CN102289510B - Method and apparatus for writing data to a production volume - Google Patents

Abstract

The embodiment of the invention provides a method and device for writing data to a production volume. The method for writing the data to the production volume comprises the following steps of: receiving a data operation request carrying data to be written and the initial address and the length of the data to be written; when copy-on-write is not carried out in a target region of the data to be written in the production volume, determining data blocks in the need of performing the copy-on-write according to the initial address and the length of the data to be written and the preset storage granularity; performing the copy-on-write on the data blocks in the need of performing the copy-on-write, and generating mapping marks corresponding to the data blocks; and writing the data to be writtento the data blocks after performing the copy-on-write. According to the method and device disclosed by the invention, the sizes of practical data blocks on the production volume corresponding to the mapping marks can be dynamically adjusted; the system performance is increased; and resource pool resources is saved.

Description

Method and apparatus for writing data to a production volume

technical field

Embodiments of the present invention relate to the field of communication technologies, and in particular, to a method and device for writing data to a production volume.

Background technique

With the continuous development of information science and technology and the wide application of the Internet, the amount of data generated is growing at an incredible speed. The design of the storage system has gradually become a problem that needs to be considered in the design of the application system. Higher requirements are put forward in terms of availability, reliability, and scalability. At the same time, green storage is also vigorously advocated, so higher requirements are put forward for the reasonable allocation and utilization of storage system resources. On the one hand, the demand for data protection, data backup, and data parallel utilization (such as testing, report generation, and/or data analysis, etc.) is becoming more and more important; on the other hand, the online business model requires the system to run continuously, and downtime means This leads to a halt in business and loss of business opportunities. As a result, the data snapshot technology came into being.

A snapshot is a fully available copy of a data object, including the image of the data object at a certain point in time. Using the data snapshot technology, you can instantly obtain a consistent image (snapshot) of fully available production volume data in an instant, and occupy less storage space; after the snapshot is created, the access to the snapshot can be the same as the access to the production volume Simultaneously so as not to affect the business continuity of production volumes. Among them, the production volume, also called the source logical unit number (Logic Unit Number; hereinafter referred to as: LUN) or the source physical volume, is a volume that provides data storage and access space for basic business applications of the storage system, and is deployed in the storage array.

When creating a snapshot, a logical copy at the time of the snapshot needs to be created, that is, a mapping table is generated to indicate the storage location of the actual data corresponding to the snapshot. When the snapshot is just created, the pointers in the mapping table all point to the production volume, and each record in the mapping table corresponds to an actual data block on the production volume.

When modifying the production volume, it is necessary to first determine whether the data to be modified has been copied before writing (Copy-On-Write). Copy the data of the part to be modified to the resource pool, and modify the mapping table so that the pointer points to the corresponding position in the resource pool, and then modify the production volume; if the data of the part to be modified has been copied before writing, it can be directly modified The data for that portion of the production volume. If the production volume has multiple snapshots, before modifying the production volume, it is necessary to judge the mapping information of each snapshot as described above, and determine whether to perform copy-before-write according to the judgment result, and then the production volume can be modified. to modify. Wherein, the resource pool is used to store snapshot data different from the production volume, including the copy-before-write data generated when the production volume is written, and the resource pool and the production volume are deployed in the same storage array.

On the premise of ensuring business continuity, the data snapshot technology generates a data consistency image of the data object at a certain moment, which effectively guarantees the reliability and integrity of the data to a certain extent. However, since the size of the actual data block on the production volume corresponding to the mapping mark in the mapping table is fixed, there are problems of reducing system performance and wasting resource pool resources, and after a long time of reading and writing, the mapping mark in the mapping table is random Therefore, multiple redirections are required when reading and writing snapshots or production volumes, and the location corresponding to each redirection is random, and the read and write performance will be greatly reduced.

Contents of the invention

Embodiments of the present invention provide a method and device for writing data to a production volume, so as to dynamically adjust the size of an actual data block on the production volume corresponding to a mapping mark, improve system performance and save resource pool resources.

An embodiment of the present invention provides a method for writing data to a production volume, including:

receiving a data operation request, the data operation request carrying data to be written, and the start address and length of the data to be written;

When the target area of the data to be written in the production volume has not been copied before writing, determine the data block that needs to be copied before writing according to the start address, length and preset storage granularity of the data to be written ;

performing copy-before-write on the data block that needs to be copied before write, and generating a mapping mark corresponding to the data block;

Writing the data to be written into the data block after the copy-before-write is performed.

An embodiment of the present invention also provides a device for writing data to a production volume, including:

A receiving module, configured to receive a data operation request, the data operation request carrying data to be written, and the start address and length of the data to be written;

A determining module, configured to determine the need according to the start address, length, and preset storage granularity of the data to be written when the target area of the data to be written in the production volume has not been copied before writing. Data blocks for copy-before-write;

A generating module, configured to perform copy-before-write on the data blocks determined by the determination module that need to be copied before write, and generate a mapping mark corresponding to the data blocks;

The writing module is configured to write the data to be written received by the receiving module into the data block after the copy-before-write is performed.

Through the embodiment of the present invention, after receiving the data to be written, when the target area of the data to be written in the production volume has not been copied before writing, according to the start address, length and preset storage of the data to be written The granularity determines the data blocks that need to be copied before writing; performs copy before writing on the above data blocks that need to be copied before writing, and generates the mapping mark corresponding to the above data blocks; writes the data to be written after copying before writing data block. Therefore, it is possible to dynamically adjust the size of the actual data block on the production volume corresponding to the mapping mark, improve system performance and save resource pool resources.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description These are some embodiments of the present invention. Those skilled in the art can also obtain other drawings based on these drawings without creative work.

FIG. 1 is a flowchart of an embodiment of a method for writing data to a production volume in the present invention;

2 is a flow chart of another embodiment of the method for writing data to a production volume in the present invention;

Fig. 3 is a schematic diagram of an embodiment of the data writing process of the present invention;

4 is a schematic diagram of another embodiment of the data writing process of the present invention;

5 is a schematic diagram of another embodiment of the data writing process of the present invention;

FIG. 6 is a schematic diagram of an embodiment of merging continuous mapping marks in the present invention;

FIG. 7 is a schematic structural diagram of an embodiment of an apparatus for writing data to a production volume according to the present invention;

FIG. 8 is a schematic structural diagram of another embodiment of an apparatus for writing data to a production volume according to the present invention.

Detailed ways

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments It is a part of embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

Fig. 1 is a flowchart of an embodiment of a method for writing data to a production volume according to the present invention. As shown in Fig. 1, the method for writing data to a production volume may include:

Step 101, receiving a data operation request, the data operation request carrying data to be written, and the start address and length of the data to be written.

Step 102, when the target area of the data to be written in the production volume has not been copied before writing, it is determined according to the starting address of the data to be written, the length of the data to be written and the preset storage granularity that need to be performed. Copy-before-write data blocks.

Specifically, determining the data blocks that need to be copied before writing according to the start address, length, and preset storage granularity of the data to be written may be: the storage array where the production volume is located is based on the length of the data to be written and the preset The set storage granularity determines the size of the data block that needs to be copied before writing, and determines the data block that needs to be copied before writing according to the starting address of the data to be written and the size of the data block that needs to be copied before writing s position.

In this embodiment, the above-mentioned preset storage granularity may be: the size of the data block that needs to be copied before writing is the power of 2 times the minimum value of the preset storage granularity; the above-mentioned preset storage granularity The minimum value of is a power of 2 times, for example: it may be the minimum input and output (Input Output; hereinafter referred to as: IO) granularity of accessing the disk, that is, the size of a sector.

Step 103, perform copy-before-write on the data blocks that need to be copied before write, and generate mapping marks corresponding to the above data blocks.

Among them, the mapping mark is used to record the relevant information of the data block that is copied before writing. In order to further explain the mapping mark, for example, if you want to borrow books in the library, you can use the index number of the record to find the corresponding book. Here, the mapping mark can be understood as an index number, and the data block for copy-before-write can be understood as a book. In addition, the mapping mark can also be a pointer, etc., and the embodiment of the present invention does not limit the form of the mapping mark, as long as the mapping mark can record the relevant information of the data block to be copied before writing.

Step 104, write the data to be written into the data block after the copy-before-write is performed.

Further, after the above-mentioned data to be written is successfully written into the data block after the copy-before-write, the storage array where the production volume is located returns a write operation success response; or, after the above-mentioned data to be written is written and the copy-before-write After subsequent data blocks fail, a write operation failure response is returned.

In addition, in this embodiment, when there is at least one data block that is copied before overwriting in the data blocks covered by the target area of data to be written in the production volume, the storage array where the production volume is located performs pre-overwriting The number of data blocks to be copied, the size of the above-mentioned data blocks copied before overwriting, and the size of the above-mentioned target area, split the data to be written into a section of data to be written, or according to the above data copied before overwriting The number of blocks, the size of the above-mentioned data blocks copied before overwriting, the position of the above-mentioned data blocks copied before overwriting in the target area, and the size of the above-mentioned target area split the data to be written into at least two sections to be written into the data segment. Wherein, the size of the target area is the length of the data to be written.

For example, when there is only one copy-before-write data block in the data block covered by the target area of the data to be written in the production volume, and the size of the copy-before-write data block is the same as that of the target area When the sizes of are the same, it is only necessary to split the above-mentioned data to be written into a section of data to be written, which is equivalent to not splitting the above-mentioned data to be written.

When there is only one copy-before-write data block in the data block covered by the target area to be written in the production volume, and the copy-before-write data block is smaller than the size of the target area, if the If the data block that is copied before overwriting is located at the endpoint of the above target area, the data to be written can be split into at least two data segments to be written; if the data block that is copied before overwriting is located in the above target area In the middle part, the above data to be written needs to be split into at least three data segments to be written.

For example: Assume that the target area of the data to be written covers 3 data blocks, which are data block 1, data block 2 and data block 3 in sequence; if data block 1 or data block 3 is copied before overwriting, that is, before overwriting If the copied data block is located at the endpoint of the above target area, the data to be written can be split into two data segments to be written, and the target area of the split data segment to be written is data block 1 or data Block 3, the target area of another data segment to be written after splitting is two data blocks other than data block 1 or data block 3; if data block 2 is copied before overwriting, copy before overwriting If the data block is located in the middle of the above target area, the data to be written needs to be split into three data segments to be written. The target area of the split data segment to be written is data block 1. After splitting The target area of another data segment to be written is data block 2, and the target area of another data segment to be written after splitting is data block 3.

In this embodiment, the data block covered by the target area of the data to be written has only one data block that has been copied before writing as an example to illustrate the splitting method of the data to be written. When the data to be written When there are two or more data blocks that have been copied before writing in the data blocks covered by the target area, the splitting method of the data to be written is similar to the above, and will not be repeated here. Simply put, it is to use At least two data segments to be written after splitting form two groups. The target area of one group of data segments to be written only covers the data blocks copied before overwriting, and the target area of the other group of data segments to be written only covers Overwrite data blocks that have not been copied before write.

Then, the storage array where the production volume is located judges whether the data block covered by the target area to be written in each of the at least two data segments to be written has been copied before writing;

For the data segment to be written that has not been copied before writing in the data block covered by the target area to be written, the storage array determines the required A data block that needs to be copied before writing; perform a pre-write copy on the data block that needs to be copied before writing, and generate a mapping mark corresponding to the data block; finally, write the data in the data segment to be written above for writing The data block after the pre-copy.

For the data segment to be written in which the data block covered by the target area to be written is copied before overwriting, the data in the data segment to be written is written into the target area to be written in the data segment to be written. of data blocks.

Further, after the at least two data segments to be written are successfully written, the storage array returns a write operation success response; or, at least one of the at least two data segments to be written is written After the failure, the above storage array returns a write operation failure response.

In this embodiment, after step 104, the storage array where the production volume is located can also detect the data block corresponding to the mapping mark, and merge the mapping marks corresponding to the physically continuous data blocks into one mapping mark; thus, the memory resource consumption can be reduced. Consumption, improve system performance. For example, after the storage array where the production volume is located detects the data block corresponding to the mapping mark, it finds that mapping mark 1 corresponds to data block 1, and mapping mark 3 corresponds to data block 2. Assume that the physical address of the area where data block 1 is located is between 0 and 9. The physical address of the area where data block 2 is located is 10-19, then data block 1 and data block 2 are physically continuous data blocks, so the mapping mark 1 corresponding to data block 1 and the mapping mark corresponding to data block 2 can be 3 merged into one map tag.

The above embodiment can dynamically adjust the size of the actual data block on the production volume corresponding to the mapping mark, improve system performance and save resource pool resources; and the above embodiment supports merging consecutive mapping marks, thereby reducing the consumption of memory resources. Improve system performance.

Fig. 2 is a flowchart of another embodiment of the method for writing data to a production volume according to the present invention. As shown in Fig. 2, the method for writing data to a production volume may include:

Step 201, the storage array where the production volume is located receives a data operation request, the data operation request carries data to be written, and the start address and length of the data to be written.

Step 202, determine whether the target area of the above-mentioned data to be written in the production volume has been copied before writing; if not, then perform step 203; if there are at least For a data block that has undergone copy-before-write, step 206 is performed.

Specifically, the production volume, also called the source LUN, can be stored in the mapping table in the memory of the host. When the storage array is running normally, the production volume is read and written in the memory, and will be finally saved to the storage array after the system is powered off. .

Determine whether there is a mapping mark corresponding to the data block covered by the target area of the data to be written in the production volume in the mapping table; if there is no mapping mark corresponding to the data block covered by the target area in the mapping table, then It can be determined that the target area of the above-mentioned data to be written in the production volume has not been copied before writing; if there is a mapping mark corresponding to at least one data block covered by the above-mentioned target area in the mapping table, it can be determined that the above-mentioned data to be written in the production volume There is at least one copy-before-write data block in the data blocks covered by the target area of the incoming data.

In step 203, the storage array where the production volume is located determines the data blocks that need to be copied before writing according to the start address, length and preset storage granularity of the data to be written.

Specifically, the above-mentioned storage array can determine the size of the data block that needs to be copied before writing according to the length of the above-mentioned data to be written and the preset storage granularity, and write according to the start address of the above-mentioned data to be written and the above-mentioned needs. The size of the data block to be copied before determines the position of the data block that needs to be copied before writing.

In this embodiment, the above-mentioned preset storage granularity may be: the size of the data block that needs to be copied before writing is the power of 2 times the minimum value of the preset storage granularity; the above-mentioned preset storage granularity The minimum value of is a power of 2 times, for example: it can be the minimum IO granularity for accessing the disk, that is, the size of a sector.

Step 204 , the storage array where the production volume is located performs copy-before-write on the data blocks that need to be copied before write, and generates a mapping mark corresponding to the above data blocks.

Further, the generated mapping mark may be recorded in the above mapping table.

Step 205, the storage array where the production volume is located writes the data to be written into the data block after the copy-before-write; and then executes step 210.

Step 206, the storage array where the production volume is located splits the data to be written into a segment to be written according to the number of data blocks copied before overwriting, the size of the data blocks copied before overwriting, and the size of the target area into the data segment, or according to the number of data blocks that are copied before overwriting, the size of the data blocks that are copied before overwriting, the position of the data blocks that are copied before overwriting in the target area, and the size of the above target area The data to be written is split into at least two data segments to be written.

For example, when there is only one copy-before-write data block in the data block covered by the target area of the data to be written in the production volume, and the size of the copy-before-write data block is the same as that of the target area When the sizes of the above-mentioned data to be written are the same, it is only necessary to split the above-mentioned data to be written into a data segment to be written, which is equivalent to not splitting the above-mentioned data to be written; when the target of the above-mentioned data to be written in the production volume There is only one copy-before-overwrite data block in the data blocks covered by the area, and the size of the copy-before-write data block is smaller than the size of the target area, if the copy-before-write data block is located in The endpoint position of the above-mentioned target area, the above-mentioned data to be written can be split into at least two sections of data to be written; if the data block to be copied before overwriting is located Split the incoming data into at least three data segments to be written; for example: suppose the target area of the data to be written covers three data blocks, which are data block 1, data block 2 and data block 3 in sequence; if data block 1 or Data block 3 is copied before overwriting, that is, the data block that is copied before overwriting is located at the endpoint of the above-mentioned target area, then the data to be written is split into two data segments to be written. The target area of the data segment to be written covers data block 1 or data block 3, and the target area of another segment of the data segment to be written after splitting covers two data blocks except data block 1 or data block 3; if the data Block 2 is copied before overwriting, that is, the data block that is copied before overwriting is located in the middle of the above-mentioned target area, so the data to be written needs to be split into three data segments to be written, and a segment to be written after splitting The target area of the data segment covers data block 1, the target area of another split data segment to be written covers data block 2, and the split target area of another data segment to be written covers data block 3.

In this embodiment, the data block covered by the target area of the data to be written has only one data block that has been copied before writing as an example to illustrate the splitting method of the data to be written. When the data to be written When there are two or more data blocks that have been copied before writing in the data blocks covered by the target area, the splitting method of the data to be written is similar to the above, and will not be repeated here. Simply put, it is to use At least two data segments to be written after splitting form two groups. The target area of one group of data segments to be written only covers the data blocks copied before overwriting, and the target area of the other group of data segments to be written only covers Overwrite data blocks that have not been copied before write.

Step 207, judging whether the data block covered by the target area to be written in each of the at least two data segments to be written has been copied before writing. For the data segment to be written that has been copied before overwriting for the data block covered by the target area to be written, step 208 is performed; data segment, go to step 209.

Wherein, please refer to the method provided in step 202 for the method of judging whether copy-before-write has been performed, and will not be repeated here.

In step 208, the storage array where the production volume is located writes the data in the data segment to be written into the data block covered by the target area to be written in the data segment to be written, and then executes step 210.

Step 209, the storage array where the production volume is located determines the data blocks that need to be copied before writing according to the start address, length and preset storage granularity of the data segment to be written; copy-before-write, and generate a mapping mark corresponding to the above-mentioned data block; write the data in the above-mentioned data segment to be written into the data block after the copy-before-write; and then perform step 210.

Specifically, the above-mentioned storage array determines the data blocks that need to be copied before writing according to the start address, length and preset storage granularity of the above-mentioned data segments to be written in the same way as that provided in step 203, and will not repeat.

Step 210, judging whether the data to be written has been split. If yes, execute step 211, and if the data to be written has not been split, execute step 212.

Specifically, after the data to be written is split, a split flag can be set for the split data segment to be written, which is used to indicate that the data in the data segment to be written is obtained after splitting. data. In this way, it is only necessary to detect whether there is a split identifier in the data to be written, if there is a split identifier, the data to be written has been split, and if there is no split identifier, the data to be written has not been split. Of course, this is just a way of judging whether the data to be written has been split, and the embodiment of the present invention is not limited thereto.

Step 211, the storage array where the production volume is located returns a write operation success response after at least two split data segments to be written are successfully written; or, in at least two split data segments to be written After at least two data segments to be written fail to be written, a write operation failure response is returned.

Step 212: After the storage array where the production volume is located successfully writes the data to be written into the data block after the copy-before-write, it returns a write operation success response; or, after writing the data to be written into the copy-before-write After subsequent data blocks fail, a write operation failure response is returned.

In this embodiment, the storage array where the production volume is located can also detect the data blocks corresponding to the mapping marks, and merge the mapping marks corresponding to the physically continuous data blocks into one mapping mark; thus, the consumption of memory resources can be reduced, and the system can be improved. performance. For example, after the above storage array detects the data block corresponding to the mapping mark, it is found that the mapping mark 1 corresponds to the data block 1, and the mapping mark 3 corresponds to the data block 2. Assuming that the physical address of the area where the data block 1 is located is 0-9, the data The physical address of the area where block 2 is located is 10 to 19, then data block 1 and data block 2 are physically continuous data blocks, so the mapping mark 1 corresponding to data block 1 and the mapping mark 3 corresponding to data block 2 can be combined into A map tag.

The above embodiment can dynamically adjust the size of the actual data block on the production volume corresponding to the mapping mark, improve system performance and save resource pool resources; and the above embodiment supports merging consecutive mapping marks, thereby reducing the consumption of memory resources. Improve system performance.

The method for writing data to the production volume provided by the embodiment of the present invention will be described below by taking the data operation request as IO and the preset minimum storage granularity as 1 KB as an example.

Scenario 1: The storage array where the production volume is located receives IO1 sent by the host, with a size of 4KB. Assume that the target area to be written to IO1 on the production volume has not been copied before writing, because the size of IO1 is 2 to the power of 2 , so it can be determined that the size of the data block that needs to be copied before writing in the above target area is 4KB, directly copy the 4KB data in the data block that needs to be copied before writing to the resource pool, and generate the mapping mark corresponding to the above data block , and record the generated mapping mark in the mapping table, then write the data to be written in IO1 to the data block after the copy before writing, as shown in Figure 3, Figure 3 is an implementation of the data writing process of the present invention Example diagram.

Scenario 2: The storage array where the production volume is located receives IO2 with a size of 1.5KB. Assume that the target area to be written to IO2 on the production volume has not been copied before writing. Since the size of IO2 is not a power of 2, therefore According to the size of IO2 and the preset storage granularity, it can be determined that the size of the data block that needs to be copied before writing is 2KB, and then the above storage array will copy the 2KB data in the data block that needs to be copied before writing to the resource pool, and Generate a mapping mark corresponding to the above data block, and record the generated mapping mark in the mapping table, and then write the data to be written in IO2 to the data block after the copy before writing, as shown in Figure 4, Figure 4 It is a schematic diagram of another embodiment of the data writing process of the present invention. The shaded area in FIG. 4 represents the area not covered by IO2 in the 2KB copy-before-write data block.

Scenario 3: The storage array where the production volume is located receives IO3 with a size of 1.5MB. Assume that part of the target area to be written to IO3 on the production volume has been copied before writing and a mapping mark is generated. At this time, IO3 needs to be split into at least two small IOs, and then determine the size of the data block according to the method provided by the embodiment of the present invention for each IO after splitting, as shown in Figure 5, the target area to be written in IO3 on the production volume covers the data block 1 to 4, in which data block 2 and data block 3 have been copied before writing. Since data block 2 and data block 3 are located in the middle of the above target area, IO3 needs to be split into three IOs, which are respectively recorded as IO3 -1, IO3-2 and IO3-3.

Among them, the data blocks covered by the target area to be written by IO3-2 are data block 2 and data block 3. Since data block 2 and data block 3 have been copied before writing, the data in IO3-2 is directly written The data block covered by the target area to be written by IO3-1 is data block 1, and the data block covered by the target area to be written by IO3-3 is data block 4. Assume that data block 1, data The sizes of block 2, data block 3, and data block 4 are 256KB, 512KB, 512KB, and 256KB respectively, so it is necessary to copy the 256KB data on data block 1 and data block 4 to the resource pool respectively, and generate 1 and data block 4 corresponding to the mapping mark, and record the generated mapping mark in the mapping table, and then respectively write the data to be written in IO3-1 and IO3-3 to data block 1 and The data block 4 is shown in FIG. 5 , which is a schematic diagram of another embodiment of the data writing process of the present invention.

Scenario 1 and Scenario 2 mentioned above are scenarios for modifying small data blocks, for example, business scenarios for modifying small data blocks such as bank depositor data database modification. Scenario 3 is for scenarios for modifying large data blocks, such as website image modification and video library application Business scenarios such as large data block modification. To sum up, the method for writing data to the production volume provided by the embodiment of the present invention can dynamically adjust the size of the snapshot data block according to the application of the actual business, and can rationally allocate and utilize storage system resources, so that the consumption of memory resources and resource pool resources is minimized. excellent.

Further, in the embodiment of the present invention, the storage array where the production volume is located can also detect the data block corresponding to the mapping mark, and combine the mapping marks corresponding to the physically continuous data blocks into one mapping mark; thus, the memory resource consumption can be reduced. Consumption, improve system performance. Fig. 6 is a schematic diagram of an embodiment of the present invention combining continuous mapping marks. As shown in Fig. 6, data block 1, data block 2, data block 3 and data block 4 are physically continuous data blocks, and data block 1 can be , data block 2, data block 3, and data block 4 corresponding mapping tags are merged into one mapping tag.

The method for writing data to the production volume provided by the embodiment of the present invention can rationally allocate and utilize storage system resources, reduce the consumption of memory resources and resource pool resources, and support the merging of mapping marks, which can improve the read performance of the storage system. The implementation of the present invention The method for writing data to the production volume provided in this example is not only applicable to virtual snapshots, but also applicable to replication services such as clone mirroring, remote replication, and LUN copy.

Those of ordinary skill in the art can understand that all or part of the steps for realizing the above-mentioned method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the It includes the steps of the above method embodiments; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk and other various media that can store program codes.

Figure 7 is a schematic structural diagram of an embodiment of the device for writing data to the production volume according to the present invention. The device for writing data to the production volume in this embodiment can be used as a storage array or a part of the storage array to implement the present invention as shown in Figure 1 Example flow. As shown in FIG. 7 , the apparatus for writing data to the production volume may include: a receiving module 71 , a determining module 72 , a generating module 73 and a writing module 74 .

Wherein, the receiving module 71 is configured to receive a data operation request, the data operation request carries the data to be written, and the start address and length of the data to be written.

The determining module 72 is configured to determine that pre-write is required according to the starting address, length and preset storage granularity of the data to be written when the target area of the data to be written in the production volume has not been copied before writing. The copied data block; specifically, the determination module 72 can determine the size of the data block that needs to be copied before writing according to the length of the above-mentioned data to be written and the preset storage granularity, and according to the above-mentioned starting address of the data to be written and the size of the data block that needs to be copied before writing determines the position of the data block that needs to be copied before writing; in this embodiment, the above-mentioned preset storage granularity can be: the size of the data block that needs to be copied before writing is The power of 2 times the minimum value of the preset storage granularity; the minimum value of the above preset storage granularity is the power of 2 times, for example: it can be the minimum IO granularity for accessing the disk, that is, one sector the size of.

The generating module 73 is configured to perform copy-before-write on the data blocks determined by the determining module 72 that require copy-before-write, and generate mapping marks corresponding to the above-mentioned data blocks.

The writing module 74 is configured to write the data to be written received by the receiving module 71 into the data block after the copy-before-write is performed.

The above-mentioned device for writing data to the production volume can dynamically adjust the size of the actual data block on the production volume corresponding to the mapping mark, improve system performance and save resource pool resources.

Fig. 8 is a schematic structural diagram of another embodiment of the device for writing data to the production volume according to the present invention. The device for writing data to the production volume in this embodiment can be used as a storage array or a part of the storage array to implement the present invention. The process flow of the embodiment shown in Figure 2. Compared with the device for writing data to the production volume shown in FIG. 7, the difference is that the device for writing data to the production volume shown in FIG. A response return module 77 , a second response return module 78 and a combination module 79 .

Wherein, the splitting module 75 is configured to, when there is at least one data block that is copied before overwriting in the data block covered by the target area of the data to be written in the above production volume, according to the data that is copied before overwriting The number of blocks, the size of the above-mentioned data blocks that are copied before overwriting, and the size of the above-mentioned target area, split the data to be written into a data segment to be written, or according to the above-mentioned number of data blocks that are copied before overwriting , the size of the above-mentioned data block copied before overwriting, the position of the above-mentioned data block copied before overwriting in the target area, and the size of the above-mentioned target area Split the data to be written into at least two data segments to be written , so that at least two data segments to be written after splitting are obtained into two groups. The target area of a group of data segments to be written only covers the data blocks copied before overwriting, and the target area of the other group of data segments to be written Regions only cover data blocks that have not been copy-before-written. Wherein, the size of the target area is the length of the data to be written.

A judging module 76, configured to judge whether the data block covered by the target area to be written in each of the at least two data segments to be written has been copied before writing.

In this embodiment, the determining module 72 can also, when the data block covered by the target area of the data segment to be written has not been copied before writing, according to the start address, length and pre-set address of the data segment to be written The set storage granularity determines the data blocks that need to be copied before writing; then the generating module 73 can also perform copy-before-writing on the data blocks that need to be copied before writing determined by the above-mentioned determining module 72, and generate a mapping corresponding to the above-mentioned data blocks Marking; the writing module 74 can also write the data in the data segment to be written into the data block after the copy-before-write; and overwrite the data block covered by the target area to be written in the data segment During pre-copying, the data in the data segment to be written is written into the data block covered by the target area to be written in the data segment to be written.

The first response returning module 77 is configured to return a write operation success response after the at least two data segments to be written are successfully written; or, at least one segment of the data to be written in the at least two data segments to be written After segment writing fails, return write operation failure response.

The second response returning module 78 is used to return the write operation success response after the data to be written is successfully written to the data block after the copy before writing; or, after the data to be written is written to the copy before writing After the data block fails, a write operation failure response is returned.

The merging module 79 is configured to detect the data block corresponding to the mapping mark, and merge the mapping marks corresponding to the physically continuous data blocks into one mapping mark.

The above-mentioned device for writing data to the production volume can dynamically adjust the size of the actual data block on the production volume corresponding to the mapping mark, improve system performance and save resource pool resources; and the above-mentioned embodiment supports the merging of continuous mapping marks, so that Reduce memory resource consumption and improve system performance.

Those skilled in the art can understand that the drawing is only a schematic diagram of a preferred embodiment, and the modules or processes in the drawing are not necessarily necessary for implementing the present invention.

Those skilled in the art can understand that the modules in the device in the embodiment can be distributed in the device in the embodiment according to the description in the embodiment, or can be located in one or more devices different from the embodiment according to corresponding changes. The modules in the above embodiments can be combined into one module, and can also be further split into multiple sub-modules.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: it can still be Modifications are made to the technical solutions described in the foregoing embodiments, or equivalent replacements are made to some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims (8)

1. one kind writes the method for data to producing volume, it is characterized in that, comprising:

Receive data operation request, described data operation request is carried data to be written, and the start address of described data to be written and length;

Before the target area that produces data to be written described in the volume did not carry out writing during copy, according to start address, length and the predefined storage granularity of described data to be written determine needs write before the data block of copy;

When existence was carried out writing the data block of preceding copy in the data block that the target area of data to be written described in the described production volume covers, according to the described quantity of carrying out writing the preceding data block that copies, described carried out writing before the size of data block of copy and the size of described target area, described data to be written are split as one section data segment to be written, perhaps according to the described quantity of carrying out writing the preceding data block that copies, the described size of carrying out writing the preceding data block that copies, described carried out writing before the data block of copy in described target area the position and the described data to be written of big young pathbreaker of described target area be split as at least two sections data segments to be written, make at least two sections data segments to be written that split the back acquisition form two groups, the data block of copy before the target area of one group of data segment to be written only covers and carried out writing, the data block of copy before another target area of organizing data segment to be written only covers and do not carry out writing;

Described needs are write the data block of preceding copy and write preceding copy, and the generation map tag corresponding with described data block;

Described data to be written are write preceding copy data block afterwards.

2. method according to claim 1 is characterized in that, described start address according to described data to be written, length and predefined storage granularity determine that the data block that copies before needs are write comprises:

According to length and the predefined storage granularity of described data to be written, the size of the data block of copy before determining to write;

The size of the data block of copy before writing according to the start address of described data to be written and described needs, determine described needs write before the position of data block of copy.

3. method according to claim 1, it is characterized in that, described according to described carried out writing before copy data block quantity, described carried out writing before copy data block size, described carried out writing before the data block of copy position and the described data to be written of big young pathbreaker of described target area in described target area be split as after at least two sections data segments to be written, also comprise:

Judge whether the data block that the target area to be written of every section data segment to be written in described at least two sections data segments to be written covers carried out writing preceding copy;

The data block that covers for the target area to be written of every section data segment to be written in described at least two sections data segments to be written is the data segment to be written of copy before not carrying out writing, according to start address, length and the predefined storage granularity of every section data segment to be written in described at least two sections data segments to be written determine needs write before the data block of copy; The data block that the described needs of determining according to start address, length and the predefined storage granularity of every section data segment to be written in described at least two sections data segments to be written are write preceding copy is write preceding copy, and the generation map tag corresponding with described data block; Data block before data in every section data segment to be written in described at least two sections data segments to be written are write after the copy;

The data block that covers for the target area to be written of every section data segment to be written at least two sections data segments to be written is the data segment to be written of copy before carrying out writing, and the data in every section data segment to be written in described at least two sections data segments to be written are write the data block that the target area to be written of every section data segment to be written in described at least two sections data segments to be written covers.

4. according to any described method of claim 1-3, it is characterized in that, described with described data to be written write before the copy after data block after, also comprise:

Data block to the map tag correspondence detects, and the map tag of continuous data block correspondence is physically merged into a map tag.

5. one kind to producing the device that volume writes data, it is characterized in that, comprising:

Receiver module is used for receiving data operation request, and described data operation request is carried data to be written, and the start address of described data to be written and length;

Determination module when being used for before the target area of data to be written described in the described production volume did not carry out writing copy, is determined the data block that copies before needs are write according to start address, length and the predefined storage granularity of described data to be written;

Split module, when being used for the data block of copy before the data block existence that the target area of data to be written described in the described production volume covers was carried out writing, according to the described quantity of carrying out writing the preceding data block that copies, described carried out writing before the size of data block of copy and the size of described target area, described data to be written are split as one section data segment to be written, perhaps according to the described quantity of carrying out writing the preceding data block that copies, the described size of carrying out writing the preceding data block that copies, described carried out writing before the data block of copy in described target area the position and the described data to be written of big young pathbreaker of described target area be split as at least two sections data segments to be written, make at least two sections data segments to be written that split the back acquisition form two groups, the data block of copy before the target area of one group of data segment to be written only covers and carried out writing, the data block of copy before another target area of organizing data segment to be written only covers and do not carry out writing;

Generation module is used for that the needs that described determination module is determined are write the preceding data block that copies and writes preceding copy, and the generation map tag corresponding with described data block;

Writing module is used for the data to be written that described receiver module receives are write preceding copy data block afterwards.

6. device according to claim 5 is characterized in that,

Described determination module, concrete be used for according to the length of described data to be written and predefined storage granularity determine needs write before the size of data block of copy, and before writing according to the start address of described data to be written and described needs the size of the data block of copy determine described needs write before the position of data block of copy.

7. device according to claim 5 is characterized in that, also comprises:

Judge module, be used for to judge according to described carried out writing before copy data block quantity, described carried out writing before copy data block size, described carried out writing before the data block of copy after the described data to be written of big young pathbreaker of the position of described target area and described target area are split as at least two sections data segments to be written, whether the data block that the target area to be written of every section data segment to be written covers carried out writing preceding copy;

Described determination module, before the data block that also is used for covering when the target area to be written of every section data segment to be written of described at least two sections data segments to be written did not carry out writing during copy, determine the data block that copies before needs are write at least according to start address, length and the predefined storage granularity of every section data segment to be written in described two sections data segments to be written;

Described generation module, be used for that also the needs that described determination module is determined according to start address, length and the predefined storage granularity of every section data segment to be written of described two sections data segments to be written are at least write the preceding data block that copies and write preceding copy, and the generation map tag corresponding with described data block;

The said write module also is used for the data in every section data segment to be written of described at least two sections data segments to be written are write preceding copy data block afterwards; And the data block that covers when the target area to be written of every section data segment to be written in described at least two sections data segments to be written carried out writing before when copy, the data in every section data segment to be written in described at least two sections data segments to be written are write the data block that the target area to be written of every section data segment to be written in described at least two sections data segments to be written covers.

8. according to any described device of claim 5-7, it is characterized in that, also comprise:

Merge module, be used for the data block of map tag correspondence is detected, the map tag of continuous data block correspondence is physically merged into a map tag.

Images