TWI693550B - Backup method for supporting multi-dsik architecture of virtual machine - Google Patents

Abstract

The present invention provides a backup method for supporting a multi-disk architecture of virtual machine, the method comprising: setting a project name, a maximum number of backups, and a selected virtual disk at the time of backup to establish a backup plan; establishing a group for each of virtual system disks and determining whether the virtual system disk of the virtual machine to which the virtual data disk in the backup plan belongs is in a group, so that placing the virtual data disk in the same group as the virtual system disk when the virtual system disk and the virtual data disk belongs the same group; creating a corresponding number of threads based on the number of groups, and the backup plan for a single group is performed by a single thread; creating and storing a virtual disk snapshot of the virtual machine; and recording the virtual disk snapshot creation process into an execution history when the execution of the backup plan is completed.

Description

Backup method supporting multi-disk structure of virtual machine

The present invention relates to a virtual machine backup mechanism. In detail, it relates to a backup method that supports a multi-disk architecture of a virtual machine.

Virtual machine (virtual machine) refers to a special software that can create an environment between a computer platform and a user terminal. The virtual system can generate a virtual image of the operating system, and at the same time has similar functions as the real operating system. It can be executed in the virtual system. All operations can be independently installed and run software, save data, but will not have any impact on the real operating system.

The startup of the existing virtual machine is through the image file, that is, the image file records the settings required for the entire virtual machine, such as system status, memory status, data content, etc. Therefore, to start the virtual machine through the image file, when it can generate the required In the virtual machine environment, there is still room for improvement in the use of such image files. For example, the overall content of the image file is huge, resulting in a long backup time, which will cause the problem of the storage device continuing to be busy. Even if you want to perform a backup of an image file with a slightly different system status, you also face problems such as long backup time and insufficient storage space.

As can be seen from the above, based on the deficiency of the existing virtual machine backup mechanism, if you can find A virtual machine backup technology is developed, in particular, it is not necessary to back up the existing image files as a whole, so as to avoid the problem of too long backup time and insufficient storage space. At the same time, if automatic backup scheduling can be provided, the backup mechanism can be made more flexible This will become the technical goal that people in this technical field are striving for.

The purpose of the present invention is to shorten the backup time of a virtual machine, that is, to completely change the traditional virtual machine architecture booted from an image file to a virtual machine with a multi-disk architecture, so that users can select one or more disks as a backup plan according to their needs Backup target, and when performing backup, the group mechanism is used to avoid the problem that the same virtual machine is repeatedly suspended for multiple times due to the selection of multiple disks, so it can improve the quality of user operations.

In order to achieve the above or other objectives, the present invention provides a backup method that supports a multi-disk architecture of a virtual machine, which includes: setting a plan name during backup, the maximum number of backups, and the selected virtual disk to create a backup plan; Let each virtual machine's virtual system disk create a group to determine whether the virtual data disk in the backup plan and the virtual system disk of the virtual machine to which it belongs are in a group, so that the virtual data disk When the virtual system disk to which it belongs is already in the group, put the virtual data disk into the same group as the virtual system disk; create a corresponding number of threads according to the number of groups created, to A single thread executes a single group backup plan; creates a virtual disk snapshot of the virtual machine to store the virtual disk snapshot; and writes the virtual disk snapshot creation process when the backup plan is completed In the execution process, it records the creation time or snapshot deletion time of the virtual disk snapshot.

In the foregoing method, setting the backup plan further includes setting an execution date to take a disk group snapshot, and wherein the execution date is executed regularly or at a specified time.

In the foregoing method, when it is determined that the virtual system disk of the virtual machine to which the virtual data disk in the backup plan belongs does not appear in any group, the virtual data disk is allowed to create a new group in units of virtual machines group.

In the foregoing method, after the virtual disk snapshot is created, when the number of disk snapshots exceeds the maximum backup quantity set by the backup plan, the oldest disk snapshot that is not set to be retained is deleted.

In the aforementioned method, when the single thread executes the backup plan of the single group, the operation of the virtual machine is stopped. In addition, the virtual disk snapshot only records data changes.

In the aforementioned method, before executing the backup plan, the method includes: selecting the image file to be used by the virtual machine, the corresponding virtual template and related network configuration; and copying the image file to the virtual disk to generate The virtual machine.

In the aforementioned method, the recovery includes restoring virtual machines in different states based on a snapshot of the virtual system disk at different points in time, and wherein the restored virtual machines in different states take at least one point in time virtual data disk according to user needs Mount to a virtual machine for data storage.

As can be seen from the above, the present invention discloses a backup method supporting a multi-disk architecture of a virtual machine, which is used to solve the problem that the traditional image-based virtual machine performs a long backup time, which leads to the continuous busy storage device. Specifically, the present invention Group backup mechanism and scheduling settings including multi-disk architecture of virtual machines It can automatically perform backup at a specified time, and the backup method based on the multi-disk architecture virtual machine can effectively reduce the backup time of the virtual machine. The backup plan proposed by the present invention divides the disk structure of the virtual machine into a virtual system disk and a virtual data disk, and provides a scheduling function to perform backup at a specified time. When performing backup, the group mechanism will be the same The virtual system disk and virtual data disk of a virtual machine are allocated to the same group, and a corresponding number of threads are generated according to the number of groups. Each thread is responsible for the backup of a single group, so that the virtual machine only needs to be suspended once You can just operate it. In addition, the virtual disk snapshot is only the recorded data transaction. The remaining unchanged blocks are referred to the original read-only disk, so the space is very small, and the backup time can be reduced from the original few minutes. It can be completed in a few seconds, so that the system resources can be used more efficiently.

501‧‧‧ group

502‧‧‧Virtual system disk

503‧‧‧Virtual Data Disk

504‧‧‧Virtual machine

505‧‧‧Virtual system disk snapshot

506‧‧‧Virtual data disk snapshot

507‧‧‧Virtual Data Disk

508‧‧‧Virtual machine from snapshot

S11~S15‧‧‧Step

201~205‧‧‧Step

301~304‧‧‧Step

401~411‧‧‧Flow

Please refer to the detailed description and the accompanying drawings of the present invention for further understanding of the technical content and purpose of the present invention. The related drawings include: FIG. 1 is a backup method of the present invention supporting a virtual machine multi-disk architecture Step diagram; Figure 2 is a step diagram of creating a multi-disk virtual machine in the backup method of the present invention supporting the multi-disk architecture of the virtual machine; Figure 3 is a creation method in the backup method of the present invention supporting the multi-disk architecture of the virtual machine Step diagram of the backup plan; Figure 4 is the execution flow chart of the backup plan in the backup method supporting the virtual machine multi-disk architecture of the present invention; and Figure 5 is the backup of the present invention supporting the virtual machine multi-disk architecture method Schematic diagram of virtual disk snapshot and restore in

The technical content of the present invention will be described below with specific specific implementation forms, and those skilled in the art can easily understand the advantages and effects of the present invention from the content disclosed in this specification. However, the present invention can also be implemented or applied by other specific embodiments.

FIG. 1 illustrates the steps of the backup method of the present invention supporting a multi-disk architecture of a virtual machine. The present invention proposes an improved virtual machine backup plan, which is used to solve the problem that the backup time of the virtual machine based on the image file is too long, so that the storage device continues to be busy. As shown in the figure, in step S11, the plan name, maximum backup quantity and selected virtual disk at the time of backup are set to create a backup plan. In this step, the user is provided with information on the maximum backup quantity and plan name of the backup plan. The maximum backup quantity is mainly used for the storage of virtual disk snapshots, and the plan name is used to search the group after grouping. In addition, users can freely select one or more virtual disks, that is, virtual disks to participate in this backup plan, and finally create a backup plan.

The aforementioned maximum number of backups is to avoid excessive virtual disk snapshots caused by virtual disks, which may overload the back-end storage device. Therefore, when creating a backup plan, you can set the maximum number of backups. When the generated virtual disk snapshots are larger than the backup plan When the maximum number of backups is set in the painting, the oldest virtual disk snapshot that is not set to be retained (reserved refers to those who are not automatically deleted) will be automatically deleted to maintain the space of the back-end storage device. The virtual disk snapshot section will be described in detail later.

In one embodiment, setting the backup plan further includes setting an execution date to take a disk group snapshot, wherein the execution date is executed regularly or at a specified time. The above provides the user with free choice of the execution date to take a virtual disk snapshot of the disk group. This execution date can be set to be executed periodically, such as minutes, hours, days, months, or specified time.

In step S12, the virtual system disks of each virtual machine are individually grouped and it is determined whether the virtual data disks in the backup plan and the virtual system disks of the virtual machines to which they belong are in a group. When the virtual system disk to which the virtual data disk belongs is already in the group, the virtual data disk is placed in the same group as the virtual system disk. The virtual disk includes a virtual system disk and a virtual data disk. During backup, if the virtual disk of the same virtual machine can perform backup at the same time, the possibility of the virtual machine being suspended multiple times can be reduced, so the present invention proposes to use the same The virtual disks of the virtual machine are backed up together. In this step, the virtual system disk and the virtual data disk of the same virtual machine are placed in the same group.

In addition, if it is determined that the virtual system disk of the virtual machine to which the virtual data disk in the backup plan belongs does not appear in any group, the virtual data disk is made to create a new group in units of virtual machines. The above is that when the virtual data disk cannot find the same virtual system disk as its own group in the backup plan, it creates another group by itself.

As can be seen from the above, first create a group for the virtual system disk of each virtual machine, and then determine whether the virtual system disk of the virtual machine to which the virtual data disk in this backup plan belongs is in the group. When the virtual data disk does not find the group of the virtual machine to which it belongs, it means the virtual system of the virtual machine If the disk is not in the backup target of this backup plan, the virtual data disk will be grouped in units of virtual machines. On the contrary, if the virtual system disk of the virtual machine to which the virtual system data disk belongs is already in the group, Then put this virtual data disk in the same group as the virtual system disk.

In step S13, a corresponding number of threads are created according to the number of groups created, so that a single thread can execute a backup plan for a single group. When performing a backup, when a single thread executes a backup plan for a single group, the operation of the virtual machine will be stopped.

The foregoing steps create a corresponding number of threads according to the number of groups, and a single thread executes the backup task of a single group, which can achieve the effect that only one operation of the virtual machine needs to be suspended when the snapshot is executed.

In step S14, a virtual disk snapshot of the virtual machine is created and stored. This step is mainly to create a virtual disk snapshot of the virtual machine to store the virtual disk snapshot, and the virtual disk snapshot only records the data transaction part, which will reduce the amount of backup data and avoid excessive backup time or too much space.

In one embodiment, after the virtual disk snapshot is created, when the number of disk snapshots exceeds the maximum backup quantity set by the backup plan, the oldest disk snapshot that is not set to be retained is deleted. In short, when the virtual disk snapshot is created, it will check whether the virtual disk snapshot exceeds the maximum number of backups set in the backup plan. If it exceeds, the oldest disk snapshot that has not been set as reserved will be deleted, and vice versa If not exceeded, the above deletion process is skipped.

The setting of the aforementioned maximum number of backups is to maintain the space utilization of the back-end storage device and avoid excessive virtual disk snapshots, which may cause Prepare for the burden. In one embodiment, after the maximum number of backups is set, the default value cannot be less than the minimum backup requirement. In addition, if the number of disk snapshots exceeds the maximum number of backups and it is not possible to find those who can be deleted (for example, all are reserved), then The backup process will warn you that the number of existing disk snapshots should be lower than the maximum number of backups before you can back up.

In step S15, when the execution of the backup plan is completed, the creation process of the virtual disk snapshot is written into the execution history, so as to record the creation time or snapshot deletion time of the virtual disk snapshot. In this step, the creation history of the virtual disk snapshot after the backup plan is written into the execution history, so as to provide detailed virtual disk snapshot creation time and snapshot deletion time.

In one embodiment, before executing the backup plan, when a virtual machine with a multi-disk architecture needs to be generated, the process is to first select the image file, corresponding virtual template and related network configuration that the virtual machine wants to use, Next, copy the image file to the virtual disk to generate the virtual machine.

In another embodiment, after the backup plan is executed multiple times, the user can restore the desired virtual machine by selecting, that is, the virtual machine in different states can be restored according to the virtual system disk snapshots at different points in time , Which includes using a virtual system disk at a different point in time, that is, you can take a virtual system disk at a certain point to restore a virtual machine, and you can also mount at least one virtual data disk at a point in time according to user needs. A virtual machine, and virtual data disks at multiple points in time can be used to find differences between before and after system data.

It can be seen from the above that after the virtual machine of the present invention adopts a multi-disk architecture, data is no longer recorded in the image file used for booting, but is stored in the disk used during booting, and because the disk is different Image file Attached to the virtual machine service, so the disk data can be kept independently after the virtual machine is deleted. In addition, the multi-disk architecture virtual machine can create data disks with different mounting numbers according to user needs. The data disks can be transferred to other virtual machines for access through mounting and unloading operations to achieve data portability .

In addition, when performing a backup, the original disk data will be changed to Read Only mode. Each time a virtual disk snapshot is created, only the data transaction part will be recorded, and the remaining unchanged blocks will be referred to The original disk that is read therefore takes up very little space, which in turn shortens the backup time of the virtual machine. In addition, when recovering through a virtual disk snapshot, the blocks stored in the disk snapshot and the data blocks that have not changed from the original disk through the reference are synthesized to create a disk at the time of the snapshot. However, since the original virtual disk snapshot still needs to be kept available after the recovery, the original disk will not be cleaned until the virtual disk snapshot is deleted, so it will be cleared together, so in order to avoid multiple virtual After the disk snapshot recovery operation, there are a large number of read-only original disks that cause a burden on the back-end storage device. The present invention proposes the concept of maximum number of backups (Quota) to avoid occupying too much space resources of the storage device.

FIG. 2 illustrates a step diagram of creating a multi-disk virtual machine in the backup method supporting the multi-disk architecture of the virtual machine of the present invention. In step 201, select the image file to be used by the virtual machine. This step explains that when creating a virtual machine, you must first select the image file that the virtual machine wants to use.

In step 202, the virtual template corresponding to the virtual machine is selected. This step is to select the virtual hardware template corresponding to the virtual machine for the virtual machine to be created.

In step 203, a virtual machine network related configuration is selected. This step explains that in addition to the aforementioned template, for network connection information, the network-related configuration is selected through this step.

In step 204, the image file is copied to the virtual disk. This step explains that the system will copy the image file selected by the user to the virtual disk, and then generate the required virtual machine in the subsequent steps.

In step 205, a virtual machine is generated using the virtual disk. This step is to boot through this virtual disk. Through the foregoing steps and processes, the creation of a disk virtual machine can be completed. In addition, users can also evaluate the storage requirements of the virtual machine and freely mount or unmount different data disks.

FIG. 3 illustrates the steps of creating a backup plan in the backup method of the present invention supporting the multi-disk architecture of a virtual machine. In step 301, basic information such as the name of the backup plan and the maximum number of backups are set. This step explains that you need to set the basic information such as the name, maximum backup quantity, and description of the backup plan before executing the backup plan.

In step 302, the virtual disk to be backed up is added to the backup plan. This step explains the concept that each virtual disk will be in the same group, so the virtual disk is added to the backup plan, and then the virtual system disk or virtual data disk in this virtual disk will be mapped to this backup plan painting.

In step 303, the execution method and execution time of the backup plan are set. This step sets the execution method of the aforementioned backup plan. For example, the execution method of the backup plan can be divided into single execution and periodic execution, and the periodic execution can be monthly, daily, weekly, hourly, or even minute. carried out.

In step 304, the backup plan is successfully established. This step explains Successfully established a backup plan.

The periodic backup function of the present invention can support users to arbitrarily select one or more virtual system disks and virtual data disks as targets of a backup plan. In addition, each time the system schedules a backup, it will automatically determine the contents of the backup target, and allocate the virtual system disk and virtual data disk of the same virtual machine to the same group through the group mechanism, and according to the number of groups Assign threads so that each thread is responsible for a single group backup. This method only needs to suspend the operation of the virtual machine once. After the backup is completed, the virtual machine can be resumed and the user can continue to operate the virtual machine. According to this, the problem that the same virtual machine is repeatedly suspended for multiple times due to selecting multiple virtual disks for snapshot can be avoided.

In addition, the backup plan can meet the needs of different situations, and provides the option of periodic backup and single backup, and the execution time can also be specified by the user. In terms of periodic backup, users can set the frequency of periodic backups, such as monthly, daily, weekly, hourly, minute, etc., while in single backup, the user first selects the backup target and specifies when to perform the backup. The procedure of manual backup by the user is omitted, and the backup mechanism is more flexible.

FIG. 4 illustrates the execution flow chart of the backup plan in the backup method supporting the multi-disk architecture of the virtual machine of the present invention. In the present invention, when executing the backup plan, the virtual disk of the virtual machine is grouped according to the virtual machine in a grouping manner.

In the process 401, the virtual system disks are grouped in units of virtual machines. This process is to create a group for each virtual machine's virtual system disk.

In process 402, whether the virtual machine virtual system disk to which the virtual data disk in the backup plan belongs is in the group. This process is to determine whether the virtual system disk of the virtual machine to which the virtual data disk in this backup plan belongs appears in the group. If the system disk group of the same virtual machine already exists in the group, proceed to Process 404, otherwise, proceed to process 403.

In the process 404, the virtual data disk is placed in the same group as the virtual system disk. This process puts the virtual data disk of the virtual machine into the system disk group belonging to the same virtual machine.

In addition, if the virtual data disk does not find the virtual machine group to which it belongs, it means that the virtual system disk of the virtual machine is not in the backup target of this backup plan, then proceed to the process 403, the virtual data disk Grouping by virtual machine means creating a group by yourself. In this way, other virtual data disks of the same virtual machine will also be added to this group.

In the process 405, a corresponding number of threads are created according to the number of groups. This process illustrates that the subsequent group will be backed up by one thread, so a corresponding number of threads must be created according to the number of groups.

In the process 406, a single thread executes the backup of a single group, and the operation of the virtual machine is suspended. In addition, in the process 407, the virtual disks in the group are backed up. The above two processes illustrate that each thread is responsible for a single group of virtual disk backup tasks. When the thread starts a backup task for a single group, in order to avoid the inconsistent virtual disk snapshot data caused by the continuous write content of the virtual machine In some cases, the state of the virtual machine will be changed to suspended first. When the virtual machine is in the suspended state, the user cannot perform any operation on the virtual machine.

In process 408, whether the number of snapshots exceeds the maximum number of backups set in this backup plan. This process means that when a virtual disk snapshot of a virtual disk is created, the number of snapshots of the current disk is compared with the maximum number of backups in this backup plan. If it exceeds, the process 409 is executed, otherwise, the process directly enters the process 410.

In the process 409, the oldest and unreserved target disk snapshot is deleted. This process is based on the fact that when the number of snapshots exceeds the maximum number of backups, the oldest disk snapshot in the virtual disk that is not set to be reserved must be deleted. Retention is whether the user can set the disk snapshot that does not want to be automatically deleted to be retained. In addition, if there is no need to delete the existing disk snapshot, this process is skipped.

In the process 410, the backup result is written into the backup history. This process writes the results into the backup history, so that users can accurately grasp the backup status of the virtual disk, including the success or failure of the virtual disk snapshot, and also write them into the backup history.

In the process 411, the state of the virtual machine is restored to running. This process is to wait for the thread to complete all the backup tasks in the group, that is, to change the state of the virtual machine back to running, the user can operate the virtual machine again, so that the overall backup process can be completed in just a few seconds. For users, it can achieve the effect of uninterrupted service.

FIG. 5 illustrates a schematic diagram of virtual disk snapshot and restoration in a backup method supporting a multi-disk architecture of a virtual machine according to the present invention. This embodiment mainly describes the related process of restoring through the virtual disk snapshot. Please refer to Table 1 for the backup plan execution history record table.

As shown in the figure, assume that the user creates a backup plan. A group 501 in the backup plan includes two backup targets from the virtual machine 504. The two backup targets are the virtual system disk 502 and the virtual data disk. 503 (that is, one system disk and one data disk), and set the backup plan to be executed every 30 minutes, and the maximum backup number of the backup plan is 3, that is, when the first For four disk snapshots, the oldest disk snapshot that has not been reserved is executed.

This backup plan performs the first backup at 10:35, and both backup targets belong to the same virtual machine, so only one thread is generated to process the backup task. This backup task is also completed at 10:35. Backup task, and regularly execute the backup task every 30 minutes. After that, the backup task executed at 12:05, the disk snapshot generated is the fourth disk snapshot, which exceeds the maximum number of backups set in this backup plan. At this time, it will automatically find this backup destination. The oldest and unreserved disk snapshot is deleted. As shown in Table 1, the disk snapshot generated at 12:05 will replace the disk snapshot generated at 10:35. In addition, in this working example, assuming that the user wants to retain the disk snapshot generated at 13:05 (the sixth run), and set the disk snapshot generated at 13:05 to be retained, the system will start at 14 : 35 minutes (the ninth run) After the disk snapshot is generated, the disk snapshot generated at 13:35 will be deleted to retain the disk snapshot that the user has set as the retention.

As shown in FIG. 5 again, the mechanism described in the present invention not only allows the user to freely select the backup target, but when performing the snapshot restore, the user can also arbitrarily select different virtual machine disk snapshots to restore according to their own needs. Taking Table 1 as an example, if the user wants to use the virtual system disk snapshot 505 generated at 13:05 and the virtual data disk snapshot 506 generated at 15:05, the virtual system generated at 13:05 can be used Disk snapshot to restore the virtual machine, and borrow Create a virtual data disk 507 from the virtual data disk snapshot generated at 15:05, and then mount the virtual data disk 507 to the restored virtual machine, that is, the virtual machine 508 generated by the snapshot.

The present invention adopts a multi-disk virtual machine architecture, so it takes only a few seconds to take a virtual disk snapshot, and allows the user to freely choose the execution time of the backup plan and the target disk to be backed up, not mandatory Back up the system disk and the data disk together; in addition, the group mechanism is used to centralize the pause operation time of the virtual machine, so that the user only feels a pause operation time of about several seconds, and establishes a relative according to the number of groups The number of threads, and then parallel processing, makes the overall backup mechanism more flexible, flexible and efficient.

It can be seen from the above that the backup plan and group scheduling mechanism proposed by the present invention based on the multi-disk architecture virtual machine can be applied to the cloud management and control system. Through the change of the virtual machine architecture, in addition to the independent separation of the virtual machine and the disk service Provides more flexibility in use, and can also effectively shorten the time for a virtual machine to create a disk snapshot, and the group scheduling function allows users to freely choose the disk to perform the disk snapshot and the execution method and time. Group management when scheduling is triggered to avoid repeated suspension of user operations on virtual machines, thus enhancing the applicability and user experience of the backup function of the cloud management and control system.

In summary, the present invention discloses a backup method that supports a multi-disk architecture of a virtual machine, including dividing the disk structure of the virtual machine into a virtual system disk and a virtual data disk, and providing a scheduling function to specify a time sheet Backups are performed at regular or regular intervals, and the backup mechanism supports virtual disk snapshots of virtual system disks and virtual data disks. According to this, disk-based backup is used in the present invention The method is different from the traditional image-based backup, only the differential part of the disk needs to be differentially backed up, so the backup time is reduced from a few minutes to a few seconds. In the backup method of the present invention, the group mechanism can arrange the backup work of a single virtual machine into the same group, and create a corresponding number of threads according to the number of groups, and then a single thread is responsible for the backup of a single group Work, this design can focus on the time when the virtual machine is suspended, to avoid the situation that the same virtual machine is repeatedly suspended for multiple times due to the selection of multiple disks for snapshots. In addition, it also provides some users with virtual machines between disks. When data linkage is required, snapshots at the same point in time can be used for restoration without data inconsistency. In summary, the backup method of the present invention supporting a multi-disk architecture of a virtual machine will provide a good backup mechanism to make the virtual machine It is easier to use and more flexible to restore.

The above detailed description is a specific description of a feasible embodiment of the present invention, but this embodiment is not intended to limit the patent scope of the present invention, and any equivalent implementation or change without departing from the technical spirit of the present invention should be included in The patent scope of this case.

S11~S15‧‧‧Step

Claims (10)

A backup method supporting a multi-disk architecture of a virtual machine, which includes: setting a plan name during backup, the maximum number of backups, and the selected virtual disk to create a backup plan; making the virtual system disk of each virtual machine Create a group to determine whether the virtual system disk of the virtual machine to which the virtual data disk in the backup plan belongs is in a group, so that the virtual system disk to which the virtual data disk belongs is already in the group When in a group, place the virtual data disk in the same group as the virtual system disk; create a corresponding number of threads based on the number of groups created to allow a single thread to perform a single group backup Plan; create a virtual disk snapshot of the virtual machine to store the virtual disk snapshot; and when the backup plan is completed, write the virtual disk snapshot creation process into the execution history to record the virtual Disk snapshot creation time or snapshot deletion time. The backup method for supporting a multi-disk architecture of a virtual machine as described in item 1 of the patent application scope, wherein setting the backup plan further includes setting an execution date to take a disk group snapshot. The backup method supporting the multi-disk architecture of a virtual machine as described in item 2 of the patent application scope, wherein the execution date is executed regularly or at a specified time. The backup method that supports the multi-disk architecture of a virtual machine as described in item 1 of the scope of the patent application, in which the virtual data in the backup plan is determined When the virtual system disk of the virtual machine to which the disk belongs does not appear in any group, make the virtual data disk create a new group in units of virtual machines. The backup method for supporting a multi-disk architecture of a virtual machine as described in item 1 of the patent scope, wherein after the virtual disk snapshot is created, the number of disk snapshots exceeds the maximum backup set by the backup plan When the number is exceeded, delete the oldest disk snapshot that has not been set as reserved. The backup method for supporting a multi-disk architecture of a virtual machine as described in item 1 of the patent scope, wherein the operation of the virtual machine is stopped when the single thread executes the backup plan of the single group. The backup method for supporting a multi-disk architecture of a virtual machine as described in item 1 of the scope of the patent application, wherein the virtual disk snapshot only records data changes. The backup method for supporting a multi-disk architecture of a virtual machine as described in item 1 of the scope of the patent application, wherein before executing the backup plan, the method includes: selecting the image file to be used by the virtual machine, the corresponding virtual template, and Related network configuration; and copy the image file to the virtual disk to generate the virtual machine. As described in item 1 of the scope of the patent application, a backup method that supports a multi-disk architecture of a virtual machine includes a virtual system disk snapshot at a different point in time to restore virtual machines in different states. As described in item 9 of the scope of the patent application, a backup method supporting a multi-disk architecture of a virtual machine, wherein restoring a virtual machine in different states includes mounting a virtual data disk with at least one time point to the virtual machine according to requirements Plan the machine for data storage.

Images