a heartbeat message to the load balancer placed in front of the workload Kubernetes cluster, and determines whether or not it responds to the workload Kubernetes. You can check whether it is available or not.

In this regard, the cluster status check unit 54 may check the unavailability status of the workload Kubernetes and generate an alarm if a response to the heart beat message is not received. Of course.

As discussed above, according to the configuration of the virtual resource operation device 50 according to an embodiment of the present invention, when implementing a managed Kubernetes service based on Mobile-Edge Computing (MEC) By managing the under cloud's virtual infrastructure in a unified container form, similar to the over cloud's workload Kubernetes, E2E (End To End) It can be seen that management is possible.

Hereinafter, a method of operating the virtual resource management device 50 according to an embodiment of the present invention will be described with reference to FIG. 5.

First, the virtual infrastructure provision unit 51 creates a virtual machine for Workload Kubernetes using a containerized hypervisor (S10-S20).

At this time, the virtual infrastructure provision unit 51 uploads the containerized hypervisor in the form of a pod (POD) using System Kubernetes, as shown in FIG. 3, and installs it within the pod. You can recreate a virtual machine to run the workload Kubernetes.

As such, the method of creating a virtual machine using a containerized hypervisor in one embodiment of the present invention is significantly different from the existing method of creating a virtual machine by directly calling the hypervisor installed in bare metal (BM). You can confirm that they are separated.

However, in order for Workload Kubernetes to operate normally, a virtual machine is essential, creating a virtual network, a virtual disk to contain Workload Kubernetes, a key pair for accessing Workload Kubernetes, and workload Kubernetes requires pre-built virtual image resources.

Accordingly, the virtual infrastructure provision unit 51 creates a virtual machine through a virtual machine management module (Virtual Machine Manager), as shown in Figure 4 above, to support the normal operation of the workload Kubernetes, and also creates a virtual machine A virtual network is created through the network management module (Virtual Network Manager), a virtual disk is created through the volume management module (Volume Manager), and a key pair is created through the key pair management module (Keypair Manager), and Create virtual image resources through the image management module (Image Manager).

Then, when a virtual machine is created for workload Kubernetes, the cluster provider 52 builds a cluster of workload Kubernetes that follows predefined specifications for workload Kubernetes within the virtual machine ( S30).

Here, the workload Kubernetes cluster built within a virtual machine includes a master node that controls the entire workload Kubernetes cluster and several worker nodes that receive commands from this master node and generate and service the actual workload. May be included.

In particular, in the case of master nodes within a workload Kubernetes cluster, two or more can be deployed for load balancing purposes.

In other words, in the case of a worker-loaded Kubernetes cluster according to an embodiment of the present invention, it can be constructed to include two or more master nodes and several worker nodes joining each of these two or more master nodes.

In this regard, the cluster provider 52 adds a load balancer for two or more master nodes and a firewall for the workload Kubernetes on a cluster basis to support HA (High Availability) of the workload Kubernetes. Create it and place it in front of the cluster (S40).

Furthermore, the cluster status check unit 54 periodically checks whether the workload Kubernetes is available using a heart beat message (S50).

At this time, the cluster status check unit 54 periodically transmits a heartbeat message to the load balancer placed in front of the workload Kubernetes cluster, and determines whether or not it responds to the workload Kubernetes. You can check whether it is available or not.

In this regard, the cluster status check unit 54 may check the unavailability status of the workload Kubernetes and generate an alarm if a response to the heart beat message is not received. Of course (S60-S70).

Afterwards, when a workload Kubernetes cluster is built in a virtual machine under the assumption that the master node is available, the bootstrap provider 53 joins the worker node to the master node of the cluster to load the workload. Activate Kubernetes into a serviceable state (S80-S90).

At this time, the bootstrap provider 53 connects to the master node using the authentication key (key pair) generated by the key pair management module (Keypair Manager) and executes all Kubernetes components in the master node.

Here, Kubernetes components within the master node may include a kublet, container runtime, container network interface (CNI), controller, and scheduler.

Furthermore, when all Kubernetes components in the master node are started and can receive external joins, the bootstrap provider 53 connects to each worker node and authenticates the worker node with an authentication key (key pair). ), the workload Kubernetes can be activated in a serviceable state by connecting to the master node.

As discussed above, according to the operation method of the virtual resource operation device 50 according to an embodiment of the present invention, a managed Kubernetes service based on Mobile Edge Computing (MEC) is implemented. By managing the under cloud's virtual infrastructure in a unified container form, similar to the over cloud's workload Kubernetes, E2E (End) utilizes the closed loop control of Kubernetes. To End) management is possible.

Meanwhile, the operating method according to an embodiment of the present invention may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc., singly or in combination. Program instructions recorded on the medium may be specially designed and constructed for the present invention or may be known and usable by those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic media such as floptical disks. -Includes optical media (magneto-optical media) and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, flash memory, etc. Examples of program instructions include machine language code, such as that produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

Although the present invention has been described in detail with reference to preferred embodiments so far, the present invention is not limited to the above-described embodiments, and the technical field to which the present invention pertains without departing from the gist of the present invention as claimed in the following claims. Anyone skilled in the art will recognize that the technical idea of the present invention extends to the extent that various changes or modifications can be made.

According to the virtual resource operation device and the operating method of the virtual resource operation device according to the present invention, under cloud when implementing a managed Kubernetes service based on Mobile-Edge Computing (MEC) By surpassing the limitations of existing technologies in that the form of virtual infrastructure deployed in the system can be managed by unifying it with the over cloud, the possibility of not only using related technologies but also marketing or selling the applied devices is increased. It is an invention that has industrial applicability because it is not only sufficient but also clearly practicable in reality.

10: Mobile Edge Computing Orchestrator
20: Mobile Edge Computing Platform
30: Mobile Edge Platform Manager
40: Virtualization infrastructure manager
50: Virtual resource operation device
51: Virtualization Infrastructure Provision Department 52: Cluster Provision Department
53: Bootstrap provision unit 54: Cluster status check unit

Claims (10)

A virtual infrastructure provision department that creates virtual machines for workload Kubernetes using a containerized hypervisor;
a cluster provider that builds a cluster of the workload Kubernetes in a form including a master node and worker nodes matching the master node within the virtual machine; and
A bootstrap provider that connects the worker node to the master node and activates the workload Kubernetes in a serviceable state,
The cluster is,
Contains two or more master nodes according to predefined specifications for the workload Kubernetes,
The cluster provider department,
A virtual resource management device, characterized in that a load balancer for the two or more master nodes and a firewall for the workload Kubernetes are additionally created and placed in front of the cluster. delete According to claim 1,
The virtual infrastructure provision department,
A virtual resource management device characterized in that the containerized hypervisor is configured in the form of a pod (POD) and the virtual machine is created within the pod using the hypervisor. delete According to claim 1,
The virtual resource operation device is,
A virtual resource management device further comprising a cluster status checker that periodically checks whether the workload Kubernetes is available using a heartbeat message transmitted to the load balancer. A virtual infrastructure provision step of creating a virtual machine for the workload Kubernetes using a containerized hypervisor;
A cluster provision step of constructing a cluster of the workload Kubernetes in a form including a master node and a worker node matching the master node within the virtual machine; and
It includes a bootstrap provision step of activating the workload Kubernetes to a serviceable state by joining the worker node to the master node,
The cluster is,
Contains two or more master nodes according to predefined specifications for the workload Kubernetes,
The cluster provision step is
A method of operating a virtual resource management device, characterized in that additionally creating a load balancer for the two or more master nodes and a firewall for the workload Kubernetes and placing them in front of the cluster. delete According to claim 6,
The virtual infrastructure provision step is,
A method of operating a virtual resource management device, characterized in that the containerized hypervisor is configured in the form of a pod (POD), and the virtual machine is created within the pod using the hypervisor. delete According to claim 6,
The method is:
The operation of the virtual resource management device further comprising a cluster status check step of periodically checking whether the workload Kubernetes is available using a heart beat message transmitted to the load balancer. method.