KR101841847B1 - Method and apparatus for managing provisioning virtual disk - Google Patents

Abstract

The present invention relates to a method and an apparatus for managing virtual disk provisioning. The method comprises: a step of receiving a provisioning request for a plurality of virtual machines; a step of generating at least one feature model for the virtual machines; a step of searching an existing virtual disk having the same feature model to generate a virtual disk on the basis of the feature model; a step of determining a provisioning procedure of the virtual disk according to a search result; and a step of generating a virtual disk by copying the existing virtual disk according to a determined provisioning procedure of the virtual disk. A virtual disk is generated on the basis of a provisioning procedure of a virtual disk, and a time required for provisioning a virtual disk can be shortened by reusing and copying a pre-generated virtual disk.

Description

≪ Desc / Clms Page number 1 > METHOD AND APPARATUS FOR MANAGING PROVISIONING VIRTUAL DISK &

The present invention relates to a virtual disk provisioning management method and apparatus, and more particularly, to a virtual disk provisioning management method and apparatus capable of speeding up disk creation by utilizing a virtual disk already created.

As network technology and storage technologies have evolved, platform technology, or virtualization technology, has been actively researched to apply analytical processing techniques for large amounts of data stored in shared storage. Particularly, in order to effectively analyze a large amount of data, a technique of processing distributed and stored data in parallel is important. However, in order to process large amount of data distributed in parallel, it is necessary to use a virtual cluster technology that utilizes one or more processing computing nodes as clusters.

A virtual cluster is a set of virtual machines installed with software that performs a specific function and is connected to a network. To create virtual clusters, you must go through the steps of provisioning virtual computing resources. Virtual disk provisioning includes virtual cluster placement provisioning and virtual cluster disk provisioning. Virtual cluster placement provisioning refers to a process for creating virtual clusters based on resource information of virtual machines included in a virtual cluster to be created by placing them in a virtualized physical system. Virtual cluster disk provisioning refers to the process of creating a virtual disk to create a virtual machine in a virtual cluster. Here, the process of installing and setting the software requested by the user in each of the created virtual disks, that is, the system software and the application software, directly affects the total virtual disk provisioning time.

[Related Technical Literature]

A virtual machine providing method, a system, and a recording medium on which the program is recorded (Patent Document 10-2014-0055481)

The virtual cluster may be a collection of a plurality of different virtual machines. Here, the virtual machine may include a virtual disk in which software is installed.

The inventors of the present invention have recognized that the process of repeatedly installing the same software and repeating the setting of the same software result in a waste of the entire virtual disk provisioning time.

Accordingly, an object of the present invention is to provide a virtual disk provisioning management method and apparatus capable of shortening a time required for virtual disk provisioning by replicating a virtual disk included in a virtual cluster already created and providing the virtual disk to virtual machines .

Another object of the present invention is to provide a virtual disk provisioning management method and device capable of reducing the number of steps of virtual disk provisioning through virtual disk duplication in response to various virtual disk provisioning requests and shortening the time of virtual disk provisioning .

The problems of the present invention are not limited to the above-mentioned problems, and other problems not mentioned can be clearly understood by those skilled in the art from the following description.

According to an aspect of the present invention, there is provided a virtual disk provisioning management apparatus including a virtual disk provisioning processor receiving unit for receiving a provisioning request for a plurality of virtual machines, at least one feature for a plurality of virtual machines A reusable virtual disk search unit configured to search for an existing virtual disk having the same feature model to create a virtual disk based on the feature model; A virtual disk procedure determination unit configured to determine a provisioning procedure of the virtual disk, and a virtual disk creation unit that creates a virtual disk by copying an existing virtual disk according to a provisioning procedure of the determined virtual disk.

According to another aspect of the invention, the feature model may include at least one of a name of software, application software and operating system software, a version information package, a module and an association.

According to another aspect of the present invention, the feature model generation unit may generate a feature model for each of the subgroups including the plurality of virtual machines, and generate a feature model associated with at least one homology or heterogeneity based on the feature model have.

According to another aspect of the invention, a feature model associated with homogeneity or heterogeneity may be a feature model for software that overlaps the feature model for each of the subgroups, or a feature model for non-overlapping software.

In accordance with another aspect of the present invention, a virtual disk provisioning management apparatus includes a virtual disk provisioning management apparatus that, when a plurality of virtual machines includes at least one redundant software, at least two virtual machines share a first disk including redundant software, The first disk is read-only, and files requiring writing can be stored in a separate second disk of each of the virtual machines.

According to still another aspect of the present invention, the reusable virtual disk search unit can search an existing virtual disk through an integer identifier to which a hash function is applied to a feature model of a virtual disk to be generated.

According to another aspect of the present invention, the virtual disk procedure determination unit may determine a provisioning procedure of a virtual disk to be created based on an asset to be searched and replicated when an existing virtual disk exists, have.

According to another aspect of the present invention, the virtual disk procedure determination unit may search an installer of software to be installed and determine a provisioning procedure of the virtual disk based on the installer of the retrieved software.

According to another aspect of the present invention, the virtual disk procedure determination unit may determine a procedure for creating a virtual disk including each of an operating system virtual disk corresponding to operating system software and an application software virtual disk corresponding to the application software.

According to another aspect of the present invention, the procedure may include the name and version information of the operating system software.

According to another aspect of the present invention, the virtual disk procedure determination unit may determine a procedure for creating one virtual disk corresponding to operating system software and application software.

According to an aspect of the present invention, there is provided a virtual cluster disk provisioning management method including receiving a provisioning request for a plurality of virtual machines, generating at least one feature model for a plurality of virtual machines, Searching an existing virtual disk having the same feature model to create a virtual disk based on the feature model, determining a provisioning procedure of the virtual disk according to the search result, And creating a virtual disk by replicating an existing virtual disk.

According to another aspect of the invention, the feature model comprises at least one of a name of software, version information package, module and association, including application software and operating system software.

According to another aspect of the present invention, generating a feature model comprises generating a feature model for each of the subgroups comprising a plurality of virtual machines, and generating a feature model associated with at least one homogeneity or heterogeneity .

According to another aspect of the invention, a feature model associated with homogeneity or heterogeneity may be a feature model for software that overlaps the feature model for each of the subgroups, or a feature model for non-overlapping software.

According to another aspect of the present invention, a virtual disk provisioning management method includes: when a plurality of virtual machines include at least one redundant software, at least two virtual machines share a first disk including redundant software, The first disk is read-only, and files requiring writing can be stored in a separate second disk of each of the virtual machines.

According to another aspect of the present invention, the step of searching for an existing virtual disk may be a step of searching for an existing virtual disk through an identifier obtained by applying a hash function to a feature model of a virtual disk to be generated.

According to another aspect of the present invention, the step of determining a provisioning procedure of a virtual disk includes the steps of: searching for a location where an existing virtual disk is stored, And may be a step of determining the provisioning procedure.

According to another aspect of the present invention, the step of determining the provisioning procedure of the virtual disk may be the step of searching the installer of the software to be installed and the step of determining the provisioning procedure of the virtual disk based on the installer of the retrieved software.

According to another aspect of the present invention, the step of determining the provisioning procedure of the virtual disk determines a procedure for creating a virtual disk including each of the operating system virtual disk corresponding to the operating system software and each of the application software virtual disks corresponding to the application software Step.

According to another aspect of the present invention, the procedure includes the name of operating system software and version information.

According to another aspect of the present invention, the step of determining the provisioning procedure of the virtual disk may be a step of determining a procedure of creating one virtual disk corresponding to the operating system software and the application software.

The details of other embodiments are included in the detailed description and drawings.

The present invention can provide a virtual disk provisioning management method and apparatus capable of shortening the time required for virtual disk provisioning by replicating a virtual disk included in a virtual cluster already created and providing it to virtual machines.

The present invention can reduce the steps of virtual disk provisioning by replicating a virtual disk in response to various virtual disk provisioning requests, and shorten the time of virtual disk provisioning. A method and apparatus for managing virtual disk provisioning can be provided.

The effects according to the present invention are not limited by the contents exemplified above, and more various effects are included in the specification.

FIG. 1 is a diagram for explaining virtual disk provisioning according to an embodiment of the present invention.
2 is an exemplary block diagram illustrating a virtual disk provisioning management apparatus according to an embodiment of the present invention.
3 illustrates a virtual disk provisioning procedure according to a virtual disk provisioning management apparatus according to an embodiment of the present invention.
4 is a diagram illustrating virtual disk provisioning for a virtual cluster according to an embodiment of the present invention.
FIG. 5 is an illustration of a feature model and a virtual cluster feature model of a virtual cluster subgroup according to another embodiment of the present invention.
FIG. 6 illustrates a process of creating a reusable virtual disk based on a feature model according to another embodiment of the present invention.
FIG. 7 illustrates a process of creating a provisioning procedure of a virtual disk according to another embodiment of the present invention.
8A-8G illustrate the effect of virtual disk provisioning in accordance with another embodiment of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS The advantages and features of the present invention, and the manner of achieving them, will be apparent from and elucidated with reference to the embodiments described hereinafter in conjunction with the accompanying drawings. The present invention may, however, be embodied in many different forms and should not be construed as being limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. Is provided to fully convey the scope of the invention to those skilled in the art, and the invention is only defined by the scope of the claims.

The shapes, sizes, ratios, angles, numbers, and the like disclosed in the drawings for describing the embodiments of the present invention are illustrative, and thus the present invention is not limited thereto. In the following description, well-known functions or constructions are not described in detail since they would obscure the invention in unnecessary detail. Where the terms 'comprises', 'having', 'done', and the like are used herein, other parts may be added as long as '~ only' is not used. Unless the context clearly dictates otherwise, including the plural unless the context clearly dictates otherwise.

In interpreting the constituent elements, it is construed to include the error range even if there is no separate description.

Although the first, second, etc. are used to describe various components, these components are not limited by these terms. These terms are used only to distinguish one component from another. Therefore, the first component mentioned below may be the second component within the technical spirit of the present invention.

Like reference numerals refer to like elements throughout the specification unless otherwise specified.

It is to be understood that each of the features of the various embodiments of the present invention may be combined or combined with each other partially or entirely and technically various interlocking and driving is possible as will be appreciated by those skilled in the art, It may be possible to cooperate with each other in association.

Hereinafter, terms used in this specification will be defined.

As used herein, the term " provisioning " means a procedure for allocating, arranging, and distributing resources according to a user's request. Here, the resource may mean a memory, CPU, etc. of the server. Specifically, provisioning refers to a procedure for creating a virtual disk in which software is installed in a storage of a server according to a user's request, and placing the virtual disk in a virtual machine so that the software can be operated.

FIG. 1 is a diagram for explaining virtual disk provisioning according to an embodiment of the present invention.

Referring to FIG. 1 (a), the virtual machine 110 includes a virtual disk 120 in which software 130 is installed. The software 130 may include system software 131, first application software 132, and second application software 133.

Referring to FIG. 1 (a), a virtual disk 120 in which a root file system is installed is a device necessary for driving the virtual machine 110. The virtual machine 110 including the virtual disk 120 can boot from the root file system. The virtual disk 120 may be created in the form of a file or a logical volume in which a file system in an operating system of a physical machine exists, unlike a general hard disk. Accordingly, the virtual disk 120 forms a file system required for each virtual machine 110 in the created file or logical volume, and is provided to the virtual machine 110. The virtual disk 120 is also provided with application software such as the desired system software 131, the first application software 132 and the second application software 133, and a version of the system software 131 different from the operating system of the physical machine And the virtual disk 120 may be provided to the virtual machine 110. [

Referring to FIG. 1 (a), a virtual disk provisioning management system provisions a virtual disk 120 by installing software 130 in a virtual disk 120. The virtual disk provisioning management system creates an empty virtual disk 120 to create a virtual disk 120 in which the software 130 is installed. Subsequently, the virtual disk provisioning management system installs Debian 8.0 as the system software 131, Open JDK 7.0 as the first application software 132, and Hadoop 2.6.0 as the second application software 133 in the empty virtual disk 120 do. Finally, the virtual disk provisioning management system sets up the software 130 installed in the virtual disk 120. Specifically, the virtual disk provisioning management system performs necessary settings to execute Debian 8.0 as the system software 131, Open JDK 7.0 as the first application software 132, and Hadoop 2.6.0 as the second application software 133 do. In FIG. 1A, the virtual disk 120 may be an integrated virtual disk, but the virtual disk may be embodied as a merged virtual disk in which a plurality of virtual disks each have one software and can be combined with each other .

Referring to FIG. 1B, the virtual disk provisioning management system may create a virtual disk 140 including the same software by replicating the virtual disk 120 created in advance.

Referring to FIG. 1B, the virtual disk provisioning management system includes Debian 8.0 as the system software 131, Open JDK 7.0 as the first application software 132, The virtual disk 120 in which the software 130 including the Hadoop 2.6.0 as the second application software 133 is installed is copied as it is to generate the virtual disk 140. [ Thus, the virtual disk 140 also includes Debian 8.0, Open JDK 7.0, which is the first application software 132, and Hadoop 2.6.0, which is the second application software 133. [ Provisioning a virtual disk through replication can be faster than creating a virtual disk through installation. Specifically, virtual disk creation through installation installs software that is actually needed for the virtual disk. On the other hand, virtual disk creation through replication is faster than virtual disk creation through installation because it only takes time to replicate software installed on a pre-created virtual disk. In addition, the creation of a virtual disk through replication is performed by setting the software installed in the virtual disk to replicate the virtual disk, thereby avoiding the step of setting up the software for creation of the virtual disk through replication. Can be created quickly.

As a result, the virtual disk provisioning management system can create a virtual disk on which software is installed according to the user's request, duplicate the created virtual disk, duplicate the virtual disk containing the duplicated software, and shorten the time required for provisioning the virtual disk can do.

Although FIGS. 1 (a) to 1 (b) illustrate shortening the time required for virtual disk provisioning by simply copying a virtual disk, replication of a virtual disk may involve various procedures and determination processes . Hereinafter, various embodiments for shortening the time required for virtual disk provisioning through replication and creation of a virtual disk will be described in detail.

2 is an exemplary block diagram illustrating a virtual disk provisioning management apparatus according to an embodiment of the present invention.

2, the virtual disk provisioning management apparatus 200 includes a virtual disk provisioning process receiving unit 210, a feature model generating unit 220, a reuse virtual disk searching unit 230, a virtual disk provisioning procedure determining unit 240, And a virtual disk creation task performing unit 250. [ The virtual disk provisioning process receiving unit 210 includes a virtual disk provisioning process performing unit 211, a virtual disk provisioning starting unit 212, and a virtual disk provisioning processor observing unit 213. The feature model generation unit 220 includes a virtual cluster subgroup feature model generation unit 221 and a virtual cluster feature model generation unit 222. The virtual disk provisioning procedure determination unit 240 includes a virtual disk provisioning procedure generation unit 241 and a reuse virtual disk provisioning procedure generation unit 242. [ The virtual disk creation task execution unit 250 includes a virtual disk creation unit 251 and a reuse metadata generation unit 252.

Referring to FIG. 2, the virtual disk provisioning process receiving unit 210 receives a provisioning request for a virtual disk in a request for building a virtual cluster using XML RPC (Remote Procedure Call). The virtual disk provisioning process receiving unit 210 drives the CML RPC server and delivers the received virtual disk provisioning request to the virtual disk provisioning process performing unit 211. At this time, the provisioning request for the virtual disk received by the virtual disk provisioning process receiving unit 210 includes the software requirements installed in the virtual machines of the virtual cluster requested by the user. The virtual disk provisioning process performing unit 211 parses the received virtual disk provisioning request. In addition, the virtual disk provisioning process performing unit 211 stores the task object parsed in the internal buffer to manage the progress status of the virtual disk provisioning task. All virtual disk provisioning processes of the present invention are performed by calling a subcomponent by the virtual disk provisioning process performing unit 211. [ Here, the sub-component is a virtual disk provisioning management device such as a feature model generation unit 220, a reuse virtual disk search unit 230, a virtual disk provisioning procedure determination unit 240, etc., which are components of the virtual disk provisioning management device 200, Quot; module " Each subcomponent receives, processes, and returns the task to be performed by the virtual disk provisioning process performing unit 211 on a task-by-task basis. In addition, each subcomponent includes a task execution unit that calls stepwise processing logic to be executed based on the status information of the task. The virtual disk provisioning initiation unit 212 is a component that drives and terminates subcomponents present in the virtual disk provisioning management device 200. [ The subcomponents of the virtual disk provisioning management device 200 are operated and managed by the virtual disk provisioning initiator 212. The virtual disk provisioning processor observing unit 213 periodically monitors the progress of a job managed by the virtual disk provisioning process performing unit 211 and transmits the progress status to the external system. Here, the external system may be a web-based system that provides a virtual disk provisioning service. Specifically, the external system may include a number of subgroups in the virtual cluster desired by the user, a number of virtual machines included in the virtual cluster subgroup, a description of virtual resources such as VCPU, memory, virtual disk space, The system and application software and setting information of the virtual cluster, and the time and space constraints of the virtual cluster. That is, the external system may transmit a virtual disk provisioning creation request to the virtual disk provisioning process receiving unit 210. [

The feature model generation unit 220 generates at least one feature model for a plurality of virtual machines. Specifically, the feature model generation unit 220 generates a feature model for a virtual cluster including a plurality of virtual machines.

On the other hand, in various embodiments, when a request for creation of two or more groups, i.e., virtual cluster subgroups, for a plurality of virtual machines is received, the virtual cluster subgroup feature model generation unit 221 generates Create a feature model for the group. In order to generate the feature model associated with the homogeneity or heterogeneity of the software included in the virtual cluster subgroup, the virtual cluster subgroup feature model generation unit 221 generates a homogeneous or heterogeneous software component based on the subgroup of the virtual cluster, Or non-overlapping software. The virtual cluster subgroup feature model generation unit 221 generates a homogeneous software component and a heterogeneous software component included in the subgroup of the virtual cluster as a feature model. The virtual cluster feature model generation unit 222 compares and analyzes the feature models of the generated subgroups, and generates a feature model of the virtual disks of the virtual clusters to be finally generated. The feature model associated with the generated homogeneity or heterogeneity may be a feature model for software that overlaps the virtual cluster subgroup feature model or a feature model for non-overlapping software. Here, the feature model includes at least one of a name of software including application software and operating system software, a version information package, a module, and an association. A detailed description of a case where a request for a virtual cluster subgroup is received will be described later with reference to FIG.

The reusable virtual disk searching unit 230 searches a reusable virtual disk for analyzing a reusable virtual disk based on the feature model generated through the feature model generation unit 220 and a virtual disk to be newly installed and created. Specifically, the reusable virtual disk searching unit 230 searches a reusable virtual disk in a reusable asset storage based on a feature model of at least one feature model for a plurality of virtual machines, newly installs a reusable virtual disk Classify the virtual disks that should be created. Here, the reusable asset repository is a repository that stores artifacts of all the software created in the virtual disk provisioning management device 200. The output that may be stored is a virtual disk provisioning request received from an external system, a feature model of a virtual cluster, a created virtual disk provisioning procedure and a reusable virtual disk provisioning procedure, a reusable virtual disk, a software installation and configuration program. The virtual disk created by the virtual disk creation unit 251 is not stored in the reuse asset storage, but is stored in a separate shared storage, since it must be utilized when a virtual machine is actually created. The reusable virtual disk searching unit 230 may assign an integer identifier to the feature model of the virtual disk to be stored in the reusable asset repository and apply the hash function to retrieve the existing virtual disk from the reusable asset repository.

The virtual disk provisioning procedure determining unit 240 determines the provisioning procedure of the virtual disk according to the search result of the reuse virtual disk searching unit 230. Specifically, the virtual disk provisioning procedure determination unit 240 determines a virtual disk provisioning procedure and a method that specify whether to create a virtual disk for a plurality of virtual machines through installation or replication. When the virtual disk is newly installed and provisioned, the virtual disk provisioning procedure generator 241 searches the reusable asset repository for the software installer required for the installation and generates a virtual disk provisioning procedure based on the installer of the retrieved software do. At this time, the virtual disk provisioning procedure generation unit 241 generates a procedure for installing a virtual disk including each of the operating system virtual disk corresponding to the operating system software and the application software virtual disk corresponding to the application software. The generated procedure includes the operating system software name and version information. The virtual disk provisioning procedure generation unit 241 may also generate a procedure for creating a virtual disk to be reusable later. When there is a reusable virtual disk, the reusable virtual disk provisioning procedure generation unit 242 searches for a location where the reusable virtual disk is stored and creates a provisioning procedure of the virtual disk to be created based on the replica set asset. That is, the reusable virtual disk provisioning procedure generation unit 242 creates a reusable virtual disk provisioning procedure in which the reusable virtual disk provisioning procedure generation unit 242 replicates the reusable virtual disk based on the reusable virtual disk or installs additional application software. Accordingly, the re-use virtual disk provisioning procedure generator 242 must create a re-use virtual disk provisioning procedure by duplicating a virtual disk to be reused, setting up necessary application software, and setting the re-use virtual disk provisioning procedure. The virtual disk to be replicated is analyzed by the reusable virtual disk searching unit 230 and retrieved as an integer identifier obtained by using the hash function. The reusable virtual disk provisioning procedure generation unit 242 parses the reused asset description of the retrieved reusable virtual disk to generate a reusable virtual disk provisioning procedure.

The virtual disk creation task execution unit 250 generates all the tasks for creating a virtual disk according to the virtual disk provisioning procedure or the reusable virtual disk provisioning procedure. The virtual disk creation unit 251 installs a virtual disk of a new virtual cluster based on the virtual disk provisioning procedure according to the generated task or creates a virtual disk of the virtual cluster through replication based on the reuse virtual disk provisioning procedure. When the virtual disk creation unit 251 installs a new virtual disk based on the virtual disk provisioning procedure, the reuse metadata creation unit 252 creates metadata of the virtual disk created according to the virtual disk provisioning procedure. That is, the reuse metadata generation unit 252 generates the metadata of the virtual disk so that the newly created virtual disk can be reused later. Here, the metadata is name and version information of system software and application software. According to various embodiments, when a plurality of virtual machines including the created virtual disk includes at least one redundant software, at least two virtual machines may share a first disk containing redundant software. For example, the shared first disk may be read only, and the files that need to be written may be stored in a separate second disk of each virtual machine.

Accordingly, the virtual disk provisioning management device generates and stores the provisioning procedure of the virtual disk to be created, and when the same virtual disk is later provisioned, the time required for provisioning the virtual disk by utilizing the provisioning procedure of the already created virtual disk Can be shortened.

3 illustrates a virtual disk provisioning procedure according to a virtual disk provisioning management apparatus according to an embodiment of the present invention. For convenience of explanation, the components will be described with reference to FIG. 2 and reference numerals.

The virtual disk provisioning process receiving unit 210 receives a provisioning request for a plurality of virtual machines (S310).

Specifically, the virtual disk provisioning process receiving unit 210 receives a provisioning request for a virtual cluster including a plurality of virtual machines. The virtual disk provisioning process performing unit 211 parses the virtual disk provisioning specification included in the received virtual disk provisioning request. At this time, the virtual machine of the virtual cluster must be able to be placed on a specific physical host so that the virtual disk provisioning specification can be parsed. The virtual disk provisioning specification is an XML-format document conforming to the OVF standard, and the parsed result is generated as an internal data structure to be used in the virtual disk provisioning management apparatus 300. In addition, the virtual disk provisioning process performing unit 211 determines whether a virtual disk of the virtual cluster existing in the parsed specification can be created. The criterion that can be determined is whether or not there is a space that can be stored based on the virtual disk size of the specified virtual cluster and whether or not provisioning is possible until the virtual disk usage time defined by the user is determined. To perform this process, there must be a shared storage that can store the virtual disk, and the result is whether the virtual disk can be provisioned.

Then, the feature model generation unit 220 generates at least one feature model for a plurality of virtual machines (S320).

The feature model generation unit 220 generates a feature model for a virtual cluster including a plurality of virtual machines.

Then, the reuse virtual disk searching unit 230 searches an existing virtual disk having the same feature model to create a virtual disk based on the feature model (S330).

The reusable virtual disk searching unit 230 searches a reusable virtual disk based on a feature model of a virtual cluster including a plurality of virtual machines. The reusable virtual disk searching unit 230 uses a unique integer identifier to which a hash function is applied to a feature model of a virtual disk to be provisioned for searching a reusable virtual disk. The integer type identifier is used to search for a location where a reusable virtual disk exists or an ID value for registering a newly created virtual disk in a reusable asset storage. Accordingly, the re-use virtual disk searching unit 230 searches for an existing virtual disk through an integer identifier to which a hash function is applied to the feature model of the virtual disk to be generated. At this time, the existing virtual disk is stored in the reusable asset storage.

Subsequently, the virtual disk provisioning procedure determination unit 240 determines a provisioning procedure of the virtual disk according to the search result (S340).

The virtual disk provisioning procedure determination unit 240 generates a provisioning procedure and a method of reusing or newly installing and creating a virtual disk. At this time, a reusable asset repository is required to retrieve software installers and configurators. If the reusable virtual disk retrieval unit 230 finds a reusable virtual disk, the reusable virtual disk provisioning procedure generation unit 242 searches for a location where an existing virtual disk is stored using a unique integer identifier value, To create a provisioning procedure. On the other hand, the virtual disk provisioning procedure determining unit 240 for the virtual disk to be installed and newly installed searches for the installer (software) of the software for installing the software, and based on the installed software installer, The non-provisioning procedure generation unit 241 generates the provisioning procedure of the virtual disk. The virtual disk provisioning procedure generator 241 and the reusable virtual disk provisioning procedure generator 242 generate an XML document so as to be utilized in various virtual disk creation systems. The created virtual disk provisioning procedure includes an operating system virtual disk corresponding to the operating system software and each of the application software virtual disks corresponding to the application software. At this time, the virtual disk provisioning procedure includes the operating system software name and version information. Accordingly, the virtual disk provisioning procedure determination unit 240 determines a procedure for creating one virtual disk corresponding to the operating system software and the application software.

Subsequently, the virtual disk creation unit 250 creates a virtual disk by copying the existing virtual disk according to the provisioning procedure of the determined virtual disk (S350).

More specifically, the virtual disk creation unit 250 creates all the tasks for creating a virtual disk. In accordance with the generated task, the virtual disk creation unit 251 creates a virtual disk according to a virtual disk provisioning procedure or a reusable virtual disk provisioning procedure. In the case of creating a reusable virtual disk, a reusable asset storage is necessary because the virtual disk creation unit 251 replicates and creates a reusable virtual disk. In addition, a shared storage is needed so that the created virtual disk can be used at the time of running the virtual machine. The virtual disk creation unit 251 creates and sets a virtual disk using software installation and configuration assets defined in the order specified in the reuse virtual disk provisioning procedure. If there is no reusable virtual disk and the virtual disk creation unit 251 installs the virtual disk according to the virtual disk provisioning procedure, the reuse metadata generation unit 252 stores the integer identifier of the virtual disk specified in the virtual disk provisioning procedure Generate reuse metadata based on location information. Reuse metadata is documented, for example, using the Reusable Asset Specification, but is not limited thereto. At this time, the reusable asset statement is stored in the reusable asset repository to enable automatic and incremental reuse management. Additionally, when a plurality of virtual machines includes at least one redundant software, at least two virtual machines may share a first disk containing redundant software. For example, the shared first disk can be read only, and the files requiring writing can be stored in a separate second disk of each virtual machine.

Accordingly, the virtual disk provisioning management apparatus determines whether the software included in the virtual cluster subgroup to be generated is duplicated, and shortens the virtual disk creation time by provisioning the virtual disk based on the provisioning procedure of the already created virtual disk have.

4 is a diagram illustrating virtual disk provisioning for a virtual cluster according to an embodiment of the present invention.

The virtual cluster may include different virtual clusters. Here, the different virtual clusters may be a set of virtual machines in which at least one other software is installed. In this specification, the different virtual clusters are referred to as virtual cluster subgroups.

Virtual Disk Provisioning Management A device is a device that, when there are two or more virtual cluster subgroups to provision, to prevent redundant provisioning of a virtual disk for a virtual cluster subgroup, a homogeneous, heterogeneous software component And determines the type of the final virtual disk. That is, the virtual disk provisioning management apparatus recognizes the duplicated software and the non-duplicated software existing in the virtual cluster subgroup, and generates a feature model associated with at least one homogeneity or heterogeneity. At this time, the feature model generated by the virtual disk provisioning management device may be a feature model for overlapping software associated with homogeneity, or a feature model for non-overlapping software associated with heterogeneity. Therefore, the virtual disk provisioning management apparatus generates feature models based on the software components of the virtual cluster subgroups, and compares and analyzes the feature models of the respective subgroups, so that the feature models of the virtual disks including the overlapped software are not duplicated Create a feature model for non-software.

Referring to FIG. 4, the first virtual cluster subgroup 410 includes a plurality of virtual machines 420. In addition, the second virtual cluster subgroup 440 includes a plurality of virtual machines 450. The virtual disk provisioning for a normal virtual cluster is performed by repeating the procedure of creating a virtual disk in which the software 430 and 460 are installed according to the number of virtual machines 420 and 450 and further performing a software setting operation for clustering it means.

Referring to FIG. 4, the virtual disk provisioning management apparatus analyzes the types and the number of virtual disks to be created in consideration of the virtual cluster subgroup 410. The procedure for analyzing the types and the number of concrete virtual disks will be described later with reference to FIG. Accordingly, the virtual disk provisioning management apparatus, in accordance with the analysis result, for example, Debian 8.0 which is the system software 431 in the virtual disk of one virtual machine 420 included in the first virtual cluster subgroup 410, Software 430 including Open JDK 7.0 which is software 432 and Hadoop 2.6.0 which is second application software 433 is installed. Subsequently, the virtual disk provisioning management system repeatedly performs the virtual disk provisioning for the virtual cluster by the number of the virtual machines 420 included in the first virtual cluster 410.

Similarly, the virtual disk provisioning management system analyzes the types and the numbers of the virtual disks included in the second virtual cluster subgroup 440, and analyzes the number and type of the virtual disks included in the second virtual cluster subgroup 440, The software 460 including Debian 8.0 which is the system software 461, Open JDK 7.0 which is the first application software 462 and Hbase 1.1.1 which is the second application software 463 is installed in the virtual disk of the server 450 in the virtual disk. Subsequently, the virtual disk provisioning management device repeatedly performs virtual disk provisioning as many times as the number of virtual machines 450 included in the second virtual cluster subgroup 440.

Comparing the software 430 and 460 included in the first virtual cluster subgroup 410 and the second virtual cluster subgroup 440, the system software 431 and 461 are identical to Debian 8.0, and the application software 432, 462) is the same as Open JDK 7.0, but application software 433 and 462 are different with Hadoop 2.6.0 and Hbase 1.1.1, respectively. Accordingly, the virtual disk provisioning management apparatus analyzes the software to be installed in the virtual disk of each virtual cluster subgroup, thereby distinguishing the overlapping software. Subsequently, when the system software and the application software to be installed in the new virtual disk are the same, the virtual disk provisioning management system replicates the virtual disk in which the system software 431 and the application software 432 are installed and creates the same system software 461 in the new virtual disk, And application software 462 are installed.

Thus, the virtual disk provisioning management device replicates the preinstalled virtual disk, thereby shortening the time required for the entire virtual disk provisioning. The process of specifically creating a virtual disk included in a virtual cluster by distinguishing duplicate software will be described later with reference to FIG. The created virtual disk can be reused. Therefore, an existing virtual disk can be used to replicate a new virtual disk that contains the same software included in an existing virtual disk.

Accordingly, the virtual disk provisioning management apparatus distinguishes duplicate software included in each virtual cluster subgroup, replicates an existing virtual disk in which software is installed, creates a new virtual disk in which the same software is installed, .

FIGS. 5A through 5C are schematic diagrams of a virtual cluster sub-group feature model and a virtual cluster feature model according to another embodiment of the present invention.

Referring to FIG. 5A, the first virtual cluster subgroup feature model 510 is a feature model for the Hadoop virtual cluster subgroup. The first virtual cluster subgroup feature model 510 includes system software 511 and application software 512, 513 based on the AMD64 architecture. Specifically, the first virtual cluster subgroup feature model 510 includes the Debian 8.0 Base package as the system software 511 and includes the Hadoop 2.6.0 application software 512 and the OpenJDK 7.0 application software 513 do.

Referring to FIG. 5B, the second virtual cluster subgroup feature model 520 is a feature model for the HBase virtual cluster subgroup. The second virtual cluster subgroup feature model 520 includes the Debian 8.0 Base package and the OpenJDK 7.0 application software 523 as the system software 521 based on the AMD64 architecture in the same manner as the first virtual cluster subgroup feature model 510 . In addition, the second virtual cluster subgroup feature model 520 further includes application software 522 with Hbase 1.1.1. That is, the first virtual cluster subgroup feature model 510 needs to install Hadoop 2.6.0 as the application software 512, and the second virtual cluster subgroup feature model 520 needs to be installed in the application software 522 as Hbase 1.1. 1.

5A and 5B, since the virtual disk provisioning management apparatus must create a virtual disk including all the features, the first virtual cluster subgroup feature model 510 and the second virtual cluster All features included in the subgroup feature model 520 have a mandatory association.

Referring to FIG. 5C, the virtual disk provisioning management apparatus integrates the first virtual cluster subgroup feature model 510 and the second virtual cluster subgroup feature model 520 to obtain a virtual cluster feature model 530 . The system software 511, 521 and the application software 513, 523, which are overlapped by the first virtual cluster subgroup feature model 510 and the second virtual cluster subgroup feature model 520, And the non-overlapping application software 512, 522 is represented by the relationship of Exclusive OR. Accordingly, the virtual cluster disk provisioning management apparatus generates metadata based on the virtual cluster feature model 530 combining the first virtual cluster subgroup feature model 510 and the second virtual cluster subgroup feature model 520 . Specifically, the virtual cluster disk provisioning management apparatus determines whether the first virtual cluster subgroup characteristic model 510 and the second virtual cluster subgroup characteristic model 520 overlap with each other based on the virtual cluster characteristic model 530, The application software 532 and the application software 533 overlapping the application software 531 and the application software 534, respectively.

Accordingly, the virtual disk provisioning management apparatus classifies each of the virtual cluster sub-group feature models to classify the overlapping software and the non-overlapping software, thereby repeating the duplicated software to create a virtual disk. You can reduce the provisioning time by changing to a procedure and a recurring replication procedure.

FIG. 6 illustrates a process of creating a reusable virtual disk based on a feature model according to another embodiment of the present invention.

6, in order to analyze a reusable virtual disk, the virtual disk provisioning management apparatus expresses features of a software configuration possessed by a virtual disk as a feature model, and reuses virtual virtual disks based on the integer ID values generated through the hash function. Manage the disk.

6, the virtual disk provisioning management apparatus includes a Debian 8.0 Base package as system software 611 in a virtual cluster feature model 610, application software 612 with Hadoop 2.6.0, application software 6 with OpenJDK 7.0 And a virtual cluster feature model document 620, which is an XML document, based on the generated virtual cluster feature model 610. Subsequently, the virtual disk provisioning management apparatus applies a hash function to the virtual cluster feature model document 620 to generate an integer ID value 630 '1956267503'. Here, the hash function uses the hash function hash_fn included in the C ++ standard library funtional. The virtual disk provisioning management apparatus searches for a reusable virtual disk based on the integer type ID value 630 of the created virtual cluster feature model document 620. Specifically, the virtual disk provisioning management apparatus searches the re-use virtual disk storage for a virtual disk having an integer ID value equal to the integer type ID value 630. If there is a reusable virtual disk among the virtual cluster characteristics models stored in the reusable virtual disk storage, the virtual disk provisioning management device determines whether the integer type ID value of the reusable virtual disk, the software component included in the virtual disk, And so on.

Accordingly, when the virtual disk provisioning management apparatus assigns an integer ID value to a virtual cluster characteristic model document and duplicates the same virtual disk, the virtual cluster characteristic model document is reused by searching the existing virtual cluster characteristic model document based on the integer type ID value, .

FIG. 7 illustrates a process of creating a provisioning procedure of a virtual disk according to another embodiment of the present invention.

Referring to FIG. 7, the virtual disk provisioning management apparatus includes a virtual disk (virtual disk) having Debian 8.0 as the system software 711, Open JDK 7.0 as the first application software 712, and Hadoop 2.6.0 as the second application software 713 710, a virtual disk provisioning procedure 740 that is a document specifying the virtual disk provisioning procedure. Here, the virtual disk provisioning procedure 740 includes a procedure for creating an empty virtual disk image file, and installing a new system and application software, as an XML-formatted document. In addition, the virtual disk provisioning procedure 740 stores the generated virtual disk in the reusable asset storage 730 for reuse in the future, replicates the generated virtual disk as many times as necessary, and sets software in the replicated virtual disk Procedure. In addition, when the virtual disk is cloned based on the feature model of the virtual disk already stored in the re-use asset storage 730, the virtual disk provisioning procedure 740 generates an empty virtual disk image file, and replays the re- And duplicating the required number of generated virtual disks and setting software in the duplicated virtual disk.

Referring to FIG. 7, the virtual disk provisioning management apparatus searches whether the same feature model document exists in the reuse asset repository 730 based on the integer ID value 720 given to the feature model document of the virtual disk 710 . That is, the virtual disk provisioning management apparatus retrieves data in the form of an installer and a configurator of the software for performing the virtual cluster disk creation procedure in the reusable asset storage 730. Thus, the virtual disk provisioning management device generates a virtual cluster disk provisioning procedure 740 in the form of XML based on the search data of the reusable asset store 730.

Accordingly, the virtual disk provisioning management apparatus can document, store, and utilize the procedures of the virtual disk provisioning to shorten the provisioning step and the time required to create the same virtual cluster subgroup.

8A-8G illustrate the effect of virtual disk provisioning in accordance with another embodiment of the present invention.

Referring to FIG. 8A, there is shown a graph 810 of a single virtual disk creation time measurement result for each non-reuse and reuse-based virtual disk type. Here, the re-use means that when a new virtual disk is created by equally including the software installed in a previously created virtual disk, a new virtual disk is created by duplicating the previously created virtual disk.

Referring to FIG. 8A, the size of the virtual disk used for checking the effect of the virtual disk provisioning is 3 GB, the system software is Debian 7.0, the application software is OpenJDK 7.0, Hadoop 2.6.0, and HBase 1.1 .1 was used. As a result, the non-reused integrated virtual disk 811 on the virtual machine allocated with 4 VCPUs and 4 GB of memory represents the generation time of 1273.85 seconds, and the non-reused combined virtual disk 812 has the generation time of 1217.69 seconds Respectively. Here, the merged virtual disk is created by the merged virtual disk provisioning, and each software exists as a separate virtual disk type. Combined Virtual Cluster Disk provisioning refers to provisioning in which virtual disks are classified by software type and a plurality of classified software are merged to be provided to a virtual machine of a virtual cluster. Thus, the virtual disk provisioning device merge each of the merge type virtual disks into a virtual machine. In addition, an integrated virtual cluster disk is created by an integrated virtual disk provisioning and exists as one virtual disk in which one software is installed by integrating all the software. Integrated virtual cluster disk provisioning refers to provisioning in which a plurality of software can be integrated and provided to a virtual machine in a virtual cluster as a single software. Thus, the virtual disk provisioning device provides an integrated virtual disk that integrates and installs each software to the virtual machine.

 The difference in the generation time of the non-reused integrated type and the merged type virtual disks 811 and 812 can be regarded as a difference that can occur because the time for installing the system and the application software uses the external storage where each software package exists. The difference in the generation time of the non-reuse integrated type and the merged type virtual disks 811 and 812 can be regarded as a slight difference in installation time depending on the network environment of the external storage where the software package exists. Therefore, it can be seen that the generation times of the non-reused integrated type and the merged type virtual disks 811 and 812 are similar regardless of the virtual disk type.

When the virtual disk on which the system software is installed is reused, the total generation time of the integrated and merged virtual disks 813 and 814 represents 629.16 seconds and 608.96 seconds, respectively. If only system software is reused, it is necessary to install application software. Similar to the generation of the non-reusable virtual disks 811 and 812, the difference in the generation time of the system software reuse integrated type and the merged type virtual disks 813 and 814 is that a package . As a result, even if the system software is reused, if the application software to be additionally installed exists, the generation times of the system software reuse integrated type and the merged type virtual disks 813 and 814 are similar. When all the virtual disks having the system and application software are reused, the generation times of the integrated type and the combined virtual disks 815 and 816 are 356.17 seconds and 233.87 seconds, respectively. Reuse of System and Application Software Compared to the integrated virtual disk 815, the reuse of system and application software is shortened because the generation time of the merged virtual disk 816 is shortened due to the difference in method of reusable virtual disk duplication. System and Application Software Reuse The integrated virtual disk 815 replicates the contents of the reusable virtual disk after creating a blank disk image corresponding to the size of the virtual disk requested by the user. If the size of the reusable virtual disk is 2.2 GB, the user can not provide the requested virtual disk size of 3 GB. In contrast, the system and application software reuse combined virtual disk 816 can create an upper disk and merge it, so that the virtual disk having the reusable software installed can be copied and used as it is. Here, the Upper disk is a disk that stores settings for software installed in the merged virtual disk and stores data to be stored due to a write operation. If the size of the reusable virtual disk is 2.2 GB, it is necessary to provide 0.8 GB of the upper disk. Duplicating the contents of a reusable virtual disk image consumes a lot of copying time compared to the time of duplicating a single disk image file because it repeatedly performs all the file-by-file copying operations. Therefore, when reusing all the software components, the time for generating the concurrent disk is shortened as compared with the integrated type.

Referring to FIG. 8B, there is shown a graph 820 of the virtual disk provisioning preprocessing time ratio as the total number of virtual machines increases. It can be seen that the overhead is negligible as the ratio of the actual creation time of the virtual disk increases compared to the temporal overhead generated in the actual preprocessing process. Of course, when the total number of virtual machines is 14 (821), it can be seen that the time ratio of the highest preprocessing process is 0.55%. However, it can be seen that as the number of total virtual machines increases, the time proportion of the preprocessing process decreases compared to the total virtual disk provisioning time. This is because the preprocessing process occurs only once per virtual cluster, and the time required to generate the virtual disk is relatively longer than the preprocessing process. Of course, if the number of software components included in the virtual cluster virtual machine is varied, it will consume more virtual disk provisioning preprocessing time than the result of the graph 820 of the virtual disk provisioning preprocessing time ratio as the total number of virtual machines increases . However, the increasing number of software components in virtual clusters means that the amount of software that needs to be installed increases accordingly. Therefore, it can be interpreted that the time generated in the pre-processing of virtual disk provisioning performed in the present invention causes a slight temporal overhead when considering the actual virtual disk creation time.

Referring to FIG. 8C, a graph 830 of the virtual disk type creation time result as the number of virtual machines included in one virtual cluster increases. This will be described with reference to FIG.

Referring to FIG. 8C, the reuse-based integrated virtual disk creation time in the virtual machine allocated with four VCPUs and 4 GB of memory shows an exponential increase in the generation time in the number of virtual machines. Especially, it can be seen that the generation time increase rate slightly increases from the time when the number of virtual machines is changed from 8 (831) to 16 (832). The reuse-based integrated virtual disk creation time is generated because the disk created by the number of virtual machines must be duplicated. That is, since the virtual machines existing in the same virtual cluster do not share the created virtual disk, the integrated system must duplicate the disk in which all necessary system and application software are installed by the number of virtual machines. In the case of replication, the same result can be obtained because it is necessary to replicate the replica by the size of the virtual disk requested by the user.

Referring to FIG. 8D, a combined virtual disk creation time reduction rate graph 840 of a virtual machine having 32 virtual disks included in the virtual cluster with an increase in the maximum number of shared virtual machines per virtual disk is included.

Referring to FIG. 8D, a combined virtual disk creation time reduction rate graph 840 of a virtual machine having 32 virtual servers included in the virtual cluster according to an increase in the maximum number of shared virtual machines per virtual disk is represented. Also, it is expressed based on the generation time reduction rate when the maximum number of shared virtual machines per virtual disk is one. As a result, the reuse-based concurrent disk creation time in a virtual machine allocated 4 VCPUs and 4 GB of memory is based on a total of 32 virtual machines (846). The maximum number of shared virtual machines per virtual disk is 1 to 4 (841, 842, 843). However, when the number of maximum shared virtual machines is 4 or more (843, 844, 845, 846), the decrease ratio of generation time is reduced to 11% or less. That is, it can be seen that the maximum number of virtual machines that can be shared per virtual disk is four (843) when the method of generating a merged virtual disk of the present invention is used. This means that even if the number of virtual machines included in the virtual cluster increases, the maximum generation time savings can be obtained when a maximum of four virtual machines (841, 842, and 843) share one virtual disk.

Referring to FIG. 8E, a generation time result graph 850 for each virtual disk type according to the number of virtual cluster subgroups increases. As a result, the integrated virtual disk 851 shows the linear disk creation time according to the increase in the number of subgroups of the virtual cluster on the virtual machine allocated with 4 VCPUs and 4 GB of memory. This is because as the number of subgroups increases, the total number of virtual disks to be created increases proportionally. In other words, when the number of subgroups is 2, the number of virtual disks to be created is 16 in total. If the number of subgroups is 4, the number of disks to be created is 32, and a total of 64 disks must be created for 8 subgroups. That is, it can be confirmed that the disk creation time increases in proportion to the number of increase in the number of virtual cluster subgroups. The merge type virtual disks 852, 853, 854, and 855 show a difference in generation time according to the number of virtual machines to be shared maximum per virtual disk. However, it shows the generation time proportional to the number of subgroups of the virtual cluster in common. And it shows that the virtual disk creation time increases in inverse proportion to the increase of the maximum number of shared virtual machines. That is, when two virtual machines (853) are shared maximum per virtual disk, 8 virtual disks and 8 virtual disks must be created when the subgroups are 2 and 4, respectively. If the maximum number of shared virtual machines per virtual disk is 4 (854), then 8 virtual disks should be created when the subgroups are 2 and 4, respectively. As the maximum number of shared virtual machines increases, the number of virtual disks to be created decreases in inverse proportion, and the generation time also decreases. Generation times of the integrated type and the combined virtual disks 851, 852, 853, 854, and 855 were 22.50% and 86.99%, respectively.

Referring to FIG. 8F, there is shown a graph 860 of a shared storage usage space comparison result of virtual disk types according to the increase in the number of virtual cluster subgroups. The integrated virtual disk 861 represents a linear shared storage usage size proportional to the increase in the number of virtual cluster subgroups. This can be interpreted as a result because virtual disks are created in proportion to the number of virtual machines existing in each virtual cluster subgroup. In addition, since the virtual machines in each subgroup do not share the virtual disk with the homogeneous software installed, it can be concluded that the shared storage is used linearly according to the number of virtual machines per subgroup. The merged virtual disks 862, 863, and 864 represent different shared storage usage spaces depending on the maximum number of virtual machines shared per virtual disk. First, if one shared virtual machine (861) is shared per merged virtual disk, it shows the same storage usage space as the integrated virtual disk. This is because the virtual machines in the virtual cluster subgroup are all the same as those in which individual virtual disks are provided. However, when the number of maximum shared virtual machines increases from 862 to 863 to 864, it shows 36.88%, 55.32% and 64.54%, respectively. Similar to the generation time reduction rate, when the maximum number of shared virtual machines per merged virtual disk is 4 or more (863, 864), the rate of reduction of shared storage space is decreased. As a result, the generation time of the merged virtual disks 862, 863, and 864 is shorter than that of the integrated virtual disk 861 even when the number of subgroups of the virtual cluster increases. . It can also be seen that the generation time of the integrated virtual disk 861 and the combined virtual disks 862, 863, and 864 is proportional to the increase of the number of virtual cluster subgroups. The space used in the shared storage also shows an increase in proportion to the number of subgroups of both the integrated virtual disk 861 and the combined virtual disks 862, 863, and 864. Particularly, when the number of maximum shared virtual machines among the merged virtual disks is one (862), it is confirmed that shared storage of the same size as the integrated type is used. Of course, in terms of the total virtual disk creation time, it is possible to create a disk faster than the integrated virtual disk 861. However, considering the use space of the shared storage, the virtual machine which has allocated 4 VCPUs and 4GB memory, It can be concluded that it is reasonable to perform disk provisioning of the virtual cluster after setting at least two (862, 863, 864).

Referring to FIG. 8G, there is shown a graph 870 of the generation time comparison result for each virtual disk type according to the number of virtual cluster subgroups in which overlapping application software exists. Experimental Results In the case of creating an integrated virtual disk (871) with redundant application software by repeatedly installing overlapping application software between subgroups in a virtual cluster on a virtual machine allocated with 4 VCPUs and 4GB memory, As the number of groups increased, it showed a tendency to increase exponentially. Especially, when the number of virtual cluster subgroups with redundant application software is more than 4, rapid disk creation time is shown. This is because the application software needs to be installed as many as the number of virtual cluster subgroups to be increased and the installed integrated virtual disk must be replicated by the number of virtual machines included in each subgroup. Similarly, if the redundant installation procedure of redundant application software is not removed in the subgroup of the virtual cluster using the merged virtual disk (872), the exponential generation time trend is increased. However, the generation of the merge type virtual disk 872 in which redundant application software exists may generate a somewhat smaller disk creation time pattern as the number of subgroups increases compared with the generation of the integrated virtual disk 872 in which redundant application software exists see. This is because the two virtual machines existing in each sub group share the virtual disk in which the duplicated software exists. Therefore, compared with the integrated virtual disk 871 in which the duplicated application software exists, The number of times of disk duplication for each virtual machine of the disk 872 decreases. Of course, the upper disk creation time for each virtual machine is proportional to the number of virtual machines in each subgroup. However, it is possible to obtain a result that the total virtual disk creation time is not significantly affected by the time of duplicating the virtual disk in which all the duplicate software is installed. Then, the integrated virtual disk 873 which analyzes the overlapping application software between the virtual cluster subgroups and removes the duplicated application software that eliminates the repetitive installation procedure, and the linearly combined virtual disk 874 that removes the overlapping application software Increase disk creation time. This is a result of analyzing the number of virtual cluster subgroups in which duplicated application software is repeatedly analyzed in advance, and then creating a new virtual disk by duplicating the virtual disk created after one installation. In other words, if duplication of application software is not analyzed, the installation process is repeated as many times as the number of subgroups. If overlapping software analysis is performed, the application software is installed only once, You only have to replicate as many as you like. And for all subgroups that do not have overlapping application software installed in the virtual cluster, the existing system software can be copied as it is. As a result, it can be interpreted that the number of replications is generated as "the number of virtual machines in the subgroup is -1" according to the increase of the virtual cluster subgroup in which the duplicated application software exists. When a merged virtual disk (874) is created by analyzing overlapping application software among subgroups of a virtual cluster and duplicated application software is removed, a linear virtual disk creation time increase pattern is shown. The merge type virtual disk 874 from which redundant application software is removed shows a virtual disk generation time increase similar to that of the integrated type virtual disk 873 in which redundant application software is removed. However, the merge type virtual disk 874 ), The generation time of the application software is relatively low because the number of the sub-groups is equal to the number of the application software installed. The merged virtual disk should provide an independent upper disk for each virtual machine, but it does not affect the overall virtual disk creation time. The analysis of overlapping application software among the subgroups of virtual clusters shows that the integrated virtual disk 871 in which redundant application software exists and the integrated virtual disk 873 in which application software which is overlapped is removed have a minimum of 30.87% and a maximum of 77.63% Generation time difference. Also, the merged virtual disk 872 in which redundant application software exists and the merged virtual disk 874 in which the overlapped application software is removed have a disc creation time difference of at least 27.23% and a maximum of 84.75%. In other words, when the duplicated application software analysis is not performed, the merged type virtual disk 872 in which the application software is overlapped proportionally and the merged type virtual disk 874 in which the duplicated application software is removed is 57.52% It can be confirmed that the disc creation time difference is shown. In the case of a merged virtual disk, it can be interpreted that the size of a duplicate virtual disk is smaller than that of an integrated virtual disk according to the number of subgroups having duplicated application software. That is, in the case of the integrated virtual disk, the created virtual disk is duplicated according to the number of virtual machines in each subgroup. At this time, the size of the virtual disk is duplicated by the size specified by the user. However, a merge virtual disk replicates as many virtual disks as the size of the actually installed software by the required number of subgroups. In other words, since the size of the file that replicates the installed virtual disk is smaller than that of the integrated virtual disk, the installation of the software in a redundant manner can make a big difference in the entire disk creation time. When duplicated application software existed in two subgroups, when the duplicate software analysis of the merged virtual disk was not performed, the generation time was somewhat higher than that of the duplicate software analysis of the integrated virtual disk. Even though the number of virtual disks to be created by a virtual machine included in a subgroup is shared by a virtual disk, it shows the generation time delay that a repetitive software installation can exhibit. Therefore, we can conclude that it is a way to reduce the virtual cluster disk provisioning time by replicating installed disks instead of minimizing the number of installs by analyzing whether there is homogeneous software between subgroups in order to create a more efficient virtual disk. In particular, the task of installing software can reveal unpredictable disk creation time because it utilizes external package repositories. This is because installation of software using external package repositories shows different installation times depending on the network conditions of the corresponding repository. Therefore, the virtual disk provisioning management method of the present invention, which prevents the repeated software installation procedure, can effectively shorten the entire virtual disk creation time by maximizing the number of copies of the created virtual disk image. This is an important virtual disk provisioning method that can achieve a higher disk creation time as the number of subgroups of a virtual cluster is large and the type of software to be installed varies.

In this specification, each block or each step may represent a part of a module, segment or code that includes one or more executable instructions for executing the specified logical function (s). It should also be noted that in some alternative embodiments, the functions mentioned in the blocks or steps may occur out of order. For example, two blocks or steps shown in succession may in fact be performed substantially concurrently, or the blocks or steps may sometimes be performed in reverse order according to the corresponding function.

The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. The software module may reside in a RAM memory, a flash memory, a ROM memory, an EPROM memory, an EEPROM memory, a register, a hard disk, a removable disk, a CD-ROM or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor, which is capable of reading information from, and writing information to, the storage medium. Alternatively, the storage medium may be integral with the processor. The processor and the storage medium may reside within an application specific integrated circuit (ASIC). The ASIC may reside within the user terminal. Alternatively, the processor and the storage medium may reside as discrete components in a user terminal.

Although the embodiments of the present invention have been described in detail with reference to the accompanying drawings, it is to be understood that the present invention is not limited to those embodiments and various changes and modifications may be made without departing from the scope of the present invention. . Therefore, the embodiments disclosed in the present invention are intended to illustrate rather than limit the scope of the present invention, and the scope of the technical idea of the present invention is not limited by these embodiments. Therefore, it should be understood that the above-described embodiments are illustrative in all aspects and not restrictive. The scope of protection of the present invention should be construed according to the following claims, and all technical ideas within the scope of equivalents should be construed as falling within the scope of the present invention.

110 virtual machines
120, 140, 710 virtual disks
130, 430, 460 software
131, 431, 461, 611, 711 System software
132, 432, 462, 712 First application software
133, 433, 463, 713 Second application software
200 Virtual Disk Provisioning Management Device
210 Virtual Disk Provisioning Process Receiver
211 Virtual Disk Provisioning Process Performing Department
212 Initiate Virtual Disk Provisioning
213 Virtual Disk Provisioning Processor Observation
220 Feature Model Generator
221 Virtual cluster subgroup creation unit
222 Virtual cluster feature model generation unit
230 Reusable Virtual Disk Search Unit
231 Duplicate Software Search Department
232 Non-overlapping software search section
240 Virtual Disk Provisioning Procedure Decision Unit
241 Virtual Disk Provisioning Procedure Generation
242 Reuse Virtual Disk Provisioning Procedure Generation Unit
250 virtual disk creation task performing part
251 Virtual Disk Creation Unit
252 Reuse metadata generation unit
410 first virtual cluster subgroup
420 Multiple virtual machines
440 Second virtual cluster subgroup
510 First Virtual Cluster Subgroup Feature Model
511, 521, 531 System Software
512, 513, 522, 523, 532, 533 Application software
520 Second Virtual Cluster Subgroup Feature Model
530 Virtual cluster feature model
610 Virtual cluster subgroup feature model
612, 613 application software
620 Virtual cluster subgroup feature model document
630, 720 Integer ID value
730 Reusable asset repository
740 Virtual Disk Provisioning Procedure
810 Graph of single virtual disk creation time measurement by type of non-reusable and reusable virtual disk
811 Non-reusable integrated virtual disk
812 Non-Reusable Mergeable Virtual Disk
813 System Software Reuse Integrated Virtual Disk
814 System Software Reuse Mergeable Virtual Disk
815 System and application software reuse Integrated virtual disk
816 System and Application Software Reuse Mergeable Virtual Disk
820 Graph of Virtual Cluster Disk Provisioning Preprocess Time Ratio with Increasing Total Virtual Machines
821 Number of virtual machines 14
830 Graph of the result of creation time for each type of virtual disk as the number of virtual machines included in the cluster subgroup increases
831 Number of virtual machines 8
832 Number of virtual machines 16
840 Graph of reduction rate of creation time of merged virtual disks of virtual machines with 32 servers included in the virtual cluster subgroup as the maximum number of shared virtual machines per virtual disk increases
841 Maximum number of shared virtual machines per virtual disk One
842 Maximum number of shared virtual machines per virtual disk 2
843 Maximum number of shared virtual machines per virtual disk 4
844 Maximum number of shared virtual machines per virtual disk 8
845 Maximum number of shared virtual machines per virtual disk 16
846 Maximum number of shared virtual machines per virtual disk 32
850 Graph of the creation time of virtual disk type as the number of virtual cluster subgroup increases
851 Integrated virtual disk
852, 853, 854, 855 Merge type virtual disk
860 Graph of the shared storage space used by virtual disk type as the number of virtual cluster subgroups increases
861 one virtual machine shared per integrated virtual disk and one virtual shared disk
862 shared virtual disks and two shared virtual machines per merged virtual disk
863 Consolidated virtual disks and 4 shared virtual machines per merged virtual disk
864 Consolidated virtual disks and 8 virtual machines shared per merged virtual disk
870 Comparison of creation time of virtual disk type according to the number of virtual cluster subgroups with redundant application software Graph
871 Integrated virtual disk with redundant application software
872 Merge type virtual disk with redundant application software
873 Integrated virtual disk with redundant application software removed
874 Mergeable virtual disk with redundant application software removed

Claims (22)

A virtual disk provisioning processor receiver for receiving a provisioning request for a plurality of virtual machines;
A feature model generation unit for generating at least one feature model for the plurality of virtual machines;
A reusable virtual disk search unit configured to search an existing virtual disk having the same feature model to create a virtual disk based on the feature model;
A virtual disk procedure determination unit configured to determine a provisioning procedure of the virtual disk according to a search result of the reusable virtual disk search unit;
And a virtual disk creation unit for creating the virtual disk by copying the existing virtual disk according to the determined provisioning procedure of the virtual disk. The method according to claim 1,
Wherein the feature model comprises at least one of a name of a software including an application software and an operating system software, a version information package, a module, and an association relationship. The method according to claim 1,
Wherein the feature model generation unit comprises:
Generate a feature model for each of the subgroups including the plurality of virtual machines and generate a feature model associated with at least one homogeneity or heterogeneity based on the feature model. The method of claim 3,
The feature model associated with the homology or heterogeneity,
A feature model for software that overlaps the feature model for each of the subgroups, and a feature model for software that is not redundant. The method according to claim 1,
If the plurality of virtual machines includes at least one redundant software,
Wherein at least two virtual machines share a first disk including redundant software, the first disk that is shared is read only, and the files that are requested to be written are stored in a separate second disk of each of the virtual machines, Provisioning management device. The method according to claim 1,
Wherein the reusable virtual disk searching unit comprises:
And searches the existing virtual disk through an identifier obtained by applying a hash function to a feature model of the virtual disk to be created. The method according to claim 1,
Wherein the virtual disk procedure determination unit determines,
Wherein the virtual disk provisioning management apparatus determines a provisioning procedure of the virtual disk to be created based on an asset to be searched and duplicated when the existing virtual disk exists. The method according to claim 1,
Wherein the virtual disk procedure determination unit determines,
Wherein the virtual disk provisioning management apparatus searches for an installer of software to be installed and determines a provisioning procedure of the virtual disk based on the installer of the searched software. The method according to claim 1,
Wherein the virtual disk procedure determination unit determines,
Wherein the virtual disk provisioning management device determines the procedure for creating the virtual disk including each of the operating system virtual disk corresponding to the operating system software and the application software virtual disk corresponding to the application software. 10. The method of claim 9,
Wherein the procedure includes the name and version information of the operating system software. The method according to claim 1,
Wherein the virtual disk procedure determination unit determines,
And determines a procedure for creating one virtual disk corresponding to operating system software and application software. A virtual disk provisioning management method implemented by a processor,
Receiving a provisioning request for a plurality of virtual machines;
Generating at least one feature model for the plurality of virtual machines;
Searching an existing virtual disk having the same feature model to create a virtual disk based on the feature model;
Determining a provisioning procedure of the virtual disk according to the search result;
And creating the virtual disk by copying the existing virtual disk according to the determined provisioning procedure of the virtual disk. 13. The method of claim 12,
Wherein the feature model comprises at least one of a name of a software including application software and operating system software, a version information package, a module, and an association relationship. 13. The method of claim 12,
Wherein the step of generating the feature model comprises:
Generating a feature model for each of the subgroups including the plurality of virtual machines; And
And generating a feature model associated with at least one homogeneity or heterogeneity based on the feature model. 15. The method of claim 14,
The feature model associated with the homology or heterogeneity,
A feature model for software that overlaps the feature model for each of the subgroups, or a feature model for software that is not redundant. 13. The method of claim 12,
If the plurality of virtual machines includes at least one redundant software,
Wherein at least two virtual machines share a first disk including redundant software, the first disk that is shared is read only, and the files that are requested to be written are stored in a separate second disk of each of the virtual machines, How to Manage Provisioning. 13. The method of claim 12,
Wherein the searching of the existing virtual disk comprises:
And searching for the existing virtual disk through an integer identifier applying a hash function to a feature model of the virtual disk to be created. 13. The method of claim 12,
Wherein the step of determining a provisioning procedure of the virtual disk comprises:
If the existing virtual disk exists, searching for a location where the existing virtual disk is stored and determining a provisioning procedure of the virtual disk to be created based on an object to be duplicated and set. 13. The method of claim 12,
Wherein the step of determining a provisioning procedure of the virtual disk comprises:
Searching for an installer of the software to be installed; And
And determining a provisioning procedure of the virtual disk based on an installer of the searched software. 13. The method of claim 12,
Wherein the step of determining a provisioning procedure of the virtual disk comprises:
Determining a procedure for creating the virtual disk including each of the operating system virtual disk corresponding to the operating system software and the application software virtual disk corresponding to the application software. 21. The method of claim 20,
Wherein the procedure includes the name and version information of the operating system software. 13. The method of claim 12,
Wherein the step of determining a provisioning procedure of the virtual disk comprises:
Determining a procedure for creating one virtual disk corresponding to operating system software and application software.

Images