WO2017132872A1

WO2017132872A1 - Virtual machine backup method, backup device and host

Info

Publication number: WO2017132872A1
Application number: PCT/CN2016/073318
Authority: WO
Inventors: 母子安
Original assignee: 华为技术有限公司
Priority date: 2016-02-03
Filing date: 2016-02-03
Publication date: 2017-08-10
Also published as: CN107003893A; CN107003893B

Abstract

A virtual machine backup method and device. The method comprises: receiving a backup command, interrupting a service of a source virtual machine, and creating a clone virtual machine and a clone hard disk for the clone virtual machine (301); establishing a mapping relationship between the clone hard disk and a storage space of a source hard disk of the source virtual machine (302); mirroring operating data of the source virtual machine to the clone virtual machine, so as to enable the clone virtual machine to start operating according to the operating data, configuring the clone virtual machine to receive an IO request, and resuming the interrupted service of the source virtual machine (303); recording, upon detecting that a data operation executed by the clone virtual machine is a modification of data in the source hard disk, the modified data to the clone hard disk, and marking a position of the recording (304); upon detecting that execution of a service of the source virtual machine has been completed, shutting down the source virtual machine and backing up the data in the source hard disk (305); upon detecting that data backup is finished, copying, according to the mapping relationship, data at the marked position in the clone hard disk to a position in the source hard disk mapped to the marked position in the clone hard disk, and configuring the source hard disk as a hard disk for the clone virtual machine (306). The method and device can ensure consistency of backed up data without interrupting the service of a virtual machine.

Description

Virtual machine backup method, backup device and host machine

Technical field

The present invention relates to the field of computers, and in particular, to a virtual machine backup method, a backup device, and a host machine.

Background technique

A virtual machine (VM) is a complete system that runs through a software and has a complete hardware system function and runs in a completely isolated environment. When the virtual machine is running, the host where the virtual machine is located, that is, the host virtual machine virtualizes various hardware units, such as virtual machine memory and virtual machine hard disk. The data in the virtual machine's memory is managed by the virtual machine itself, while the data on the virtual machine's hard disk is stored and managed by the host.

1 is a schematic structural diagram of a virtual machine in the prior art. As shown in FIG. 1, all resources used by the virtual machine 101, such as CPU, memory, network, and storage resources, are all derived from the host 102, and the host 102 will itself. Resources, such as virtual multiprocessors, virtual machine cluster file systems, virtual network cards and switches, storage protocol stacks, network protocol stacks, and device drivers are provided to the virtual machine 101 in a shared manner to enable the virtual machine 101 to provide external Business services. Generally, the external service provided by the virtual machine 101 is implemented by an application (aplication, abbreviated as APP) 103 inside the virtual machine.

To ensure that the running data is not lost, or to ensure that the system recovers as soon as possible, the virtual machine needs to be backed up frequently, and the backup data is consistent with the running data. The backup data of the virtual machine is consistent with the running data and can also be called the consistency of the virtual machine. In the prior art, there are two backup methods for a virtual machine. One is cold backup, the cold backup is to back up data when the virtual machine is shut down, and the other is hot backup. The hot backup is when the virtual machine keeps working. The data is backed up.

Cold backup can guarantee the consistency of the virtual machine. However, the cold backup needs to shut down the virtual machine. When the virtual machine is shut down, it cannot provide external services. Therefore, cold backup is less used. Hot backup does not need to shut down the virtual machine, but when the virtual machine runs, the data of the internal application of the virtual machine is first saved in the cache, and then written into the virtual machine hard disk, that is, the data in the virtual machine hard disk is read, written, and updated at any time. During the backup process, the data is also updated at any time. Therefore, the hot backup method in the prior art cannot back up the updated data in the virtual machine hard disk in real time, resulting in data loss during the virtual machine backup process, and the virtual machine backup cannot be guaranteed.

It can be seen that there is no backup method in the prior art, that is, a method for providing virtual machine functions externally during backup and ensuring consistency of virtual machine backup.

Summary of the invention

In view of this, the purpose of the embodiments of the present invention is to provide a virtual machine backup method and a backup device, which ensure consistency of backup data when backing up data in the case of providing virtual machine functions externally.

In order to solve the above technical problem, the embodiment of the present invention discloses the following technical solutions:

In a first aspect, a virtual machine backup method is provided, which is applied to a host machine, and a plurality of virtual machines are created on the host machine, and when the user needs to back up data in the virtual machine on the host machine, the method is released. A virtual machine backup command, when the host receives the backup command, interrupts the service of the source virtual machine to be backed up, and creates a cloned virtual machine and a cloned hard disk of the cloned virtual machine.

After the cloned hard disk is built, the mapping relationship between the cloned hard disk and the storage space of the source virtual disk of the source virtual machine is established, and then the running data of the source virtual machine is mirrored to the cloned virtual machine, so that the The clone virtual machine starts running according to the running data, and sets the clone virtual machine to receive an IO request, so that the cloned virtual machine can take over the service of the source virtual machine. At this time, the service of the source virtual machine that is interrupted can be resumed, and the IO request that has not been executed when the backup command of the user is received is executed to ensure data consistency.

During the running of the cloned virtual machine, if a data operation is performed to the cloned virtual machine to modify data in the source hard disk, the modified data is recorded into the cloned hard disk, and the recorded data is recorded. The location is marked. After detecting that the execution of the service of the source virtual machine is completed, the source virtual machine is shut down, and data in the source hard disk is started to be backed up.

After detecting the end of the data backup, copying the data at the marked location in the cloned hard disk to the location of the source hard disk mapped at the marked location of the cloned hard disk according to the mapping relationship, and setting the location The source hard disk is the hard disk of the cloned virtual machine.

In this way, when the data in the virtual machine is backed up, the service executed by the virtual machine during the backup process is transferred to the cloned virtual machine by creating a cloned virtual machine including the cloned hard disk, and the virtual machine is backed up during the backup process. The modification of the data stored in the source virtual machine is recorded in the cloned hard disk, and the modified data is marked, so that after the data backup of the virtual machine is completed, the source may be updated by the modified data recorded in the cloned hard disk. The data in the virtual machine, in this way, while ensuring the consistency of the backup data, it does not interrupt the business of the virtual machine, thereby improving the operating efficiency of the virtual machine.

Optionally, when establishing a mapping relationship between the cloned hard disk and the storage space of the source hard disk, the source hard disk and the clone hard disk may be separately divided into multiple storage blocks according to a preset size, and then the source storage block is established. The mapping between the address and the address of the cloned block.

Through such division, each storage block in the source hard disk can be corresponding to one clone storage block, so that the modified source storage block and the modified data recorded in the clone storage block can be accurately marked.

Optionally, the modified source storage block may be marked by setting an update identifier for an address mapping relationship of an address of each source storage block in the mapping table and a clone storage block, wherein the modified method is to record the modified The update flag corresponding to the mapping relationship of the address of the cloned storage block of the data is set to be updated.

The updated identifier can conveniently identify the modified data in the source hard disk.

Modifying the data in the source hard disk may be writing new data on the basis of the data stored in the source hard disk at the backup time point, or updating the data stored in the source hard disk at the backup time point. It is also possible to delete data on the basis of the data stored in the source hard disk at the backup time point.

Optionally, after recovering the service of the terminal of the source virtual machine, when it is detected that the data operation performed by the source virtual machine updates the data in the source hard disk, determining that the data update request is updated Whether the update identifier corresponding to the mapping relationship of the address of the source storage block is updated, and the data is determined when the update identifier corresponding to the mapping relationship of the address of the source storage block updated by the data update request is not updated. And updating the updated data in the updated source storage block to the clone storage block corresponding to the source storage block updated by the data update request, and writing the data update in the source storage block updated by the data update request The update data in the request; and the update identifier corresponding to the source storage block updated by the data update request is set to be updated.

Since the service executed in the source virtual machine is also mirrored to the cloned virtual machine, the same service as the source virtual machine is also performed in the cloned virtual machine, so when the data in the source virtual machine is updated The updated data in the source virtual machine also needs to be transferred to the cloned storage block, so that the cloned virtual machine subsequently uses the data.

Optionally, during the running of the cloned virtual machine, the method for processing the read IO request is as follows:

Receiving a read IO request;

Determining whether the update identifier corresponding to the clone storage block read by the read IO request is updated;

Determining, when the update identifier corresponding to the clone storage block read by the read IO request is not updated, reading data from the source storage block corresponding to the read clone storage block;

When it is determined that the update identifier corresponding to the clone storage block read by the read IO request is updated, data is read from the read clone storage block.

Optionally, when copying the data at the marked location in the cloned hard disk to the location of the source hard disk mapped at the marked location of the cloned hard disk according to the mapping relationship, first determining the updated update identifier pair that is set to be updated. a clone storage block in the mapping relationship; copying data stored in the clone storage block in the mapping relationship corresponding to the updated update identifier to the mapping relationship corresponding to the updated update identifier set to be updated The source storage block in ; then set the update identifier set to updated to not updated.

In this way, after the data copy is completed, the modification flag is also set, so that when the read IO request is received during this period, the position of the read data can be determined according to the modification flag.

Optionally, after the data backup request is detected, when the data write request is received, the data to be written in the data write request is written into the source hard disk, and the write to be written is written. The update identifier corresponding to the source storage block of the incoming data is set to not updated.

Optionally, after the data in the cloned hard disk is all copied to the source hard disk, the cloned hard disk and the mapping table may be deleted.

In a second aspect, a virtual machine backup device is provided, where the backup device includes a creation unit, a marking unit, a backup unit, a reading unit, and an update unit, the creation unit, the marking unit, the backup unit, the reading unit, And the functions performed by the update unit correspond to the methods provided in the first aspect, and are not described herein again.

In a third aspect, a host is provided. The host includes: a processor, a memory, a hard disk, a communication interface, and a system bus. The processor, the memory, the hard disk, and the communication interface are connected through a system bus and complete each other. Communication in the memory for storing computer execution instructions, when the host is running, the processor executes the computer execution instructions in the memory to establish a virtual machine with hardware resources in the host machine, and executes The method provided by the first aspect. With the embodiment of the present invention, when the data in the virtual machine is backed up, the service executed by the virtual machine during the backup process is transferred to the cloned virtual machine by creating a cloned virtual machine including the cloned hard disk, and During the virtual machine backup process, the modification of the data stored in the source virtual machine is recorded in the cloned hard disk, and the modified data is marked, so that after the virtual machine data backup is completed, the cloned hard disk can be recorded. The modified data updates the data in the source virtual machine, so that the consistency of the backup data is ensured, and the service of the virtual machine is not interrupted, thereby improving the operating efficiency of the virtual machine.

DRAWINGS

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below. Obviously, the drawings in the following description are only It is a certain embodiment of the present invention, and other drawings can be obtained from those skilled in the art without any creative work.

1 is a schematic structural diagram of a virtual machine in the prior art;

2 is a structural diagram of a host machine according to an embodiment of the present invention;

FIG. 3 is a flowchart of a virtual machine backup method according to an embodiment of the present invention;

FIG. 4 is a schematic diagram showing a mapping relationship between a source hard disk and a storage space of a cloned hard disk according to an embodiment of the present invention;

FIG. 5 is a schematic diagram of updating data in a source hard disk during a running of a cloned virtual machine according to an embodiment of the present invention; FIG.

6 is a schematic diagram of data read by a cloned virtual machine according to an embodiment of the present invention;

FIG. 7 is a schematic diagram of updating data in a source hard disk during operation of a source virtual machine according to an embodiment of the present invention; FIG.

FIG. 8 is a schematic diagram of copying data in the cloned hard disk to the source hard disk according to an embodiment of the present invention; and

FIG. 9 is a block diagram of a data backup device according to an embodiment of the present invention.

The embodiments of the present invention have been shown in the foregoing drawings and are described in detail herein. The drawings and the written description are not intended to limit the scope of the present invention in any way,

detailed description

The technical solutions in the embodiments of the present invention are clearly and completely described in the following with reference to the accompanying drawings in the embodiments of the present invention. It is obvious that the described embodiments are only a part of the embodiments of the present invention, but not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts are within the scope of the present invention.

Numerous specific details are set forth in the Detailed Description of the invention in the following detailed description. In other instances, well-known methods, procedures, components, and circuits are not described in detail to avoid unnecessarily obscuring the embodiments.

When the data in the source hard disk of the source virtual machine is backed up, a clone virtual machine and a clone hard disk may be created, and the running data in the source virtual machine is transferred to the clone virtual machine and the subsequent access source virtual machine is accessed. The IO request is redirected, and the cloned virtual machine takes over the service of the source virtual machine, and after the data backup is completed, overwrites data in the source hard disk with data inconsistent with the source hard disk in the cloned hard disk, and the data is Source hard disk as the clone virtual machine The hard disk is such that when the data in the source virtual machine is backed up, the consistency of the backup data is ensured without interrupting the service performed by the virtual machine.

The technical solutions provided by the embodiments of the present invention will be described in detail below with reference to the embodiments.

As shown in FIG. 2, it is a structural diagram of a host machine 20 according to an embodiment of the present invention. The host machine 20 can include a processor 201, a memory 202, a hard disk 204, a communication interface 207, and a system bus 200. The processor 201, the memory 202, the hard disk 204, and the communication interface 207 are connected by the system bus 200 and complete communication with each other.

Processor 201 may be a single core or multi-core central processing unit, or a particular integrated circuit, or one or more integrated circuits configured to implement embodiments of the present invention.

The memory 202 can be a high speed RAM memory or a non-volatile memory, such as at least one disk memory, and the memory 202 is used to store computer execution instructions. Specifically, the program code may be included in the computer execution instruction.

When the host machine 20 is running, the processor 201 executes the in-memory computer execution instructions 203 to establish a plurality of virtual machines, such as the source virtual machine 208 and the clone virtual machine 209, using hardware resources in the host machine. A source hard disk 205 and a clone hard disk 206 are respectively created in the hard disk 204 for the source virtual machine 208 and the clone virtual machine 209. The clone virtual machine 209 and the clone hard disk 206 are created when the data in the source hard disk 205 of the source virtual machine 208 is backed up to the backup memory 30. The following will be combined with the virtual machine backup method shown in FIG. The flowchart illustrates the backup process in the source virtual machine.

As shown in FIG. 3, the virtual machine backup method of the embodiment of the present invention includes:

Step 301: Receive a backup command, interrupt the service of the source virtual machine, and create a cloned virtual machine and a cloned hard disk of the cloned virtual machine.

In the embodiment of the present invention, the service of the source virtual machine 208 is an I/O request being executed by the virtual machine, and the service of the interrupt source virtual machine 208 suspends the IO request being executed by the source virtual machine 208.

In actual application, after receiving the backup command, you can interrupt the service of the source virtual machine, create a cloned virtual machine and clone the virtual machine's cloned hard disk, or create a cloned virtual machine and clone the virtual machine's cloned hard disk before interrupting. The service of the source VM can be set according to the actual situation.

Step 302: Establish a mapping relationship between the cloned hard disk and a storage space of the source hard disk of the source virtual machine.

When the mapping relationship between the cloned hard disk and the storage space of the source hard disk is established, the source hard disk and the clone hard disk are respectively divided into multiple storage blocks according to a preset size, and then the address and clone of the source storage block are established. The mapping relationship of the addresses of the storage blocks.

When the clone virtual machine 209 and the clone hard disk 206 are created, the parameters of the source virtual machine 208 and the source hard disk 205 may be first obtained, and the clone virtual machine 209 is created according to the parameters of the source virtual machine 208 and the source hard disk 205. And cloning the hard disk 206, the capacity of the clone hard disk 206 is equal to the capacity of the source hard disk 205.

When the mapping relationship between the cloned hard disk 206 and the storage space of the source hard disk 205 is established, the source hard disk 205 and the clone hard disk 206 are first divided into a plurality of storage blocks according to a preset size, and then the source storage block is established. The mapping relationship between the address and the address of the cloned storage block.

4 is a schematic diagram showing a mapping relationship between the cloned hard disk and the storage space of the source hard disk. The source hard disk 205 and the clone hard disk 206 are both divided into a plurality of storage blocks, for example, source storage blocks 2051 and 2052 of the source hard disk 205. And, clone the memory blocks 2061 and 2062 of the hard disk 206.

The mapping relationship between the cloned hard disk 206 and the storage space of the source hard disk 205 includes an address of the source storage block and an address of the clone storage block corresponding to the source storage block, and each of the correspondences may be set with an update identifier. As shown in FIG. 4, the update identifier corresponding to each corresponding relationship may be indicated by a separate list 401, where the list 401 includes an update identifier corresponding to the storage block index number and the index number. The storage block index numbers Pa to Pe point to respective corresponding correspondences. In other embodiments, the update identifier may also be recorded with the correspondence between the source storage block and the clone storage block without being recorded by the independent list 401. The update identifier includes updated or not updated. In the embodiment shown in FIG. 4, the data is not updated with 0, and the data is updated with 1. In other embodiments of the present invention, it may also be indicated by 1 that the data is not updated, and that the data is updated with 0. The role of the update identifier will be described in detail below.

The mapping relationship between the cloned hard disk and the storage space of the source hard disk, and the corresponding update identifier may be in the form of a mapping table, and may also be in other forms, and details are not described herein again.

Step 303: Mirror the running data of the source virtual machine to the cloned virtual machine, so that the cloned virtual machine starts running according to the running data, and the cloned virtual machine is set to receive an IO request, and the execution is interrupted. The service of the source virtual machine.

The operational data includes data in memory in the source virtual machine 208, data in CPU registers, IO requests in the IO request stack, and data in the network stack. The data in the memory in the source virtual machine 208 is data required for the source virtual machine to run, the data in the CPU register is status bit information of the currently processed process or thread, and the IO request in the IO request stack Including the IO request being processed by the source virtual machine 208 and the IO request that has been received but not yet processed, the data stored in the network stack is the data to be sent to the external source 208 by the source virtual machine 208 and received from the outside. The data. Mirroring the operational data to the clone virtual machine 209 After that, the cloned virtual machine needs to be set to receive the IO request, so that the subsequently received IO request is sent to the cloned virtual machine for processing. When the cloned virtual machine starts to perform the service start, the service of the source virtual machine that is interrupted can be resumed. Since the backup time point that the user instructs to perform the backup starts from the time when the user issues the backup command (that is, the time when the backup command is received), the data required by the user is also all the data received before the backup time point, but due to the processor. The delay of processing the IO request, the IO request before the backup time point has not been processed in time, for example, is being executed, or has not been processed in time in the IO request stack, so the start of the business start of the cloned virtual machine is also The service of the source virtual machine that is interrupted needs to be resumed, so that data that has not been written to the source hard disk is written to the source hard disk after the backup time point, thereby ensuring data consistency.

Step 304: When it is detected that the clone virtual machine performs a data operation to modify data in the source hard disk, the modified data is recorded into the clone hard disk, and the recorded position is marked. Here you can use the update flag above to mark the location of the source hard drive or clone hard drive data update.

If the data operation is to write new data, store the new data to be written into the free clone storage block of the clone hard disk, and set the update flag corresponding to the clone storage block in which the data is written to Updated; if the data operation is to update data in the source hard disk, the update data is written into the clone hard disk and the source to be updated according to a mapping relationship between the source storage block and the clone storage block The cloned storage block corresponding to the storage block is set and the update identifier of the source storage block to be updated (or the cloned storage block to which the update data is written) is set to be updated. If the data operation is to delete the data in the source hard disk, set the update identifier corresponding to the source storage block where the data to be deleted is located (or the clone storage block corresponding to the source storage block where the data to be deleted is located) It has been updated.

The process of marking when the data stored in the backup time point in the source hard disk is modified is described below through a specific application scenario.

The application scenario shown in FIG. 5 is to perform a data operation on the cloned virtual machine to update data in the source hard disk. As shown in FIG. 5, the source virtual machine includes a source hard disk 205, and the clone virtual machine includes a clone hard disk 206.

The data operation is that the data E in the source storage block 2053 in the source hard disk 205 needs to be updated to F, and the first The update identifier corresponding to the source storage block 2053, that is, the update identifier corresponding to the index number Pe, is first found in the markup list 501, and it is determined whether the update identifier is updated (update identifier is 1), which can be seen from the figure. If the update identifier is not updated, that is, not updated (update identifier is 0), the data update data F is written into the clone storage block 2063 of the clone hard disk 206 corresponding to the source storage block 2053, and the mark list is The update identifier corresponding to the index number Pe recorded in 501 is modified to be updated (modified from 0 to 1). In another case, when the update identifier is updated, only the update data F needs to be written to the clone storage block 2063 without changing the update identifier.

The scenario in which the new data is deleted and the data is deleted on the basis of the data stored in the source hard disk at the backup time point is similar to the scenario of the updated data, and is not illustrated here.

Step 305: After detecting that the service execution of the source virtual machine is completed, shutting down the source virtual machine and backing up data in the source hard disk.

In the process of data backup, if the cloned virtual machine receives a data operation request, the modified data is also marked by the above method.

If a read IO request is received while the clone virtual machine is running, it is executed by the following method:

First, the cloned virtual machine receives a read IO request;

Then, determining whether the update identifier corresponding to the clone storage block read by the read IO request is updated;

When it is determined that the update identifier corresponding to the clone storage block read by the read IO request is not updated, reading data from the source storage block corresponding to the read clone storage block;

The method of data reading during the operation of the cloned virtual machine will be described below with reference to FIG.

In the application scenario shown in FIG. 6, the source virtual machine includes a source hard disk 205, and the clone virtual machine includes a clone hard disk 206.

As shown in FIG. 6, when the cloned virtual machine receives the data read request, it first determines the update flag corresponding to the cloned storage block 2065 corresponding to the read data (that is, the update identifier corresponding to the source storage block 2056). When the update identifier corresponding to the clone storage block 2065 is 0, indicating that the data in the source storage block 2055 corresponding to the clone storage block 2065 has not been updated, the data data B is directly read from the source storage block 2054. . When the update identifier corresponding to the clone storage block 2064 corresponding to the data read by the other data read request is 1, it indicates that the data in the source storage block 2054 corresponding to the clone storage block 2064 has been modified, then Clone memory block The data F is read in 2064.

After the execution of the interrupted operation of the source virtual machine, the source virtual machine and the clone virtual machine simultaneously execute the same because the service executed by the source virtual machine is also mirrored to the cloned virtual machine. Business, but because the cloned virtual machine is not synchronized with the source virtual machine, the source virtual machine may update the data in the source hard disk earlier than the cloned virtual machine, so that the cloned virtual machine needs to read the source hard disk. In the case of data, since the data of the source hard disk has been updated, an error is caused, so in this case, the data in the source hard disk needs to be migrated to the cloned virtual machine first, and an update flag is set for marking. The specific method is:

When it is detected that the data operation performed by the source virtual machine is to update the data in the source hard disk, determining whether the update identifier corresponding to the mapping relationship of the address of the source storage block updated by the data update request is updated;

Determining, when the update identifier corresponding to the mapping relationship of the address of the source storage block updated by the data update request is not updated, copying the updated data in the source storage block updated by the data update request to the data update request a cloned storage block corresponding to the updated source storage block;

Writing update data in the data update request in a source storage block updated by the data update request;

The update identifier corresponding to the source storage block updated by the data update request is set to be updated.

As shown in FIG. 7, when the data B in the source storage block 2056 is to be updated to F, the modification flag corresponding to the source storage block 2056 is first determined, when the modification flag corresponding to the source storage block 2056 is 0. Then, the data B stored in the source storage block 2056 is migrated to the clone storage block 2066 corresponding to the source storage block 2056, and then the update data F is written into the source storage block 2056. Conversely, if the corresponding modification flag of the source storage block 2056 is 1, the data B in the source storage block 2056 is directly updated to F, and there is no need to migrate the data B into the clone storage block 2066, the source. The update flag corresponding to the storage block 2056 also does not have to be modified. Step 306: After detecting the end of the data backup, copy the data at the marked location in the cloned hard disk to the location of the source hard disk mapped at the marked location of the cloned hard disk according to the mapping relationship. And setting the source hard disk to be a hard disk of the cloned virtual machine.

When copying the data at the marked location in the cloned hard disk to the location of the source hard disk mapped at the location where the cloned hard disk is marked according to the mapping relationship, it is preferred to determine the clone storage in the mapping relationship corresponding to the updated update identifier. And copying, by the block, the data stored in the clone storage block in the mapping relationship corresponding to the updated update identifier to the source storage block in the mapping relationship corresponding to the update identifier set to be updated; The update ID set to be updated is set to not updated.

The cloned hard disk and the mapping table can then be deleted.

After the data in the cloned hard disk is copied to the source hard disk, the clone virtual machine 209 will replace the source virtual machine 208 to provide external services. At this time, the clone virtual machine 209 uses the source hard disk 205.

FIG. 8 is a schematic diagram of copying data in the cloned hard disk to the source hard disk according to an embodiment of the present invention.

As shown in FIG. 8, the source virtual machine includes a source hard disk 205, and the clone virtual machine includes a clone hard disk 206. When the data is copied, the clone storage block with the update identifier 1 is obtained through the identifier list 901, such as the clone storage block 9022 and the clone storage block 9021. The data stored in the clone storage block 9022 is F, and the clone storage block 9021 does not. The data is stored, indicating that the data operation performed on the source memory block 9011 corresponding to the clone memory block 9021 is a delete operation, and thus, when copying, the data in the clone memory block with the update flag of 1 is overwritten with the clone storage. The data of the source storage block corresponding to the block, that is, the data in the source storage block 9012 is covered by the data F in the clone storage block 9022, and the data C in the source storage block 9011 is overwritten by the null data in the clone storage block 9021. The data C is deleted directly. After copying the data in the cloned memory block to the source memory block, the update flag can be set from 1 to 0.

After all the data in the cloned hard disk marked as the updated clone storage block is copied to the source hard disk, the cloned hard disk and the mapping table are deleted, and the cloned virtual machine provides the external service.

In the embodiment of the present invention, after the data backup request is detected, when the data write request is received, the data to be written in the data write request is written into the source hard disk, and the The update identifier corresponding to the source storage block to be written data is set to not updated.

According to the method of the embodiment of the present invention, when the data in the virtual machine is backed up, the cloned virtual machine including the cloned hard disk is created, and the service executed by the virtual machine in the backup process is transferred to the cloned virtual machine for execution, and During the virtual machine backup process, the data stored in the source virtual machine is recorded in the cloned hard disk, and the modified data is marked. Thus, after the virtual machine data backup is completed, the cloned hard disk may be recorded. The data is updated to update the data in the source virtual machine, so that the consistency of the backup data is ensured, and the service of the virtual machine is not interrupted, thereby improving the operating efficiency of the virtual machine.

As shown in FIG. 9, the embodiment of the present invention further provides a data backup device 10, which includes a creating unit 11, a marking unit 12, a backup unit 13, a reading unit 14, and an updating unit 15. The creating unit 11 is configured to receive a backup command, interrupt the service of the source virtual machine, create a cloned virtual machine, and a cloned hard disk of the cloned virtual machine, and establish a storage space of the cloned hard disk and the source hard disk of the source virtual machine. Mapping the relationship and mirroring the running data of the source virtual machine to the clone virtual machine, so that the cloned virtual machine is based on the running data The operation starts, and the cloned virtual machine is configured to receive an IO request, and resume the execution of the interrupted service of the source virtual machine.

The creating unit 11 is configured to divide the source hard disk and the clone hard disk into multiple storage blocks according to a preset size when establishing a mapping relationship between the cloned hard disk and the storage space of the source hard disk; The mapping relationship between the address of the block and the address of the cloned block.

The method for establishing a mapping relationship between the cloned hard disk and the storage space of the source hard disk of the source virtual machine and the method for mirroring the running data of the source virtual machine to the cloned virtual machine are described in detail in the method embodiment. , will not repeat them here.

The marking unit 12 is configured to: when detecting that the clone virtual machine performs a data operation to modify data in the source hard disk, record the modified data to the cloned hard disk, and record the The location is marked.

When the marking unit 12 marks the recorded position, the marking unit 12 is specifically configured to set the update flag corresponding to the mapping relationship of the address of the cloned storage block in which the modified data is recorded to be updated.

The marking unit 12 is further configured to: when detecting the data operation performed by the source virtual machine, update the data in the source hard disk, determine a mapping of an address of the source storage block updated by the data update request. Whether the update identifier corresponding to the relationship is updated; determining that the update identifier corresponding to the mapping relationship of the address of the source storage block updated by the data update request is not updated, and the data storage request is updated in the source storage block And the updated data is copied to the clone storage block corresponding to the source storage block updated by the data update request; the update data in the data update request is written in the source storage block updated by the data update request; The update identifier corresponding to the source storage block updated by the data update request is set to be updated.

The backup unit 13 is configured to shut down the source virtual machine and back up data in the source hard disk after detecting that the service execution of the source virtual machine is completed.

The method further includes a reading unit 14 configured to receive a read IO request received to read data in the cloned virtual machine, and determine whether an update identifier corresponding to the cloned storage block read by the read IO request is updated; Reading the data from the source storage block corresponding to the read clone storage block when the update identifier corresponding to the clone storage block read by the IO request is not updated; determining the read by the read IO request When the update identifier corresponding to the cloned storage block is updated, data is read from the read cloned storage block.

The updating unit 15 is configured to, after detecting the end of the data backup, copy the data at the marked location in the cloned hard disk to the source mapped at the marked location of the cloned hard disk according to the mapping relationship. Position the hard disk and set the source hard disk to be the hard disk of the cloned virtual machine.

The updating unit 15 is specifically configured to determine a clone storage block in a mapping relationship corresponding to the updated update identifier, and store the clone storage block in the mapping relationship corresponding to the updated update identifier. The data is copied to the source storage block in the mapping relationship corresponding to the updated update identifier set; the update identifier set to be updated is set to not updated.

After the data copying in the cloned hard disk is completed, the update unit deletes the clone hard disk and the mapping table.

For the system embodiment, since it basically corresponds to the method embodiment, it can be referred to the partial description of the method embodiment. The system embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, ie may be located A place, or it can be distributed to multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the embodiment. Those of ordinary skill in the art can understand and implement without any creative effort.

The invention may be described in the general context of computer-executable instructions executed by a computer, such as a program module. Generally, program modules include routines, programs, objects, components, data structures, and the like that perform particular tasks or implement particular abstract data types. The invention may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are connected through a communication network. In a distributed computing environment, program modules can be located in both local and remote computer storage media including storage devices.

One of ordinary skill in the art can understand that all or part of the steps in implementing the above method embodiments can be completed by a program to instruct related hardware, and the program can be stored in a computer readable storage medium, which is referred to herein. Storage media, such as: ROM, RAM, disk, CD, etc.

It should also be noted that, in this context, relational terms such as first and second, etc. are used merely to distinguish one entity or operation from another entity or operation, without necessarily requiring or implying such entities or operations. There is any such actual relationship or order between them. Moreover, the terms "including", "comprising" or any other variant thereof are intended to be A non-exclusive inclusion, such that a process, method, article, or device that includes a series of elements includes not only those elements but also other elements not specifically listed or included in the process, method, article, or device The inherent elements. An element that is defined by the phrase "comprising a ..." does not exclude the presence of additional equivalent elements in the process, method, item, or device that comprises the element.

The above is only the preferred embodiment of the present invention and is not intended to limit the scope of the present invention. The principles and embodiments of the present invention have been described in detail herein. The description of the above embodiments is only for the purpose of understanding the method of the present invention and the core idea thereof. The idea of the invention will vary in the specific embodiments and applications. In summary, the content of the specification should not be construed as limiting the invention. Any modifications, equivalents, improvements, etc. made within the spirit and scope of the invention are intended to be included within the scope of the invention.

Claims

A virtual machine backup method, the method comprising:

Receiving a backup command, interrupting the service of the source virtual machine, and creating a cloned virtual machine and a cloned hard disk of the cloned virtual machine;

Establishing a mapping relationship between the cloned hard disk and a storage space of the source hard disk of the source virtual machine;

Mirroring the running data of the source virtual machine to the cloned virtual machine, so that the cloned virtual machine starts running according to the running data, setting the cloned virtual machine to receive an IO request, and restoring execution of the interrupted source Virtual machine business;

After detecting that the data operation performed by the cloned virtual machine is to modify data in the source hard disk, the modified data is recorded into the cloned hard disk, and the recorded location is marked;

After detecting that the service execution of the source virtual machine is completed, shutting down the source virtual machine and backing up data in the source hard disk;

After detecting the end of the data backup, copying the data at the marked location in the cloned hard disk to the location of the source hard disk mapped at the marked location of the cloned hard disk according to the mapping relationship, and setting the location The source hard disk is the hard disk of the cloned virtual machine.
The method of claim 1, wherein the mapping relationship between the cloned hard disk and the storage space of the source hard disk is as follows:

Separating the source hard disk and the clone hard disk into a plurality of storage blocks according to a preset size;

Establish a mapping relationship between the address of the source storage block and the address of the cloned storage block.
The method according to claim 2, wherein the address mapping relationship between the address of each source storage block and the clone storage block in the mapping table further corresponds to an update identifier;

The marking of the recorded location specifically includes:

The update flag corresponding to the mapping relationship of the address of the clone storage block in which the modified data is recorded is set to be updated.
The method of claim 3, wherein the method further comprises:

When it is detected that the data operation performed by the source virtual machine is to update the data in the source hard disk, Whether the update identifier corresponding to the mapping relationship of the address of the source storage block updated by the data update request is updated;

Determining, when the update identifier corresponding to the mapping relationship of the address of the source storage block updated by the data update request is not updated, copying the updated data in the source storage block updated by the data update request to the data update request a cloned storage block corresponding to the updated source storage block;

Writing update data in the data update request in a source storage block updated by the data update request;

The update identifier corresponding to the source storage block updated by the data update request is set to be updated.
The method of claim 3, wherein the method further comprises:

Receiving a read IO request to read data in the cloned virtual machine;

Determining whether the update identifier corresponding to the clone storage block read by the read IO request is updated;

Determining, when the update identifier corresponding to the clone storage block read by the read IO request is not updated, reading data from the source storage block corresponding to the read clone storage block;

When it is determined that the update identifier corresponding to the clone storage block read by the read IO request is updated, data is read from the read clone storage block.
The method according to claim 3, wherein the copying the data at the marked location in the cloned hard disk to the location of the source hard disk mapped at the marked location of the cloned hard disk according to the mapping relationship comprises:

Determining a clone storage block in a mapping relationship corresponding to the updated update identifier;

Copying data stored in the clone storage block in the mapping relationship corresponding to the updated update identifier to the source storage block in the mapping relationship corresponding to the updated update identifier;

Set the update ID set to updated to not updated.
The method of claim 3, wherein the method further comprises:

After detecting the data backup end, when receiving the data write request, writing the data to be written in the data write request to the source hard disk, and writing the source of the data to be written The update ID corresponding to the storage block is set to not updated.
A virtual machine backup device, characterized in that the device comprises:

The unit is configured to receive a backup command, interrupt the service of the source virtual machine, create a cloned virtual machine, and clone the hard disk of the cloned virtual machine, and establish a mapping relationship between the cloned hard disk and the storage space of the source virtual disk of the source virtual machine. And mirroring the running data of the source virtual machine to the clone virtual machine, so that the cloned virtual machine starts running according to the running data, and the cloned virtual machine is configured to receive an IO request, and resume execution of the interrupted operation. Describe the business of the source virtual machine;

a marking unit, configured to: when detecting that the clone virtual machine performs a data operation to modify data in the source hard disk, record the modified data to the cloned hard disk, and perform the recorded location mark;

a backup unit, configured to: after detecting that the service execution of the source virtual machine is completed, shutting down the source virtual machine, and backing up data in the source hard disk;

And an updating unit, configured to: after detecting the end of the data backup, copy the data at the marked location in the cloned hard disk to the source hard disk mapped at the marked location of the cloned hard disk according to the mapping relationship; Position and set the source hard disk to be the hard disk of the cloned virtual machine.
The device according to claim 8, wherein the creating unit is specifically configured to establish a mapping relationship between the cloned hard disk and a storage space of the source hard disk.

Separating the source hard disk and the clone hard disk into a plurality of storage blocks according to a preset size;

Establish a mapping relationship between the address of the source storage block and the address of the cloned storage block.
The apparatus according to claim 9, wherein the address mapping relationship of each source storage block in the mapping table and the address mapping relationship of the clone storage block further corresponds to an update identifier;

When the marking unit marks the recorded position, it is specifically used,

The update flag corresponding to the mapping relationship of the address of the clone storage block in which the modified data is recorded is set to be updated.
The device according to claim 10, wherein said marking unit is further configured to:

When it is detected that the data operation performed by the source virtual machine updates the data in the source hard disk, it is determined whether the update identifier corresponding to the mapping relationship of the address of the source storage block updated by the data update request is Update

Determining an update identifier corresponding to a mapping relationship of an address of the source storage block updated by the data update request When not updated, the updated data in the source storage block updated by the data update request is copied to the clone storage block corresponding to the source storage block updated by the data update request;

Writing update data in the data update request in a source storage block updated by the data update request;

The update identifier corresponding to the source storage block updated by the data update request is set to be updated.
The device of claim 10, further comprising a reading unit for:

Receiving a read IO request receipt for reading data in the cloned virtual machine;

Determining whether the update identifier corresponding to the clone storage block read by the read IO request is updated;

Determining, when the update identifier corresponding to the clone storage block read by the read IO request is not updated, reading data from the source storage block corresponding to the read clone storage block;

When it is determined that the update identifier corresponding to the clone storage block read by the read IO request is updated, data is read from the read clone storage block.
The device according to claim 10, wherein the updating unit is specifically configured to:

Determining a clone storage block in a mapping relationship corresponding to the updated update identifier;

Copying data stored in the clone storage block in the mapping relationship corresponding to the updated update identifier to the source storage block in the mapping relationship corresponding to the updated update identifier;

Set the update ID set to updated to not updated.
The device according to claim 10, wherein the updating unit is further configured to: when the data write request is received after detecting the end of the data backup, write the data to the data to be written in the request Writing to the source hard disk, and setting an update identifier corresponding to the source storage block to which the data to be written is written is not updated.
A host machine, comprising: a processor, a memory, a hard disk, a communication interface, and a system bus, wherein the processor, the memory, the hard disk, and the communication interface are connected through a system bus and complete each other Communication, in the memory for storing computer execution instructions, when the host is running, the processor executes the computer execution instructions in the memory to establish a virtual machine with hardware resources in the host machine, and execute the right The method of any of 1-7 is claimed.