CN107783816A

CN107783816A - The method and device that creation method and device, the big data cluster of virtual machine create

Info

Publication number: CN107783816A
Application number: CN201610794476.6A
Authority: CN
Inventors: 费辉
Original assignee: Alibaba Group Holding Ltd
Current assignee: Alibaba Group Holding Ltd
Priority date: 2016-08-31
Filing date: 2016-08-31
Publication date: 2018-03-09

Abstract

A kind of method and device created this application discloses creation method of virtual machine, device, a kind of big data cluster.Wherein, the creation method of the virtual machine, including：Mirror image corresponding to being chosen according to the demand information to virtual machine from the mirror image being pre-configured with；The particular data application and configuration information of the corresponding demand information are integrated with the mirror image in advance；Virtual machine is created based on the mirror image.Using above-mentioned technical proposal when creating virtual machine, by using the mirror image establishment virtual machine for the particular data application and configuration information for being integrated with the corresponding demand information in advance, eliminate the process that corresponding data application and configuration data application are downloaded after virtual machine is created, it can directly make just to create the virtual machine offer calculating service completed, solves the problem of virtual machine creating consuming time length, waste of resource.

Description

The method and device that creation method and device, the big data cluster of virtual machine create

Technical field

The application is related to field of cloud computer technology, and in particular to a kind of creation method of virtual machine and a kind of large data sets The method that group creates；The device that the application is related to a kind of creating device of virtual machine simultaneously and a kind of big data cluster creates.

Background technology

With the arrival in data processing epoch, big data technology fast development, as the data processing technique hadoop that increases income, Hive, pig, hbase, spark etc. have obtained tremendous expansion, are increasingly widely used by industry.Using these big datas Handle software when, it is necessary to create big data cluster and to the big data cluster carry out with postpone could start accordingly service into The services such as row big data calculates, storage.

And at this stage, storage separates with calculating increasingly to be received by people, and storage can after being separated with calculating Effectively utilize resource.Under the background of cloud computing, the carrier of cluster application has been transferred to Virtual Cluster from physical cluster, Maximum of which improvement is exactly the use of virtual machine, and big data technology will calculate and referred to as elastic calculation is separated with storage, needing Cluster is created when being calculated and carries out calculating service；Terminate to discharge the resource created in cluster afterwards calculating.At present, on cloud Create elastic calculation cluster scheme be：Apply and start the virtual machine of the scale of calculating demand；Above and below each virtual machine Load demand big data processing software (such as：Hadoop, hive, pig, hbase, spark etc.)；When virtual machine download and install After complete corresponding big data processing software, each virtual machine is configured；Each started in current cluster is virtual The service processes of machine, the cluster virtual machine is provided corresponding calculate and service.

As can be seen here, under the scheme of existing establishment elastic calculation cluster, apply for that virtual machine builds big data on cloud Cluster starts virtual machine, downloads the steps such as big data processing software, configuration virtual machine, startup big data service, it is necessary to pass through, After the completion of above-mentioned steps, big data cluster could provide service.The establishment of cluster expends time length, and creates cluster in elasticity In, it is required for re-executing above-mentioned steps, waste of resource when creating cluster every time.

The content of the invention

It is existing to solve this application provides a kind of method that a kind of creation method of virtual machine and big data cluster create There is the above mentioned problem in technology；The application provides a kind of creating device of virtual machine simultaneously and a kind of big data cluster creates Device.

This application provides a kind of creation method of virtual machine, the creation method of the virtual machine, including：

Mirror image corresponding to being chosen according to the demand information to virtual machine from the mirror image being pre-configured with；In the mirror image in advance It is integrated with the particular data application and configuration information of the corresponding demand information；

Virtual machine is created based on the mirror image.

Optionally, the demand information to virtual machine, including：

The information of the service type of virtual machine offer is provided.

Optionally, the service type, is comprised at least：Data calculate service.

Optionally, mirror image corresponding to being chosen in the basis to the demand information of virtual machine from the mirror image being pre-configured with Before step, including：

Create the mirror image of the particular data application for being integrated with corresponding demand information and configuration information.

Optionally, the establishment is integrated with the particular data application of corresponding demand information and the mirror image of configuration information, bag Include：

Create virtual machine；

Operative configuration is carried out to the particular data application of corresponding demand information in the virtual machine；

The mirror image of presently described virtual machine is obtained, should using the mirror image as the specific data for being integrated with corresponding demand information With and configuration information mirror image.

Optionally, the particular data application progress operative configuration in the virtual machine to corresponding demand information is： Operative configuration is carried out to particular data application using shell script in the virtual machine.

Optionally, it is described to particular data application progress operative configuration, including：

Download particular data application and configurations are carried out to the data application.

Optionally, it is described to download particular data application and configurations are carried out to the data application, including：

Download an at least data application and each data application is configured.

Optionally, after the step of mirror image of the presently described virtual machine of acquisition, including：

By the mirrored storage of the current virtual machine of acquisition in database.

Optionally, when carrying out operative configuration to particular data application using shell script in the virtual machine, if described The configurations of data application or the data application are updated, then change the order in the shell script.

Optionally, it is described modification shell script in order the step of after, including：

Order in the shell script of record modification.

Accordingly, present invention also provides a kind of creating device of virtual machine, the creating device of the virtual machine, including：

Mirror image chooses unit, for the mirror according to corresponding to the demand information to virtual machine is chosen from the mirror image being pre-configured with Picture；The particular data application and configuration information of the corresponding demand information are integrated with the mirror image in advance；

Virtual machine creating unit, for creating virtual machine based on the mirror image.

Optionally, the mirror image acquiring unit, specifically for according to making the information of service type that virtual machine provides from pre- Mirror image corresponding to being chosen in the mirror image first configured.

Optionally, the mirror image acquiring unit, at least be used for according to make virtual machine provide data calculate service information from Mirror image corresponding to being chosen in the mirror image being pre-configured with.

Optionally, the creating device of the virtual machine, in addition to：Mirror image creating unit；

The mirror image creating unit, for being selected in the basis from the mirror image being pre-configured with the demand information of virtual machine Before mirror image corresponding to taking, the mirror image of the particular data application for being integrated with corresponding demand information and configuration information is created.

Optionally, the mirror image creating unit, including：

Virtual machine creating subelement, for creating virtual machine；

Data application configures subelement, for being carried out in the virtual machine to the particular data application of corresponding demand information Operative configuration；

Mirror image generate subelement, for obtaining the mirror image of presently described virtual machine, using the mirror image as be integrated with correspondence The particular data application of demand information and the mirror image of configuration information.

Optionally, the data application configuration subelement, specifically for using shell script in the virtual machine to spy Determine data application and carry out operative configuration.

Optionally, the data application configuration subelement, specifically for being downloaded in the virtual machine using shell script Particular data application and to the data application carry out configurations.

Optionally, the data application configuration subelement, specifically for being downloaded in the virtual machine using shell script At least one data application and each data application is configured.

Optionally, the creating device of the virtual machine, in addition to：

Mirror image memory cell, for after the mirror image for obtaining presently described virtual machine, by the current virtual of acquisition The mirrored storage of machine is in database.

Optionally, the data application configuration subelement, specifically for using shell script in the virtual machine to spy When determining data application progress operative configuration, if the configurations of the data application or the data application are updated, Change the order in the shell script.

Optionally, the creating device of the virtual machine, in addition to：

Command record unit, after the order in the modification shell script, the script journey of record modification Order in sequence.

Accordingly, present invention also provides the side that a kind of method that big data cluster creates, the big data cluster create Method, including：

According to the demand information of cluster, the mirror image corresponding to selection from the mirror image being pre-configured with；Collect in advance in the mirror image Into the particular data application and configuration information of the demand information for having corresponding cluster；

Start the virtual machine created in cluster based on the mirror image.

Optionally, the demand information of the cluster, including：

The information of service type and the calculation scale of cluster that the cluster is provided.

Optionally, in the demand information according to cluster, the step of mirror image corresponding to selection from the mirror image being pre-configured with Before rapid, including：

Create the particular data application of demand information and the mirror image of configuration information for being integrated with corresponding cluster.

Optionally, the establishment is integrated with the particular data application of demand information and the mirror of configuration information of corresponding cluster The mirror image of picture, including：

Create virtual machine；

Optionally, it is described that operative configuration, bag are carried out to the particular data application of corresponding demand information in the virtual machine Include：

Download an at least data application and each data application is configured.

Optionally, it is described start cluster in based on the mirror image create virtual machine when, including：

Main and subordinate node is configured in the cluster.

Optionally, it is described to configure main and subordinate node in the cluster, including：

Virtual machine is chosen in the cluster；

Using the virtual machine as host node, using other virtual machines in the cluster as from node；

The virtual machine in the cluster is set to update the characteristic information of main and subordinate node.

Optionally, after the step of virtual machine created in the startup cluster based on the mirror image, including：

Start the service processes of data application in each virtual machine of the cluster.

Discharge the virtual machine in the cluster.

Accordingly, present invention also provides the dress that the device that a kind of big data cluster creates, the big data cluster create Put, including：

Virtual machine creating unit, for the demand information according to cluster, the mirror corresponding to selection from the mirror image being pre-configured with Picture；The particular data application and configuration information of the demand information of corresponding cluster are integrated with the mirror image in advance；

Virtual machine start unit, for starting the virtual machine created in cluster based on the mirror image.

Optionally, the virtual machine creating unit, specifically for the information according to the service type for making the cluster offer And the calculation scale of cluster, the mirror image corresponding to selection from the mirror image being pre-configured with；It is integrated with advance in the mirror image correspondingly The particular data application and configuration information of the demand information of cluster.

Optionally, the device that described big data cluster creates, in addition to：

Mirror image creating unit, in the demand information according to cluster, the selection correspondence from the mirror image being pre-configured with Mirror image before, create the particular data application of demand information and the mirror image of configuration information for being integrated with corresponding cluster.

Optionally, the mirror image creating unit, including：

Virtual machine creating subelement, for creating virtual machine；

Optionally, the data application configuration subelement, specifically for downloading an at least data application and to each Data application is configured.

Optionally, the virtual machine start unit, specifically for starting the virtual machine created in cluster based on the mirror image When, configure main and subordinate node in the cluster.

Optionally, the virtual machine start unit, in addition to：

Virtual machine chooses subelement, for choosing virtual machine in the cluster；

Main and subordinate node generates subelement, for using the virtual machine as host node, by the cluster other are virtual Machine is used as from node；

Feature updates subelement, for making the virtual machine in the cluster update the characteristic information of main and subordinate node.

Process initiation unit, after the virtual machine based on mirror image establishment in the startup cluster, start institute State the service processes of data application in each virtual machine of cluster.

Virtual machine releasing unit, for starting the void created in the cluster according to the demand information of the cluster described After plan machine, the virtual machine in the cluster is discharged.

Compared with prior art, the application has advantages below：

Creation method, the device for a kind of virtual machine that the application provides, by according to the demand information of virtual machine from pre- Mirror image corresponding to being chosen in the mirror image first configured；The specific data for being integrated with the corresponding demand information in the mirror image in advance should With and configuration information；Virtual machine is created based on the mirror image.The technical scheme is when creating virtual machine, by using configuration Virtual machine is created to the mirror image of the application program for the demand information for tackling virtual machine, eliminates and downloads phase after virtual machine is created Answer application program and configure the process of application program, the virtual machine for just creating completion is provided the service of calculating, solution Virtual machine creating of having determined expends time length, the problem of waste of resource.

The method, apparatus that a kind of big data cluster that the application provides creates, according to the demand information of cluster, matches somebody with somebody from advance Mirror image corresponding to being chosen in the mirror image put；The particular data application of the demand information of corresponding cluster is integrated with the mirror image in advance And configuration information；Start the virtual machine created in cluster based on the mirror image.The technical scheme creates big data on cloud During cluster, virtual machine is created by using the mirror image for the application program for being configured with the demand information to tackling virtual machine, eliminated Corresponding application programs are downloaded after virtual machine is created and configure the process of application program, make the virtual machine in cluster in startup Afterwards, it is only necessary to carry out that after easy configuration main and subordinate node calculating service can be provided, the establishment for solving cluster expends time length, The problem of waste of resource, the creation time of cluster is saved, improve the utilization ratio of resource.

Brief description of the drawings

, below will be to embodiment or existing in order to illustrate more clearly of the embodiment of the present application or technical scheme of the prior art There is the required accompanying drawing used in technology description to be briefly described, it should be apparent that, drawings in the following description are only this Some embodiments described in application, for those of ordinary skill in the art, other can also be obtained according to these accompanying drawings Accompanying drawing.

Fig. 1 shows the flow chart of the creation method of the virtual machine provided according to embodiments herein；

Fig. 2 shows the schematic diagram of the creating device of the virtual machine provided according to embodiments herein；

Fig. 3 shows the flow chart for the method that the big data cluster provided according to embodiments herein creates；

Fig. 4 shows the schematic diagram of the big data cluster provided according to embodiments herein；

Fig. 5 shows the schematic diagram for the device that the big data cluster provided according to embodiments herein creates.

Embodiment

It is below in conjunction with the accompanying drawings and specific real in order to be more clearly understood that the above-mentioned purpose of the application, feature and advantage Mode is applied the application is further described in detail.It should be noted that in the case where not conflicting, the implementation of the application Feature in example and embodiment can be mutually combined.

Many details are elaborated in the following description in order to fully understand the application.But the application can To be much different from other manner described here to implement, those skilled in the art can be in the feelings without prejudice to the application intension Similar popularization is done under condition, therefore, the application is not limited by following public specific implementation.

Embodiments herein provides a kind of method that a kind of creation method of virtual machine and big data cluster create； A kind of creating device of virtual machine and a kind of device of big data cluster establishment itself are please provided simultaneously.In following implementation It is described in detail one by one in example.

Under the background of cloud computing, the carrier of cluster application has been transferred to Virtual Cluster, maximum of which from physical cluster Improvement is exactly the use of virtual machine, and big data technology will calculate and referred to as elastic calculation is separated with storage, when needing to be calculated Create cluster and carry out calculating service；Terminate to discharge the resource created in cluster afterwards calculating.At present, elastometer is created on cloud Calculate cluster scheme be：Apply and start the virtual machine of the scale of calculating demand；The big of demand is downloaded on each virtual machine Data processing software spend time can it is very long (such as：It can be taken after the decompressions such as hadoop, hive, pig, hbase, spark At least 1GB memory space, under the preferable environment of network environment, big data processing software is downloaded on each virtual machine, Also the time of a few minutes is wanted, if network environment is poor, all virtual machine installations are wanted more than ten minutes or even spent more long Time)；When virtual machine download and after install corresponding big data processing software, each virtual machine is configured；Start The each virtual machine in current cluster, the virtual machine is performed corresponding calculate and service.As can be seen here, in existing wound Build under the scheme of elastic calculation cluster, apply for that virtual machine builds big data cluster, it is necessary to by starting virtual machine, downloading on cloud Big data processing software, configuration virtual machine, start the steps such as big data service, after the above step is finished, big data cluster is Service can be provided.The establishment of cluster expends time length, and in elasticity creates cluster, is required for again when creating cluster every time Perform above-mentioned steps, waste of resource.For this problem, the technical scheme that itself please be provided is when creating virtual machine, by making Virtual machine is created with the mirror image for the application program for being configured with the demand information to tackling virtual machine, is eliminated after virtual machine is created Download corresponding application programs and configure the process of application program, provide directly the virtual machine for just creating completion and calculate clothes Business, solves the problem of virtual machine creating consuming time length, waste of resource.Meanwhile created when creating big data cluster on cloud During the cluster of virtual machine, created by using the mirror image for the application program for being configured with the demand information to tackling virtual machine virtual Machine, eliminate and download corresponding application programs after virtual machine is created and configure the process of application program, make virtual in cluster Machine is after start-up, it is only necessary to carries out that after easy configuration main and subordinate node calculating service can be provided, solves the establishment of cluster Time length is expended, the problem of waste of resource, the creation time of cluster is saved, improves the utilization ratio of resource.

Before the specific steps of the present embodiment are described in detail, the understanding to the technical program for convenience, first to existing Big data cluster be briefly described.

Cloud computing (cloud computing) is a kind of calculation based on internet, is usually directed to and passes through internet To provide dynamic easily extension and the often resource of virtualization.In this way, shared software and hardware resources and information can be with Computer and other equipment are supplied on demand.

It is the operating system template that virtual machine starts in a kind of virtualization that mirror image (Image), which is also known as image, is taken based on cloud The snapshot at business device system disk a certain moment, the scope of application only system disk.One template can be quickly started up by mirror image The virtual machine changed.The mirror-image format of main flow has vhd, raw, qcow2 etc. at present.

Big data cluster refers to put together many virtual machines carries out same service together, In the view of client just as It is an only server.Cluster can carry out parallel computation so as to obtain very high calculating speed using multiple virtual machines.

In architecture of the virtual machine (Virtual Machine) in computer science, refer to a kind of special software, A kind of environment can be created between computer platform and terminal user, and terminal user is then created based on this software Environment operates software.It is two steps using cloud computing resources to start and stop virtual machine.Disposing virtual machine in cloud can To assemble a customized virtual machine image as desired.

Embodiments herein provides a kind of creation method of virtual machine.The creation method of the virtual machine is implemented for example Under：

Fig. 1 is refer to, it illustrates the flow chart of the creation method of the virtual machine provided according to embodiments herein.

The creation method of the virtual machine includes：

Step S101, chosen according to the demand information to virtual machine from the mirror image being pre-configured with corresponding to mirror image；It is described The particular data application and configuration information of the corresponding demand information integrated in advance in mirror image.

In the present embodiment, the basis demand information of virtual machine is chosen from the mirror image being pre-configured with corresponding to mirror Picture, it can realize in the following way：The calculating that determining the virtual machine according to the demand information of the virtual machine needs to provide takes The type of business, and choose from database the particular data application and configuration information of the demand information integrated corresponding in advance Mirror image.

It should be noted that the particular data application that the corresponding demand information integrated in advance is chosen from database And configuration information mirror image when, due to storing the multiple mirror images mirror image of demand information (corresponding each) in the database, So the type for the calculating service that the virtual machine needs to provide can be determined to database by the demand information of the virtual machine The mirror image of middle storage is screened, and rejects the corresponding mirror image calculated outside service, and show remaining corresponding institute to user State the mirror image of the service of calculating, and receive user to the selection operation of the mirror image that is shown (such as：Hadoop mirror images from displaying, Selected in spark mirror images, hive mirror images).

It should be noted that the demand information of the virtual machine is：The meter for needing to provide by the virtual machine of user's input Calculate the type of service.By providing a user the input interface of service, the demand information of the virtual machine of user's input is obtained. Such as：The virtual machine of user's input needs the type of the calculating service provided to be calculated for data.

It should be understood that the virtual machine needs the type of the calculating service provided to include but is not limited to：Using Design, application and development, using test, AH, mail service, Web conference and fax through internet.

In the present embodiment, if the virtual machine of user's input needs the type of the calculating service provided to be calculated for data When, it is corresponding, in the particular data application integrated in advance of acquisition and the mirror image of configuration information, integrated particular data application Just it is data handling utility.The data handling utility comprises at least one kind in following data handling utility：Hadoop, Hive, pig, hbase, spark.

It should be noted that due to being downloaded in the mirror image of the particular data application integrated in advance and configuration information and The species of the data application of configuration and the difference of quantity, a variety of corresponding same requirements can be formed, but the data application configured is different Mirror image, so the mirror image of the integrated particular data application also has signature identification (such as specific data application title etc. The information that can uniquely represent), the naming rule of the mirror image of the integrated particular data application is used as by signature identification, to protect Card can distinguish the mirror image of each integrated particular data application.

Step S103, virtual machine is created based on the mirror image.

In the present embodiment, it is described that virtual machine is created based on the mirror image, it can realize in the following way：It is to work as first The corresponding computing resource of preceding virtual machine distribution (such as：The resources such as CPU, internal memory and network are distributed for current virtual machine), and set The Hostname of current virtual machine and ip addresses are put, then installation system and has been configured with corresponding data application by specifying one Mirror image according to the installation procedure of virtual machine, loaded by virtual platform and virtual machine be progressively installed, it is empty when reaching input During the interface of plan machine login password, the login password as corresponding to keeper or user's input, the establishment of virtual machine is completed.

As a preferred embodiment, in the creation method for the virtual machine that the embodiment of the present application provides, due in step Rapid S101 is chosen according to the demand information to virtual machine from the mirror image being pre-configured with corresponding to mirror image when, be to be integrated according to advance The particular data application of the corresponding demand information creates virtual machine, so performing step S101 according to the need to virtual machine Ask information chosen from the mirror image being pre-configured with corresponding to be integrated with corresponding each demand information, it is necessary to be pre-created before mirror image Particular data application and configuration information mirror image, the establishment be integrated with corresponding demand information particular data application and The mirror image of configuration information, it can realize in the following way：It is to the virtual machine for being integrated with data application and configuration information Disk of uniting carries out snapshot, the data in the system disk at the virtual machine current time is copied, the system that will be copied out Data in disk are as integrated corresponding data application and the mirror image of configuration information.

Specifically, the establishment is integrated with the particular data application of corresponding demand information and the mirror image of configuration information, bag Include following steps：

Create virtual machine；

Wherein, the establishment virtual machine, can be realized in the following way：Virtual machine is created using foundation image.

It should be noted that the foundation image (base image) includes：The mirror of the operating systems such as centos, ubuntu As file.It is that the virtual machine distributes corresponding calculating money first specifically, during the establishment virtual machine using foundation image Source, the environment in the virtual machine is then built by foundation image file, according to the installation procedure of virtual machine, progressively installation is empty Plan machine, when reaching the interface of input virtual machine entry password, the login password as corresponding to keeper or user's input, complete empty The establishment of plan machine.

Wherein, operative configuration is carried out to the particular data application of corresponding demand information in the virtual machine, can passed through Operative configuration is carried out to data application using shell script in the virtual machine.Such as：Using shel l shell scripts in institute State in virtual machine and operative configuration is carried out to data application.

It should be noted that shell shell scripts are similar to the batch processing under Windows/Dos, that is, with all kinds of lives Order is placed in advance in into a file, the convenient program file disposably performed, mainly facilitates keeper to be configured Or management.But because it has used the order under Linux/Unix, so more stronger than the batch processing under Windows Greatly, it is higher than the program efficiency with other programs editor.

Specifically, it is described to data application progress operative configuration, including：

Download the data application and configurations are carried out to the data application.

It should be noted that when carrying out operative configuration to data application, it can download and match somebody with somebody in the virtual machine Put a kind of data application；Or the combination of any number of data application is downloaded and configured in the virtual machine.Such as：It is right When virtual machine needs the type of the calculating service provided to create mirror image for data calculating, the application program of configuration is just hadoop, One kind in the data processing softwares such as hive, pig, hbase, spark, then hadoop is downloaded and configured in the virtual machine Software；Or hadoop softwares and spark softwares are downloaded and configured in the virtual machine.

Refer to it should be understood that carrying out configurations to the data application downloaded in the virtual machine：To the number According to building for application runtime environment.

Wherein, the mirror image for obtaining presently described virtual machine, can be realized in the following way：To the system of virtual machine Disk carries out snapshot, and the data in the system disk at the virtual machine current time are copied, obtain the virtual machine Mirror image.

It should be noted that the snapshot is：On specifying a completely available copy of data acquisition system, the copy includes Image of the corresponding data at some time point (time point that copy starts).Snapshot can be a pair of the data represented by it One duplicate of sheet or data.The effect of snapshot mainly can carry out backup and recovery.Work as storage device Quick data recovery can be carried out when generation application and trouble or file corruption, some available time point by data recovery State, under user data mistake or lost condition, to carry out data rewind to the data mode of the last snapshot.

It should be understood that because the virtual machine is created by foundation image, and after the completion of establishment Downloading data is applied and the running environment of the data application is configured in the virtual machine, so when current Carve after carrying out snapshot to the system disk of virtual machine, be to get downloading data application and the operation to the data application The mirror image of the system disk for the virtual machine that environment is configured, i.e.,：After foundation image document creation virtual machine, from described The mirror image obtained on virtual machine is the increment mirror image based on the foundation image file generated, and using the increment mirror image as collection Into the particular data application and the mirror image of configuration information for having corresponding demand information.

As a preferred embodiment, in the creation method for the virtual machine that the embodiment of the present application provides, due to data Consider not comprehensive or program function imperfection in the period of writing, designer using application program, after data application distribution, meeting New function is modified or added to data application by the form of patch, make data application from lowest version to highest version more Newly, so after data application is updated, the size of data application and the configurations of data application can change, then When carrying out operative configuration to data application using shell script in the virtual machine, if the data application or the data should Configurations are updated, then change the order in the shell script.

It should be noted that when carrying out operative configuration to data application using shell script, it is placed in advance in by performing Being downloaded to data application in shell script and the order configured to data application, make shell script automatically exist Operative configuration is carried out to data application in the virtual machine.Due to be placed in advance in shell script be download lowest version number According to the order of application, so after data application is updated, the data application after renewal can not be then downloaded using shell script.

Such as：After data application carries out version updating, change and data application is downloaded and matched somebody with somebody in shell script The order put：The order for being downloaded original version data application in shell script is revised as to download redaction data application Order (such as：Replace download command described in data application version number information), by shell script to original version number The order configured according to application is revised as configuring the order of redaction data application.

As a preferred embodiment, in the creation method for the virtual machine that the embodiment of the present application provides, when data should , can be by shell script after changing the order that data application is downloaded and configured in shell script with version updating is carried out The content of the order of middle modification is recorded.Specifically, can use Project Management Software (such as：Svn, git) record script The change process of program, it is convenient to investigate mistake.

As a preferred embodiment, in the creation method for the virtual machine that the embodiment of the present application provides, institute is being obtained After the mirror image for stating virtual machine, due to the data application species and the difference of quantity downloading and configure in the mirror image, it can form A variety of mirror images, so needing to be stored in database after setting signature identification to the mirror image of acquisition.

It should be noted that be referred to as the naming rule of mirror image with the name of data application, will be downloaded in the mirror image and The name of the data application of configuration is referred to as the signature identification of the mirror image, with guarantee to distinguish be stored in it is each default in database Mirror image.Such as：When the type for needing the calculating provided to service virtual machine is that data calculating creates mirror image, corresponding data should With for data processing software, then the signature identification for downloading and configuring hadoop mirror image is hadoop, download and configures spark's The signature identification of mirror image is hadoop, download and to configure the signature identification of hive and pig mirror image be hadoop, hive.

The creation method of a kind of virtual machine provided by the embodiment of the present application, when creating virtual machine, by using pre- First it is integrated with the particular data application of the corresponding demand information and the mirror image of configuration information creates virtual machine, eliminates and creating The process that corresponding data is applied and configuration data is applied is downloaded after building virtual machine, can directly make just to create the virtual machine completed The service of calculating is provided, solves the problem of virtual machine creating consuming time length, waste of resource.

In the above-described embodiment, there is provided a kind of creation method of virtual machine, the creation method phase with above-mentioned virtual machine Corresponding, present invention also provides a kind of creating device of virtual machine.Because the embodiment of device is substantially similar to the reality of method Example is applied, so describing fairly simple, related part illustrates referring to the part of the creation method embodiment of virtual machine.It is following The creating device embodiment of the virtual machine of description is only schematical.The creating device embodiment of the virtual machine is as follows：

Fig. 2 is refer to, it illustrates the schematic diagram of the creating device of the virtual machine provided according to embodiments herein.

The creating device of the virtual machine, including：Mirror image acquiring unit 201 and virtual machine creating unit 203；

The mirror image chooses unit 201, for being chosen according to the demand information to virtual machine from the mirror image being pre-configured with Corresponding mirror image；The particular data application and configuration information of the corresponding demand information are integrated with the mirror image in advance；

The virtual machine creating unit 203, for creating virtual machine based on the mirror image.

Optionally, the mirror image acquiring unit 201, specifically for according to make virtual machine provide service type information from Mirror image corresponding to being chosen in the mirror image being pre-configured with.

Optionally, the mirror image acquiring unit 201, at least it is used for according to the letter for making virtual machine provide data calculating service Breath mirror image corresponding to selection from the mirror image being pre-configured with.

Optionally, the creating device of the virtual machine, it is characterised in that also include：Mirror image creating unit；

Optionally, the mirror image creating unit, including：Virtual machine creating subelement, data application configuration subelement and Mirror image generates subelement；

The virtual machine creating subelement, for creating virtual machine；

The data application configures subelement, in the virtual machine to the particular data application of corresponding demand information Carry out operative configuration；

The mirror image generates subelement, for obtaining the mirror image of presently described virtual machine, using the mirror image as being integrated with The particular data application of corresponding demand information and the mirror image of configuration information.

Optionally, the creating device of the virtual machine, in addition to：Mirror image memory cell；

The mirror image memory cell, for after the mirror image for obtaining presently described virtual machine, by the current of acquisition The mirrored storage of virtual machine is in database.

Optionally, the creating device of the virtual machine, in addition to：Command record unit；

The command record unit, after the order in the modification shell script, the pin of record modification Order in this program.

In the above-described embodiment, there is provided a kind of creation method of virtual machine and a kind of creating device of virtual machine, In addition, present invention also provides a kind of method that big data cluster creates；The embodiment of the method that the big data cluster creates is such as Under：

Fig. 3 and Fig. 4 are refer to, the method created it illustrates the big data cluster provided according to embodiments herein The schematic diagram for the big data cluster that flow chart and embodiments herein provide.

The method that the big data cluster creates, including：

Step 301, according to the demand information of cluster, the mirror image corresponding to selection from the mirror image being pre-configured with；The mirror image In be integrated with advance corresponding cluster demand information particular data application and configuration information.

In the present embodiment, the demand information of the cluster is：By the service class for making the cluster offer of user's input The information of type and the calculation scale of cluster.By providing a user the input interface of service, the collection of user's input is obtained The demand information of group.Such as：The type that the virtual machine of user's input needs the calculating provided to service is calculated for data and cluster Calculation scale be 10 virtual machines.

Need what is illustrated, the calculation scale of the cluster can input the size of amount of calculation scale, and will meter by backstage Calculation gauge mould is converted to corresponding virtual machine quantity or rule of thumb directly inputted by those skilled in the art and needed The quantity for the virtual machine asked.

In the present embodiment, the demand information according to cluster, the mirror image corresponding to selection from the mirror image being pre-configured with, It can realize in the following way：Determine that the virtual machine needs the class for calculating service provided according to the demand information of the cluster Type, and chosen from database and integrate the particular data application of the corresponding demand information and the mirror image of configuration information in advance.

It should be noted that the particular data application that the corresponding demand information integrated in advance is chosen from database And configuration information mirror image when, due to storing the multiple mirror images mirror image of demand information (corresponding each) in the database, So the type for the calculating service that the cluster needs to provide can be determined to being deposited in database by the demand information of the cluster The mirror image of storage is screened, and rejects the corresponding mirror image calculated outside service, and show the remaining correspondence meter to user Calculate the mirror image of service, and receive user to the selection operation of the mirror image that is shown (such as：From hadoop mirror images, the spark of displaying Selected in mirror image, hive mirror images).

It should be understood that the cluster needs the type of the calculating service provided to include but is not limited to：Using setting Meter, application and development, using test, AH, mail service, Web conference and fax through internet.

In the present embodiment, if the cluster of user's input needs the type of the calculating service provided to be calculated for data, Corresponding, in the particular data application integrated in advance of acquisition and the mirror image of configuration information, integrated particular data application is just For data handling utility.The data handling utility comprises at least one kind in following data handling utility：Hadoop, hive, Pig, hbase, spark.

Step S303, start the virtual machine created in cluster based on the mirror image.

In the present embodiment, the virtual machine created based on the mirror image is collected in the demand information according to the cluster The calculation scale of group, number corresponding with the calculation scale of the cluster is created based on the mirror image obtained in step S301 Virtual machine.

It should be understood that due to cluster be have a number of virtual robot arm into, so when creating big data cluster, The number of the virtual machine created in the cluster is often both greater than one, so being parallel create virtually when creating virtual machine Machine.

In the specific implementation, the virtual machine based on mirror image establishment is：According to the input for the service that provides a user User's input that interface is got makes the information of service type and the calculation scale of cluster that the cluster provides, and determines institute The calculating service of the number and the cluster demand of the virtual machine for needing to create in cluster is stated, and according to each in the cluster Particular data application and the configuration of the demand information of corresponding cluster are integrated with corresponding to the particular use selection of virtual machine in advance The mirror image of information, virtual machine is created according to the mirror image.

Such as：The number for the virtual machine for determining to need to create in the cluster from the demand information of cluster is two, if with When the virtual machine of family input needs the type of the calculating service provided to be calculated for data, and according to virtual machine in the cluster Particular use determines have a virtual machine selection to download and configured hadoop default mirror image in the cluster, described There is a virtual machine selection to download and configured spark default mirror image in cluster, then basis has been downloaded and configured hadoop Default mirror image create a virtual machine, according to downloaded and configure spark default mirror image create a virtual machine.

In the present embodiment, it is described to start the virtual machine created in cluster based on the mirror image, can be in the following way Realize：Start the virtual machine created in the cluster according to the demand information of the cluster using virtualization technology.

It should be noted that the virtualization technology (Virtualization) is a kind of resource management techniques, it is to count The various actual resources of calculation machine, such as server, network, internal memory and storage, be abstracted, change after show, break reality The interstructural not cleavable obstacle of body, allows user than the configuration better way of script to apply these resources.Virtually Change technology transparence bottom physical hardware, so as to maximumlly utilize physical hardware.

In the present embodiment, KVM (Kernel-based Virtual Machine) can be used to start in the cluster The virtual machine that demand information according to the cluster creates.KVM is a kind of virtualization technology, a system virtualization mould increased income Block, it is integrated in Linux each main release version.It is managed using Linux itself scheduler, so relative to Xen, its core source code are seldom.KVM is at present as one of main flow VMM of academia.KVM virtualization needs hardware supported, It is hardware based Full-virtualization.

In the present embodiment, in the virtual machine based on mirror image establishment in starting cluster, by the virtual machine institute created In the cluster of composition, it is that each virtual machine of the establishment in the cluster links together, it is rendered as user single Computer, the virtual machine connected in a cluster can with share workload, realize single-point operation/management.

Due to being separate between each virtual machine in the cluster, so needing to match somebody with somebody the cluster Put, it is described configuration is carried out to the cluster to refer to：Main and subordinate node is configured in the cluster.It is described that master is configured in the cluster From node, can realize in the following way：It is respectively main and subordinate node by the virtual machine configuration in cluster.

It should be noted that the host node (master) is responsible for being communicated with from node, decomposition computation service is simultaneously Being calculated from node distribution in the cluster services, the calculating for being responsible for performing the host node distribution from node (salve) Service, and the result of calculating is reported to the host node.

It is described to configure main and subordinate node in the cluster, specifically comprise the following steps：

Virtual machine is chosen in the cluster；

Wherein, it is described to choose virtual machine in the cluster, it can realize in the following way：Arbitrarily chosen in the cluster One virtual machine.

It should be noted that when choosing virtual machine in the cluster, a virtual machine can also be additionally being chosen, by this Virtual machine is as the standby virtual machine of host node.When host node virtual machine delays machine, start the virtual machine as standby host node.

Wherein, it is described using the virtual machine as host node, can using other virtual machines in the cluster as from node To realize in the following way：The virtual machine configuration that the cluster is chosen is host node, is then believed according to the demand of the cluster Breath by remaining virtual machine configuration in the cluster be calculate node (i.e.：From node).

Such as：10 virtual machines are created in the cluster, then are virtual at this when being host node by the virtual machine configuration NameNode and JobTracker role is configured in machine, is responsible for the execution of house steward's data message and decomposition computation service；Then Will in the cluster remaining virtual machine (9) be configured to from node when, be to configure DataNode in remaining virtual machine With TaskTracker role, it is responsible for data storage and performs the calculating service of the host node distribution.

Wherein, the characteristic information of the main and subordinate node in the renewal cluster, can be realized in the following way：Obtain The characteristic information of host node in the cluster and the characteristic information from node, by the characteristic information of the main and subordinate node of acquisition In all nodes in synchronized update to the cluster.

It should be noted that the characteristic information refers to：Hostname, ip addresses and the Hostname of current virtual machine With the corresponding relation between ip addresses.Configured in the cluster after main and subordinate node, it is necessary to configure the Hostname of each node With determining ip addresses, and the corresponding relation between each node host title and ip addresses is recorded, such as：Match somebody with somebody in the cluster A host node and two are put from node, and the data processing software downloaded and configured in each node is hadoop, then will The Hostname of host node is configured to Master.Hadoop and determines that the ip addresses of host node are 192.168.1.141, by two Salve1.Hadoop and Salve2.Hadoop are configured to from the Hostname of node and determination is from the ip addresses of node And 192.168.1.143 192.168.1.142.

Specifically, if the operating system of virtual machine is linux, when configuring host node or the Hostname from node, Hostname is changed under Ubuntu, the value in file/etc/hostname is changed with regard to Hostname can be changed, is used after successfully modified Hostname orders check whether current hosts title sets success；When configuring the ip addresses of each node, if virtual machine is adopted Network is connected with NAT modes, then can distribute the ip addresses of each node automatically.

As a preferred embodiment, in the creation method for the virtual machine that the embodiment of the present application provides, due in step When rapid S303 starts the virtual machine based on mirror image establishment in cluster, believed according to the demand of the corresponding cluster integrated in advance The particular data application of breath creates virtual machine, so performing step S301 according to the demand information of cluster from being pre-configured with Corresponding to being chosen in mirror image before mirror image, it is necessary to be pre-created the particular data application of the demand information for being integrated with corresponding cluster with And the mirror image of configuration information.

In the present embodiment, the demand information of corresponding cluster is integrated with due to being pre-created in the big data cluster on cloud Particular data application and the process of mirror image of configuration information be integrated with being pre-created in the creation method of above-mentioned virtual machine The particular data application of corresponding demand information and the process of the mirror image of configuration information are essentially identical, due in preceding embodiment Detailed description is compared to this, here is omitted.

As a preferred embodiment, in the method that the big data cluster that the embodiment of the present application provides creates, in step After starting the virtual machine created in cluster based on the mirror image described in rapid S303, the cluster just can be according to the cluster demand Calculating service calculated accordingly.After the step of virtual machine created in the startup cluster based on the mirror image, Including：Start the service processes of data application in each virtual machine of the cluster.

If it should be understood that corresponding data processing software is hadoop in virtual machine in the cluster, start Hadoop service processes；If corresponding data processing software is hive in the virtual machine in the cluster, start hive clothes Business process；If corresponding data processing software is spark in the virtual machine in the cluster, start spark service processes；If Corresponding data processing software is hadoop and spark in virtual machine in the cluster, then starts hadoop service simultaneously The service processes of process and spark.

It should be noted that start the service processes of corresponding data processing software in each virtual machine of the cluster, Each calculating service that host node distribution can be received from node in the cluster, and holds each from node according to service processes Row is corresponding to be calculated, and the result of calculating is reported to the host node.

As a preferred embodiment, in the method that the big data cluster that the embodiment of the present application provides creates, work as institute After stating the calculating service that cluster completes offer, the result for performing calculating is saved in specified path by the cluster, and is shown and held The result that row calculates, and discharge the virtual machine in the cluster.

It should be understood that in the virtual machine for starting and being created in the cluster according to the mirror image, and complete to calculate clothes After business, in order to reduce the resource consumption shared by the virtual machine in cluster, the resource created in cluster can be discharged, i.e.,：Release The resource that each virtual machine in cluster is created.

The method that a kind of big data cluster provided by the embodiment of the present application creates, creates big data cluster on cloud When, virtual machine is created by using the mirror image for the application program for being configured with the demand information to tackling virtual machine, eliminates and is creating Build after virtual machine and download corresponding application programs and configure the process of application program, make virtual machine in cluster after start-up, Only need that calculating service can be provided after carrying out easy configuration main and subordinate node, the establishment for solving cluster expends time length, wave The problem of taking resource, the creation time of cluster is saved, improve the utilization ratio of resource.

In the above-described embodiment, there is provided a kind of method that big data cluster creates, created with above-mentioned big data cluster Method it is corresponding, present invention also provides a kind of big data cluster create device.Due to the basic phase of the embodiment of device Be similar to the embodiment of method, so describe fairly simple, related part referring to the creation method embodiment of virtual machine part Explanation.The device embodiment that big data cluster described below creates is only schematical.The big data cluster wound The device embodiment built is as follows：

Fig. 5 is refer to, the signal of the device created it illustrates the big data cluster provided according to embodiments herein Figure.

The device that the big data cluster creates, including：Virtual machine creating unit 501 and virtual machine start unit 503；

The virtual machine creating unit 501, for the demand information according to cluster, the selection pair from the mirror image being pre-configured with The mirror image answered；The particular data application and configuration information of the demand information of corresponding cluster are integrated with the mirror image in advance；

The virtual machine start unit 503, for starting the virtual machine created in cluster based on the mirror image.

Optionally, the virtual machine creating unit 501, specifically for the letter according to the service type for making the cluster offer The calculation scale of breath and cluster, the mirror image corresponding to selection from the mirror image being pre-configured with；It is integrated with advance in the mirror image pair Answer the particular data application and configuration information of the demand information of cluster.

Optionally, the device that described big data cluster creates, in addition to：Mirror image creating unit；

The mirror image creating unit, in the demand information according to cluster, being chosen from the mirror image being pre-configured with Before corresponding mirror image, the particular data application of demand information and the mirror image of configuration information for being integrated with corresponding cluster are created.

The virtual machine creating subelement, for creating virtual machine；

Optionally, the virtual machine start unit 503, specifically for starting in cluster based on the virtual of mirror image establishment During machine, main and subordinate node is configured in the cluster.

Optionally, the virtual machine start unit 503, in addition to：

In a typical configuration, computing device includes one or more processors (CPU), input/output interface, net Network interface and internal memory.

Internal memory may include computer-readable medium in volatile memory, random access memory (RAM) and/or The forms such as Nonvolatile memory, such as read-only storage (ROM) or flash memory (flashRAM).Internal memory is showing for computer-readable medium Example.

1st, computer-readable medium can be by any side including permanent and non-permanent, removable and non-removable media Method or technology realize that information stores.Information can be computer-readable instruction, data structure, the module of program or other numbers According to.The example of the storage medium of computer includes, but are not limited to phase transition internal memory (PRAM), static RAM (SRAM), dynamic random access memory (DRAM), other kinds of random access memory (RAM), read-only storage (ROM), Electrically Erasable Read Only Memory (EEPROM), fast flash memory bank or other memory techniques, read-only optical disc are read-only Memory (CD-ROM), digital versatile disc (DVD) or other optical storages, magnetic cassette tape, tape magnetic rigid disk storage or Other magnetic storage apparatus or any other non-transmission medium, the information that can be accessed by a computing device available for storage.According to Herein defines, and computer-readable medium does not include non-temporary computer readable media (transitorymedia), such as modulation Data-signal and carrier wave.

2nd, it will be understood by those skilled in the art that embodiments herein can be provided as method, system or computer program production Product.Therefore, the application can use the embodiment in terms of complete hardware embodiment, complete software embodiment or combination software and hardware Form.Moreover, the application can use the computer for wherein including computer usable program code in one or more can use The computer program product that storage medium is implemented on (including but is not limited to magnetic disk storage, CD-ROM, optical memory etc.) Form.

Although the application is disclosed as above with preferred embodiment, it is not for limiting the application, any this area skill Art personnel are not being departed from spirit and scope, can make possible variation and modification, therefore the guarantor of the application Shield scope should be defined by the scope that the application claim is defined.

Claims

A kind of 1. creation method of virtual machine, it is characterised in that including：

Mirror image corresponding to being chosen according to the demand information to virtual machine from the mirror image being pre-configured with；Integrated in advance in the mirror image There are the particular data application and configuration information of the corresponding demand information；

Virtual machine is created based on the mirror image.
2. the creation method of virtual machine according to claim 1, it is characterised in that the demand information to virtual machine, Including：

The information of the service type of virtual machine offer is provided.
3. the creation method of virtual machine according to claim 2, it is characterised in that the service type, comprise at least：Number Serviced according to calculating.
4. the creation method of virtual machine according to claim 2, it is characterised in that the demand in the basis to virtual machine Information chosen from the mirror image being pre-configured with corresponding to mirror image the step of before, including：

Create the mirror image of the particular data application for being integrated with corresponding demand information and configuration information.
5. the creation method of virtual machine according to claim 3, it is characterised in that the establishment is integrated with corresponding demand letter The particular data application of breath and the mirror image of configuration information, including：

Create virtual machine；

Operative configuration is carried out to the particular data application of corresponding demand information in the virtual machine；

Obtain the mirror image of presently described virtual machine, using the mirror image as the particular data application for being integrated with corresponding demand information with And the mirror image of configuration information.
6. the creation method of virtual machine according to claim 5, it is characterised in that it is described in the virtual machine to corresponding The particular data application of demand information carries out operative configuration：Using shell script to particular data application in the virtual machine Carry out operative configuration.
7. the creation method of virtual machine according to claim 6, it is characterised in that described to be grasped to particular data application Configure, including：

Download particular data application and configurations are carried out to the data application.
8. the creation method of virtual machine according to claim 7, it is characterised in that it is described download particular data application and Configurations are carried out to the data application, including：

Download an at least data application and each data application is configured.
9. the creation method of virtual machine according to claim 5, it is characterised in that obtain presently described virtual machine described Mirror image the step of after, including：

By the mirrored storage of the current virtual machine of acquisition in database.
10. the creation method of virtual machine according to claim 6, it is characterised in that script is used in the virtual machine When program carries out operative configuration to particular data application, if the configurations of the data application or the data application are carried out Renewal, then change the order in the shell script.
11. the creation method of virtual machine according to claim 10, it is characterised in that in the modification shell script After the step of order, including：

Order in the shell script of record modification.
A kind of 12. creating device of virtual machine, it is characterised in that including：

Mirror image chooses unit, for the mirror image according to corresponding to the demand information to virtual machine is chosen from the mirror image being pre-configured with； The particular data application and configuration information of the corresponding demand information are integrated with the mirror image in advance；

Virtual machine creating unit, for creating virtual machine based on the mirror image.
13. the creating device of virtual machine according to claim 12, it is characterised in that the mirror image acquiring unit is specific to use According to make virtual machine provide service type information chosen from the mirror image being pre-configured with corresponding to mirror image.
14. the creating device of virtual machine according to claim 13, it is characterised in that the mirror image acquiring unit, at least use Mirror image corresponding to being chosen in the information serviced according to virtual machine offer data are calculated from the mirror image being pre-configured with.
15. the creating device of virtual machine according to claim 13, it is characterised in that also include：Mirror image creating unit；

The mirror image creating unit, for being chosen in the basis to the demand information of virtual machine from the mirror image being pre-configured with pair Before the mirror image answered, the mirror image of the particular data application for being integrated with corresponding demand information and configuration information is created.
16. the creating device of virtual machine according to claim 15, it is characterised in that the mirror image creating unit, including：

Virtual machine creating subelement, for creating virtual machine；

Data application configures subelement, for being operated in the virtual machine to the particular data application of corresponding demand information Configuration；

Mirror image generate subelement, for obtaining the mirror image of presently described virtual machine, using the mirror image as be integrated with correspondence demand The particular data application of information and the mirror image of configuration information.
17. the creating device of virtual machine according to claim 16, it is characterised in that the data application configures subelement, Specifically for using shell script to carry out operative configuration to particular data application in the virtual machine.
18. the creating device of virtual machine according to claim 17, it is characterised in that the data application configures subelement, Specifically for using shell script to download particular data application in the virtual machine and carrying out basis to the data application Configuration.
19. the creating device of virtual machine according to claim 17, it is characterised in that the data application configures subelement, Specifically for using shell script to download an at least data application in the virtual machine and each data application being carried out Configuration.
20. the creating device of virtual machine according to claim 16, it is characterised in that also include：

Mirror image memory cell, for after the mirror image for obtaining presently described virtual machine, by the current virtual machine of acquisition Mirrored storage is in database.
21. the creating device of virtual machine according to claim 17, it is characterised in that the data application configures subelement, During specifically for using shell script to carry out operative configuration to particular data application in the virtual machine, if the data application Or the configurations of the data application are updated, then the order in the shell script is changed.
22. the creating device of virtual machine according to claim 21, it is characterised in that also include：

Command record unit, after the order in the modification shell script, in the shell script of record modification Order.
23. a kind of method that big data cluster creates, it is characterised in that including：

According to the demand information of cluster, the mirror image corresponding to selection from the mirror image being pre-configured with；It is integrated with advance in the mirror image The particular data application and configuration information of the demand information of corresponding cluster；

Start the virtual machine created in cluster based on the mirror image.
24. the method that big data cluster according to claim 23 creates, it is characterised in that the demand letter of the cluster Breath, including：

The information of service type and the calculation scale of cluster that the cluster is provided.
25. the method that big data cluster according to claim 24 creates, it is characterised in that in the need according to cluster Seek information, from the mirror image being pre-configured with choose corresponding to mirror image the step of before, including：

Create the particular data application of demand information and the mirror image of configuration information for being integrated with corresponding cluster.
26. the method that big data cluster according to claim 25 creates, it is characterised in that the establishment is integrated with correspondingly The mirror image of the particular data application of the demand information of cluster and the mirror image of configuration information, including：

Create virtual machine；

Operative configuration is carried out to the particular data application of corresponding demand information in the virtual machine；

Obtain the mirror image of presently described virtual machine, using the mirror image as the particular data application for being integrated with corresponding demand information with And the mirror image of configuration information.
27. the method that big data cluster according to claim 26 creates, it is characterised in that described in the virtual machine Operative configuration is carried out to the particular data application of corresponding demand information, including：

Download an at least data application and each data application is configured.
28. the method that big data cluster according to claim 26 creates, it is characterised in that be based in the startup cluster During the virtual machine that the mirror image creates, including：

Main and subordinate node is configured in the cluster.
29. the method that big data cluster according to claim 28 creates, it is characterised in that described to match somebody with somebody in the cluster Main and subordinate node is put, including：

Virtual machine is chosen in the cluster；

Using the virtual machine as host node, using other virtual machines in the cluster as from node；

The virtual machine in the cluster is set to update the characteristic information of main and subordinate node.
30. the method that the big data cluster according to claim 23-29 any one creates, it is characterised in that described After the step of starting the virtual machine based on mirror image establishment in cluster, including：

Start the service processes of data application in each virtual machine of the cluster.
31. the method that big data cluster according to claim 30 creates, it is characterised in that the base in the startup cluster After the step of virtual machine that the mirror image creates, including：

Discharge the virtual machine in the cluster.
A kind of 32. device that big data cluster creates, it is characterised in that including：

Virtual machine creating unit, for the demand information according to cluster, the mirror image corresponding to selection from the mirror image being pre-configured with；Institute State the particular data application and configuration information for the demand information for being integrated with corresponding cluster in mirror image in advance；

Virtual machine start unit, for starting the virtual machine created in cluster based on the mirror image.
33. the device that big data cluster according to claim 32 creates, it is characterised in that the virtual machine creating list Member, the information of service type and the calculation scale of cluster of the cluster offer are made specifically for basis, from what is be pre-configured with Mirror image corresponding to being chosen in mirror image；Be integrated with advance in the mirror image demand information of corresponding cluster particular data application and Configuration information.
34. the device that big data cluster according to claim 33 creates, it is characterised in that also include：

Mirror image creating unit, in the demand information according to cluster, the mirror corresponding to selection from the mirror image being pre-configured with Before picture, the particular data application of demand information and the mirror image of configuration information for being integrated with corresponding cluster are created.
35. the device that big data cluster according to claim 34 creates, it is characterised in that the mirror image creating unit, Including：

Virtual machine creating subelement, for creating virtual machine；

Data application configures subelement, for being operated in the virtual machine to the particular data application of corresponding demand information Configuration；

Mirror image generate subelement, for obtaining the mirror image of presently described virtual machine, using the mirror image as be integrated with correspondence demand The particular data application of information and the mirror image of configuration information.
36. the device that big data cluster according to claim 35 creates, it is characterised in that data application configuration Unit, specifically for downloading an at least data application and being configured to each data application.
37. the device that big data cluster according to claim 36 creates, it is characterised in that the virtual machine starts single Member, specifically for start cluster in based on the mirror image create virtual machine when, configure main and subordinate node in the cluster.
38. the device that big data cluster according to claim 36 creates, it is characterised in that the virtual machine starts single Member, in addition to：

Virtual machine chooses subelement, for choosing virtual machine in the cluster；

Main and subordinate node generates subelement, for using the virtual machine as host node, other virtual machines in the cluster to be made For from node；

Feature updates subelement, for making the virtual machine in the cluster update the characteristic information of main and subordinate node.
39. the device that the big data cluster according to claim 32-38 any one creates, it is characterised in that also include：

Process initiation unit, after the virtual machine based on mirror image establishment in the startup cluster, start the collection The service processes of data application in each virtual machine of group.
40. the device that the big data cluster according to claim 39 creates, it is characterised in that also include：

Virtual machine releasing unit, for starting the virtual machine created in the cluster according to the demand information of the cluster described Afterwards, the virtual machine in the cluster is discharged.