WO2016086819A1

WO2016086819A1 - Method and apparatus for writing data into shingled magnetic record smr hard disk

Info

Publication number: WO2016086819A1
Application number: PCT/CN2015/095997
Authority: WO
Inventors: 张雷; 肖垒
Original assignee: 华为技术有限公司
Priority date: 2014-12-05
Filing date: 2015-11-30
Publication date: 2016-06-09
Also published as: CN104461390A; CN104461390B

Abstract

A method and apparatus for writing data into a shingled magnetic record SMR hard disk relate to the technical field of computers. The method for writing data into a shingled magnetic record SMR hard disk comprises: receiving a write data command, wherein the write data comprises a key, data and a flag, the key is used for identifying the data, the flag is a write flag, a deletion flag or a modification flag (201); according to the length of a data record, applying to chunk which is being used for a space, wherein the data record comprises the key, data and flag in the write data command (202); writing the data record into the applied space (203); and saving addresses in the chunk which is being used where the key, the flag and the data are written (204). By writing the data record comprising the key, the data and the flag into the chunk which is being used, the method and apparatus mentioned above solve the problem that the layout scheme of a distributed storage system cannot be directly transplanted onto the SMR hard disk; the effect of improving the flexibility in writing data into an SMR hard disk is realized.

Description

Method and device for writing data into shingled magnetic recording SMR hard disk

This application claims priority to Chinese Patent Application No. 201410736477.6, entitled "Method and Apparatus for Writing Data into a Stacked Magnetic Recording SMR Hard Disk", on December 5, 2014, all of which is filed on December 5, 2014. The content is incorporated herein by reference.

Technical field

The present invention relates to the field of computer technologies, and in particular, to a method and apparatus for writing data into a stacked magnetic recording SMR hard disk.

Background technique

SMR (Shingled Magnetic Record) technology can reduce the track width by partially overlapping the existing hard disk tracks, thereby increasing the recording density.

The distributed Key-Value storage system has the advantages of fast query speed, large amount of data storage, and high concurrency support. It is very suitable for querying through the primary key.

In the process of implementing the present invention, the inventors have found that the prior art has at least the following problem: since the track of the SMR hard disk is very narrow, data in a certain number of tracks adjacent thereto will be written when data is written into a certain track. Write coverage is generated, causing data in adjacent tracks to fail. Therefore, the Key-Value storage system layout scheme cannot be directly transplanted to the SMR hard disk, resulting in low write performance of the SMR hard disk.

Summary of the invention

In order to solve the problem that the Key-Value storage system layout solution cannot be directly transplanted to the SMR hard disk, the embodiment of the present invention provides a method and device for writing data into the stacked magnetic recording SMR hard disk. The technical solution is as follows:

In a first aspect, a method for writing data to a shingled magnetic recording SMR hard disk is provided. The SMR hard disk includes a plurality of chunks obtained by virtualizing a track band, the method comprising:

Receiving a write data instruction, the write data instruction including a key, a data, and a mark, the key being used to identify the data, the mark being a write mark, a delete mark, or a modify mark, the write mark being used to indicate that the Data is written into the SMR hard disk, and the delete flag is used to delete the SMR hard disk. Data having the key, the modified mark being used to indicate that data having the same key and having been written to the SMR hard disk is modified with the data;

Requiring a space to the chunk being used according to the length of the data record, the data record including the key, the data and the token in the write data instruction;

Writing the data record into the space of the application;

The address is saved, the tag and the data are written to the address in the chunk being used.

In a first possible implementation manner of the first aspect, the chunk being used includes a data area for storing a data record, and the space is applied to the chunk being used according to the length of the data record, including:

A space is requested from the data area in the chunk being used according to the length of the data record.

With reference to the first aspect or the first possible implementation manner of the first aspect, in a second possible implementation, the chunk being used includes an index area for storing an index record, where the key is saved The tag and the data are written to the address in the chunk being used, including:

Forming an index record by the key, the mark, and an address in which the data is written into the chunk being used;

Saving the index record to a first storage area in memory corresponding to the chunk being used; or, when the remaining space in the chunk being used is insufficient to allocate data to be written in the write data instruction, The index record in the first storage area in the memory is stored in the index area in the chunk being used.

With reference to the first aspect, the first possible implementation of the first aspect, or the second possible implementation of the first aspect, in a third possible implementation, the After the index record in a storage area is stored in the index area in the chunk being used, the method further includes:

Modify the attribute of the chunk being used to used, and apply for an unused chunk as the used chunk for the next write data;

And storing the index record in the first storage area into a second storage area in the memory corresponding to the used chunk;

Clearing the index record in the first storage area.

Combining the first aspect, the first possible implementation manner of the first aspect, the second In a fourth possible implementation manner, the method further includes:

When the hard disk is restarted, the used chunk and the chunk being used are determined according to the attributes of the chunk;

And storing an index record stored in the used chunk into a second storage area in the memory corresponding to the used chunk;

Reconstructing an index record corresponding to the data record according to the data record stored in the chunk being used, and writing the reconstructed index record into the first storage area in the memory corresponding to the chunk being used .

With reference to the first aspect, the first possible implementation of the first aspect, to any one of the fourth possible implementation manners of the first aspect, in the fifth possible implementation, the reconstruction is obtained The index record includes: a key of data in the data record, a tag corresponding to the data, and an address in which the data is written in the chunk being used.

In a second aspect, there is provided an apparatus for writing data to a shingled magnetic recording SMR hard disk, the SMR hard disk comprising a plurality of chunks obtained by virtualizing a track band, the device comprising:

a receiving module, configured to receive a write data instruction, where the write data instruction includes a key, a data, and a mark, wherein the key is used to identify the data, and the mark is a write mark, a delete mark, or a modify mark, and the write mark is used by And indicating that the data is written into the SMR hard disk, the deletion mark is used to indicate deleting data having the key in the SMR hard disk, and the modification mark is used to indicate that the same key is modified by using the data and Data that has been written to the SMR hard disk;

An application module, configured to apply for a space to a chunk being used according to a length of the data record, where the data record includes the key, the data, and the mark in the write data instruction;

a first writing module, configured to write the data record into the space of the application;

And a saving module, configured to save the key, the mark, and the address in which the data is written in the chunk being used.

In a first possible implementation manner of the second aspect, the chunk being used includes a data area for storing a data record, and the application module is configured to use the length according to the length of the data record. The data area application space in the chunk.

With reference to the second aspect, or the first possible implementation manner of the second aspect, in a second possible implementation, the chunk being used includes an index area for storing an index record, where Storage module, including:

a component unit, configured to form an index record of the key, the mark, and an address in which the data is written into the chunk being used;

a saving unit, configured to save the index record to a first storage area in the memory corresponding to the chunk being used; or, the remaining space in the chunk being used is insufficient to allocate a write data instruction to be written And storing, in the data, the index record in the first storage area in the memory to the index area in the chunk being used.

With reference to the second aspect, the first possible implementation of the second aspect, or the second possible implementation of the second aspect, in a third possible implementation, the apparatus further includes:

a modifying module, configured to modify the attribute of the used chunk to be used, and apply an unused chunk as a used chunk for writing data next time;

a storage module, configured to store the index record in the first storage area to a second storage area in the memory corresponding to the used chunk;

And a clearing module, configured to clear the index record in the first storage area.

With reference to the second aspect, the first possible implementation of the second aspect, the second possible implementation of the second aspect, or the third possible implementation of the second aspect, in a fourth possible implementation The device further includes:

a determining module, configured to determine, according to the attribute of the chunk, the used chunk and the chunk being used when the hard disk is restarted;

a second writing module, configured to write an index record stored in the used chunk to a second storage area in the memory corresponding to the used chunk;

a reconstruction module, configured to reconstruct an index record corresponding to the data record according to the data record stored in the chunk being used, and write the reconstructed index record into the memory to correspond to a chunk being used The first storage area.

With reference to the second aspect, the first possible implementation manner of the second aspect, to any one of the fourth possible implementation manners of the second aspect, in the fifth possible implementation manner, the reconstruction is obtained The index record includes: a key of data in the data record, a tag corresponding to the data, and an address in which the data is written in the chunk being used.

In a third aspect, an apparatus for writing data into a stacked magnetic recording SMR hard disk is provided, which is applied to a device including the SMR hard disk, wherein the SMR hard disk includes a plurality of virtual bands according to a track band The resulting chunk, the apparatus comprising: a receiver and a processor, wherein the processor is coupled to the receiver;

The receiver is configured to receive a write data instruction, the write data instruction includes a key, a data and a mark, the key is used to identify the data, and the mark is a write mark, a delete mark or a modify mark, and the write Marking is used to indicate that the data is written into the SMR hard disk, and the delete flag is used to delete data having the key in the SMR hard disk, and the modified mark is used to indicate that the data modification has the same Key and data that has been written to the SMR hard disk;

The processor, configured to apply for a space to a chunk being used according to a length of the data record, where the data record includes the key, the data, and the mark in the write data instruction;

The processor is further configured to write the data record into the space of the application;

The processor is further configured to save the key, the mark, and an address in which the data is written into the chunk being used.

In a first possible implementation manner of the third aspect, the chunk being used includes a data area for storing a data record, and the processor is further configured to use the length according to the length of the data record The data area application space in the chunk.

With reference to the third aspect, or the first possible implementation manner of the third aspect, in a second possible implementation, the chunk being used includes an index area for storing an index record, and the processor further Forming an index record for the address, the mark, and an address in which the data is written into the chunk being used;

The processor is further configured to save the index record to a first storage area in a memory corresponding to a chunk being used; or, the remaining space in the chunk being used is insufficient to allocate a write data instruction When the data is written, the index record in the first storage area in the memory is stored to the index area in the chunk being used.

With reference to the third aspect, the first possible implementation manner of the third aspect, or the second possible implementation manner of the third aspect, in a third possible implementation, the processor is further configured to: The attribute of the chunk being used is modified to be used, and an unused chunk is requested as the used chunk for the next write of the data;

The processor is further configured to store the index record in the first storage area to a second storage area in the memory corresponding to a used chunk;

The processor is further configured to clear the index record in the first storage area.

Combining the third aspect, the first possible implementation manner of the third aspect, and the second A possible implementation or a third possible implementation of the third aspect. In a fourth possible implementation, the processor is further configured to determine, according to an attribute of the chunk, the used Chunk and the chunk being used;

The processor is further configured to write an index record stored in the used chunk to a second storage area in the memory corresponding to the used chunk;

The processor is further configured to reconstruct an index record corresponding to the data record according to the data record stored in the chunk being used, and write the reconstructed index record into the memory and in use The chunk corresponds to the first storage area.

With reference to the third aspect, the first possible implementation manner of the third aspect, or any one of the fourth possible implementation manners of the third aspect, in the fifth possible implementation manner, the reconstruction is obtained The index record includes: a key of data in the data record, a tag corresponding to the data, and an address in which the data is written in the chunk being used.

The beneficial effects brought by the technical solutions provided by the embodiments of the present invention are:

By writing data records containing keys, data, and tags into the chunks being used; since the data records generated according to the write data instructions contain keys and data, it is solved that the distributed storage system layout scheme cannot be directly transplanted to On the SMR hard disk, the write performance of the SMR hard disk is low. Since the chunk is obtained according to the band virtualization, data can be randomly written between the chunks, thereby achieving the flexibility of writing data on the SMR hard disk. effect.

DRAWINGS

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly described below. It is obvious that the drawings in the following description are only some embodiments of the present invention. Other drawings may also be obtained from those of ordinary skill in the art in light of the inventive work.

1 is a schematic diagram showing the spatial layout of an SMR hard disk provided in an embodiment of the present invention;

2 is a flow chart of a method for writing data into a stacked magnetic recording SMR hard disk according to an embodiment of the present invention;

3A is a flow chart of a method for writing data into a stacked magnetic recording SMR hard disk according to another embodiment of the present invention;

FIG. 3B is a schematic diagram of a process of switching a chunk provided in an embodiment of the present invention; FIG.

4 is a schematic diagram of a garbage cleaning process provided in an embodiment of the present invention;

5 is a flow chart of a method for writing data into a stacked magnetic recording SMR hard disk according to still another embodiment of the present invention;

6 is a block diagram showing the structure of an apparatus for writing data into a stacked magnetic recording SMR hard disk according to an embodiment of the present invention;

7 is a block diagram showing the structure of an apparatus for writing data into a stacked magnetic recording SMR hard disk according to another embodiment of the present invention;

FIG. 8 is a block diagram showing the structure of an apparatus including an SMR hard disk according to an embodiment of the present invention; FIG.

FIG. 9 is a block diagram showing the structure of an apparatus including an SMR hard disk according to another embodiment of the present invention.

detailed description

The embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

Please refer to FIG. 1 , which illustrates a spatial layout of an SMR hard disk provided in an embodiment of the present invention. The SMR hard disk may be logically divided into a system metadata area 110 and a recording area 120 .

The system metadata area 110 includes at least one super record block 130, and the super record block 130 includes at least one super block 140 for recording attributes and usage information of each chunk 150 in the record area 120.

Since the super block 140 needs to frequently update the attributes and usage information of each chunk 150 in the recording area 120, if a super block 140 is used for recording, the disk space of the super block 140 may be faulty. Therefore, a plurality of super blocks 140 are required to record the attributes and usage information of each chunk 150 in the recording area 120. The size of the super block 140 can be determined according to the size of the information in the recording area 120. The number of the super blocks 140 can be determined according to the number of times the super block 140 can be repeatedly written, and the more the super block 140 can be repeatedly written. More, the number of super blocks 140 can be set smaller. The number of super record blocks 130 can be determined according to the number and size of the super blocks 140. In the actual application scenario, the number of super blocks 140 in the super record block 130 and the information contained in the super block 140 may be changed according to the actual use.

Optionally, the super block 140 in the super record block 130 may further include a first global variable, where the first global variable is used to identify the sequential use order of each super block 140. For example, you can apply for each After a super block 140, the first global variable is incremented by 1 as the first global variable of the applied super block 140. For example, currently, after applying to the super block 140, if the first global variable of the super application 140 of the previous application is found to be 100, the first global variable is incremented by 1 to obtain a new first global variable 101. A global variable 101 is taken as the first global variable of the current super block 140. It should be noted that the value added by the first global variable may also be other values, such as 2, 3, etc., only need to ensure that the first global variable of the super-block 140 applied later is greater than that of the super-block 140 of the previous application. The first global variable.

The recording area 120 includes at least one chunk 150, and the chunk 150 here is obtained by virtualizing the track band in the SMR hard disk. According to the usage of chunk150, chunk 150 can be divided into used chunks, chunks in use, and unused chunks. The recording area 120 is for storing data records and index records corresponding to the data records.

It should be noted that the type of the chunk 150 in the recording area 120 may be changed according to the actual use. Optionally, the recording area 120 may further include other types of chunks.

Optionally, the chunk 150 in the recording area 120 includes a data area 160 and an index area 170.

The data area 160 can be used to store at least one data record, each data record including at least the first two or all of the keys, tags, and data.

The type of information contained in the data record can be determined based on the type of tag of the data record. For example, when the mark of the data record is a delete mark, the delete record may include only the key and the mark; when the mark type of the data record is a write mark or a modified mark, the write record or the modified record may include a key and a mark. And data.

Optionally, the data record may further include other information, such as the check information, the key length information, and the data length information. The other information included in the data record is not specifically limited in this embodiment.

Optionally, the data record may further include a second global variable, where the second global variable is used to identify a sequence in which each data record is written. For example, each time a data record is written, the second global variable of the previously written data record can be incremented by one as the second global variable of the currently written data record. For example, when writing the current data record, if the second global variable of the data record written last time is 200, the second global variable is incremented by 1 to obtain a new second global variable 201, and the second The global variable 201 acts as the second global variable of the current data record. It should be noted that the value added by the second global variable may also be other values, such as 2, 3, etc., only need to ensure that the second global variable of the data record written later is greater than the data record of the previous write. The second global variable.

It should be noted that the relative positions of the keys, the markers, and the data in the data record may be changed according to the actual use.

The index area 170 is configured to store index records corresponding to the data records one by one, and the index records corresponding to the data records include at least the same keys, tags, and addresses recorded in the data records on the SMR hard disk.

The data records in the data area 160 have a one-to-one correspondence with the index records in the index area 170. For example, as shown in FIG. 1, the data record 1 corresponds to the index record 1, and the data record 2 corresponds to the index record 2. It should be noted that the relative positions of the data record 1, the data record 2, the index record 1, and the index record 2 in FIG. 1 are only schematic, which is not limited in this embodiment.

Optionally, the index record may further include a second global variable corresponding to the data record. The relative positions of the keys, tags, and addresses in the index record can be changed according to the actual use.

It should be noted that the relative positions of the system metadata area 110 and the recording area 120 on the SMR hard disk may be changed according to the actual use.

In summary, the SMR hard disk provided in this embodiment divides the SMR hard disk into a system metadata area and a recording area, and the recording area is used for storing data records and index records corresponding to the data records, and the system metadata area is used for recording. The attribute and usage information of each chunk in the recording area; since the index record is established for each data record in the SMR hard disk, the solution of the distributed storage system layout cannot be directly transplanted to the SMR hard disk, resulting in the writing of the SMR hard disk. The problem of lower performance is achieved; since the chunk is obtained according to the band virtualization, data can be randomly written between the chunks, thereby achieving the effect of improving the flexibility of writing data on the SMR hard disk.

Please refer to FIG. 2, which is a flowchart of a method for writing data into a stacked magnetic recording SMR hard disk according to an embodiment of the present invention. The method for writing data into a stacked magnetic recording SMR hard disk can be used for In a device including an SMR hard disk, the method for writing data into a stacked magnetic recording SMR hard disk may include:

Step 201: Receive a write data instruction, where the write data instruction includes a key, a data, and a mark, the key is used to identify the data, and the mark is a write mark, a delete mark, or a modify mark, and the write mark is used to indicate that the data is written into the SMR hard disk, and is deleted. The mark is used to indicate that the data having the key in the SMR hard disk is deleted, and the modification mark is used to indicate that the data having the same key and written to the SMR hard disk is modified by using the data;

Step 202: Apply space to the chunk being used according to the length of the data record, and record the data packet. Including the keys, data, and tags in the data instructions;

Step 203, writing a data record into the space of the application;

In step 204, the save key, the tag, and the data are written to the address in the chunk being used.

In summary, the method for writing data into the shingled magnetic recording SMR hard disk provided by the embodiment is to write the data record containing the key, the data and the mark into the chunk being used; since the data is generated according to the write data instruction. The data record contains keys and data. Therefore, it solves the problem that the distributed storage system layout scheme cannot be directly transplanted to the SMR hard disk, resulting in low write performance of the SMR hard disk; since the chunk is obtained according to the band virtualization, Data can be randomly written between chunks, thus achieving the flexibility of writing data on the SMR hard disk.

In actual use, in order to facilitate the reading of the data stored in the SMR hard disk, the device in which the SMR is located needs to persist the index record of the chunk. For details, please refer to FIG. 3A, which shows another embodiment of the present invention. A flow chart of a method of writing data into a stacked magnetic recording SMR hard disk, the method of writing data into a stacked magnetic recording SMR hard disk can be used in a device including an SMR hard disk, which writes data into a stacked magnetic field The method of recording the SMR hard disk may include:

Step 301: Receive a write data instruction.

The write data instructions here include keys, data, and tags. The keys are used to identify data, marked as write marks, delete marks, or modified marks. Write marks are used to indicate that the data is written to the SMR hard disk, and the delete mark is used to indicate that the SMR is deleted. The data in the hard disk has the key, and the modification flag is used to indicate that the data having the same key and written to the SMR hard disk is modified by the data.

Step 302: Apply space to the data area in the chunk being used according to the length of the data record;

Data records include keys, data, and tags in a write data instruction.

Optionally, before step 302, it may also generally include generating a data record according to the write data instruction.

After receiving the write data command, the device where the SMR is located can generate a data record according to the data record format according to the write data command. When the data command does not include data or the included data is empty, the data record may include only keys and tags. For example, when the write data instruction is marked as a delete tag, the write data instruction usually does not include data.

After receiving the write data command, the device in which the SMR is located determines the type of the write data command, and generates a corresponding data record according to the mark type of the write data command. For the specific implementation, refer to the following three implementation manners.

In one implementation, when the write data instruction includes a mark as a delete mark, the key and the mark carried in the write data command are combined into one data record according to the format of the data record.

In another implementation, when the write data instruction includes a mark as a write mark, the data, keys, and tags carried in the write data command are combined into one data record according to the format of the data record.

In still another implementation, when the write data instruction includes a mark as a modification mark, the data, the key, and the mark carried in the write data command are combined into one data record according to the format of the data record.

That is, when the write data instruction includes a mark as a delete mark, the data record generated according to the write data instruction usually includes a key and a mark; and when the write data instruction includes a mark as a write mark or a modified mark, according to the Data records generated by write data instructions typically contain keys, tags, and data.

The device where the SMR is located can calculate the length of the data record, and obtain the attributes of each chunk in the record area through the super block in the system metadata area of the SMR hard disk, and determine the chunk currently being used according to the attributes of each chunk, and in the memory. Finding an index record corresponding to the data record in the chunk being used, determining, according to the address of the last written data in the found index record and the length of the last written data, determining that the chunk being used is finally The location of the write. The last determined position to be written here refers to the last position at which data is last written.

The device in which the SMR is located allocates the same space as the length of the generated data record from the last written position in the chunk currently being used, based on the length of the generated data record.

It should be noted that when the remaining space in the chunk being used is smaller than the length of the generated data record, it is necessary to apply for an unused chunk for space allocation.

Step 303, writing a data record into the space of the application;

Step 304, forming an index record by combining the key, the tag and the address in which the data is written into the chunk being used;

The keys and tags here are the same as the keys and tags in the data record. The device where the SMR is located determines the combination of the keys, tags, and addresses written to the chunks being used as the index record for the data record.

It should be noted that the relative position of the key, the mark, and the address in the index record is not limited in this embodiment.

Step 305: Save the index record to a first storage area in the memory corresponding to the chunk being used;

The first storage area here is used to temporarily store the number of pieces written in the chunk currently being used. According to the index record corresponding to the record, the memory here can be the system memory of the device where the SMR is located.

In actual use, when it is detected that the space in the chunk being used is allocated, the device where the SMR is located needs to persist the index record of the chunk being used, and needs to apply for a new chunk being used for subsequent The data is written, and correspondingly, a new super block needs to be applied to record the corresponding record information. For the specific process, refer to steps 306 to 309 below.

Step 306, when the remaining space in the chunk being used is insufficient to allocate the data to be written in the write data instruction, storing the index record in the first storage area in the memory to the index area in the chunk being used;

When the remaining space in the data area of the chunk being used is insufficient to store the data corresponding to the currently received write data instruction, the device in which the SMR is located determines that the space of the chunk being used is allocated. At this time, the device where the SMR is located stores the index record in the first storage area in the memory to the index area in the chunk being used.

Step 307, modify the attribute of the chunk being used to be used, and apply for an unused chunk as the used chunk for the next write data;

After the SMR device detects that the space of the chunk being used is allocated, and writes each index record stored in the first storage area to the index area of the chunk being used, the chunk being used is no longer used. The data record continues to be written, so the property of the chunk being used can be changed to used at this time, so that the chunk being used becomes the used chunk.

At the same time, the device where the SMR is located needs to apply for the next super block in the system metadata area, and records the attributes and usage information of each chunk in the current recording area in the applied super block.

For example, please refer to FIG. 3B , which is a schematic diagram of a process of switching a chunk provided in an embodiment of the present invention. As shown in FIG. 3B, after detecting that the data area 321 of the chunk 320 being used is allocated, the device in which the SMR is located records the index corresponding to the data record of the data area 321 of the chunk 320 being used from the memory. A storage area is stored in the index area 322 of the chunk 320 being used, and the attribute of the chunk 320 being used is changed to used, and becomes the used chunk 330. At the same time, the device in which the SMR is located requests an unused chunk 340, changes the attribute of the unused chunk 340 to be used, becomes the chunk 350 being used, and starts to allocate space in the data area 351 of the chunk 350 being used. After the SMR is located, the device changes the attributes of the chunk 320 and the unused chunk 340 that are being used, and applies a super block 360, and records the attributes and usage information of each chunk in the current recording area in the newly applied super block 360.

Step 308, storing an index record in the first storage area into a second storage area in the memory corresponding to the used chunk;

The second storage area here is used to store index records in all used chunks in the SMR hard disk.

In order to find the valid data saved in the SMR hard disk, the device where the SMR is located needs to write the index records after the index records stored in the current first storage area are written to the index area of the chunk being used. Go to the second storage area in memory.

When the index record in the first storage area is written into the second storage area in the memory, the device in which the SMR is located needs to sequentially read the index record to be written into the second storage area from the first storage area, and according to the read The tag in the obtained index record is operated accordingly. For details, please refer to the following three cases:

In the first case, when the mark in the read index record is a delete mark, the same index record in the second storage area as the key in the read index record is deleted, and the read index record is discarded;

In the second case, when the mark in the read index record is a modified mark, the same index record in the second storage area as the key in the read index record is deleted, and the read index record is written into the first In the second storage area;

In the third case, when the mark in the read index record is a write mark, the read index record is written in the second storage area.

It should be noted that, after all the index records in the first storage area are read and saved to the second storage area, the data records corresponding to the respective index records in the second storage area are valid data. At this time, the device where the SMR is located may determine the proportion of invalid data in the chunk according to an index record in each chunk that is not saved in the second storage area.

Step 309, clearing an index record in the first storage area;

After the index record in the first storage area of the memory is written to the second storage area in the memory, the index record in the first storage area needs to be cleared to reduce the memory usage.

That is to say, after the first storage area is applied, it is usually used to store the index record generated in the chunk currently being used, and if the index record recorded in the first storage area is persisted to the chunk that is being used, the first storage area Index records in a storage area are usually cleared. Thus, the first storage area being requested typically corresponds to a chunk being used.

In summary, the method for writing data into the shingled magnetic recording SMR hard disk provided by this embodiment is to write the data record containing the key, the data and the mark into the chunk being used; The data record generated by the data instruction contains keys and data. Therefore, the problem that the distributed storage system layout scheme cannot be directly transplanted to the SMR hard disk is solved, resulting in low write performance of the SMR hard disk; since the chunk is obtained according to the band virtualization The data can be randomly written between the various chunks, thus achieving the effect of improving the flexibility of writing data on the SMR hard disk.

In addition, the index record in the first storage area is stored in the second storage area corresponding to the used chunk in the memory; since the index record of the used chunk in the SMR hard disk is persisted into the memory, the solution is solved. The problem of low reading efficiency when sequentially reading the SMR hard disk is achieved; the effect of improving the reading efficiency of the SMR hard disk is achieved.

In actual use, in order to ensure the efficient use of the SMR hard disk, the device where the SMR is located needs to clean the SMR hard disk. Please refer to the following steps for details:

First, when the space usage rate of the SMR hard disk reaches the specified usage threshold, the super block with the largest global variable is found;

It should be noted that the super block in the SMR hard disk may further include a first global variable. Generally, the first global variable of the super block applied later is greater than the first global variable of the super block that was previously applied. For example, when the device where the SMR is located requests the next super block in the system metadata area, the current first global variable is incremented by 1 and used as the first global variable of the super block obtained by the application.

The usage information in the super block may include an occupation ratio of an invalid data record in each used chunk, and the invalid data record is a data record in which the index record is not recorded in the second storage area of the memory.

It should be noted that the detailed acquisition process of the proportion of the invalid data records herein can be referred to step 308 in the corresponding embodiment of FIG. 3A, and is not described here.

The device where the SMR is located detects the space usage of the current SMR hard disk. When it is detected that the space usage rate of the SMR hard disk is higher than the specified use threshold, the super block with the largest global variable is searched in the system metadata area of the SMR hard disk. Since the first global variable of the super block applied later is greater than the first global variable of the super block that was previously applied, the maximum chunk of the first global variable records each chunk in the record area of the current SMR hard disk. Properties and usage information.

For example, if the specified usage threshold is 85%, the device where the SMR is located detects that the current SMR hard disk space usage is 90%, and then the SMR hard disk space usage rate is higher than the specified usage threshold. The device where the SMR is located will find the super block with the largest global variable in the system metadata area of the SMR hard disk.

Optionally, since the garbage cleaning process may need to be frequently written and read, therefore, in order not to Affects the normal use of the device where the SMR is located. The device where the SMR resides can detect the usage of the system first. When the system is detected to be idle, the garbage cleaning process is performed.

Second, from the usage information in the found super block, read the occupation ratio of the invalid data record in each used chunk;

The device where the SMR is located can obtain the usage information of each used chunk according to the found super block, and determine the occupation ratio of the invalid data record in each used chunk according to the usage information of each used chunk.

Third, it is determined that the used chunk of the invalid data record reaches a predetermined ratio threshold, and the used chunk is garbage-cleaned.

The device in which the SMR is located determines the index record in the used chunk after determining that the occupied proportion of the invalid data record reaches the predetermined threshold, and searches for the index record in the second storage area in the memory. If the index record is found, the data record corresponding to the index record is written into the current chunk being used. If the index record is not found, the data record corresponding to the index record is garbage-cleaned.

For example, please refer to FIG. 4, which is a schematic diagram of a garbage cleaning process provided in an embodiment of the present invention. As shown in FIG. 4, after the device in which the SMR is located determines that the occupied proportion of the invalid data record in the used chunk 420 reaches a predetermined ratio threshold, the valid data record in the data area of the used chunk 420 is 0, and the valid data record is recorded. 2 and the valid data record 4 is written in the data area of the chunk 430 being used, and an index record corresponding to the valid data record 0, the valid data record 2, and the valid data record 4 is generated, and these index records are saved to the first storage of the memory. In the area. After the SMR device writes all the valid data in the used chunk420 to the chunk430 being used, the attribute of the used chunk420 is changed to unused, and then becomes an unused chunk440, and a super block 450 is applied. The attribute and usage information of each chunk in the current recording area is recorded in the super block 450.

In actual use, after the SMR hard disk is restarted, in order to ensure the normal use of the SMR hard disk, the device in which the SMR is located needs to reconstruct data in the memory. For details, please refer to FIG. 5, which illustrates that it will be provided in another embodiment of the present invention. A flow chart of a method for writing data into a stacked magnetic recording SMR hard disk, the method of writing data into a stacked magnetic recording SMR hard disk can be used in a device including an SMR hard disk, which writes data into a stacked magnetic recording The method of the SMR hard disk may include:

Step 501: Receive a write data instruction.

Step 502: Apply space to the data area in the chunk being used according to the length of the data record;

Step 503, writing a data record into the space of the application;

Step 504, forming an index record by combining the key, the tag, and the address in which the data is written into the chunk being used;

Step 505: Save the index record to a first storage area in the memory corresponding to the chunk being used;

Step 506, when the remaining space in the chunk being used is insufficient to allocate the data to be written in the write data instruction, storing the index record in the first storage area in the memory to the index area in the chunk being used;

Step 507, modifying the attribute of the chunk being used to be used, and applying an unused chunk as the used chunk for the next write data;

Step 508: Store an index record in the first storage area into a second storage area in the memory corresponding to the used chunk;

Step 509: Clear an index record in the first storage area.

The foregoing steps 501 to 509 are processes in which the device in which the SMR is located writes a data record to the SMR hard disk and persists the index record of the chunk. For details, refer to the embodiment corresponding to FIG. 3A, and the description thereof will not be repeated here.

In actual use, after the hard disk is restarted, in order to ensure the normal use of the SMR hard disk, the device where the SMR is located needs to rebuild data in the memory. For details, see steps 510 to 512 below.

Step 510, when restarting the hard disk, determining the used chunk and the chunk being used according to the attribute of the chunk;

After the SMR hard disk is restarted, the device where the SMR is located can determine the used chunk and the chunk being used according to the attributes of the chunk.

Optionally, the device where the SMR is located can obtain the attributes of each chunk from the latest super block in the system metadata area of the SMR hard disk, thereby determining the used chunk and the chunk being used.

Optionally, the super block may further include a first global variable, and the first global variable of the super block applied later is greater than the first global variable of the super block that was previously applied. For example, when the device where the SMR is located requests the next super block in the system metadata area, the current first global variable is incremented by 1 and used as the first global variable of the super block obtained by the application. After detecting the hard disk restart, the SMR device searches for the super block with the largest global variable in the system metadata area of the SMR hard disk. The first global The super-block with the largest variable records the attributes and usage information of each chunk in the record area of the current SMR hard disk. The device where the SMR is located determines the used chunk and the chunk being used according to the attributes and usage information of each chunk recorded in the superblock.

Step 511: Write an index record stored in the used chunk to a second storage area in the memory corresponding to the used chunk;

It should be noted that the index record further includes a second global variable uniquely corresponding to the data record, and the second global variable of the index record written later is greater than the second global variable of the previously recorded index record.

After acquiring the index records in the used chunks, the SMR records the index records in the used chunks into the second storage area of the memory according to the second global variable recorded in each index. When the device where the SMR is located writes the index records in sequence, the tags of each index record are extracted, and corresponding operations are performed according to the types of the tags, which can be classified into the following three cases.

In the first case, when the index record extracted by the device where the SMR is located is marked as a delete mark, the drop key is the same as the key of the extracted index record, and the second global variable is smaller than the index record of the second global variable of the extracted index record. The extracted index record is discarded until all the keys are identical to the extracted index record and the second global variable is smaller than the index record of the second global variable of the extracted index record.

In the second case, when the index of the index record extracted by the device where the SMR is located is a modified mark, the drop key is the same as the key of the extracted index record, and the second global variable is smaller than the index record of the second global variable of the extracted index record. The extracted index record is written to the second storage area until all the keys are identical to the extracted index record and the second global variable is smaller than the index record of the second global variable of the extracted index record.

In the third case, when the index of the index record extracted by the device where the SMR is located is a write mark, the extracted index record is written into the second storage area.

Step 512: Rebuild an index record corresponding to the data record according to the data record stored in the chunk being used, and write the reconstructed index record into the first storage area in the memory corresponding to the chunk being used.

The reconstructed index record may include: a key of data in the data record, a tag corresponding to the data, and an address in which the data is written in the chunk being used.

In addition, by writing the index record stored in the used chunk to the second storage area in the memory, and writing the reconstructed index record to the first storage area in the memory; since the data can be reconstructed in time after the hard disk is restarted, Therefore, the problem that the SMR hard disk cannot work normally after being restarted is solved; the effect of improving the performance of the SMR hard disk is achieved.

Please refer to FIG. 6 , which is a structural block diagram of an apparatus for writing data into a stacked magnetic recording SMR hard disk according to an embodiment of the present invention. The device for writing data into a stacked magnetic recording SMR hard disk can be used. In the device including the SMR hard disk, the device for writing data into the stacked magnetic recording SMR hard disk may include: a receiving module 601, an application module 602, a first writing module 603, and a saving module 604;

The receiving module 601 is configured to receive a write data instruction, where the write data instruction includes a key, a data and a mark, the key is used to identify the data, the mark is a write mark, a delete mark or a modify mark, and the write mark is used to indicate that the data is written into the SMR hard disk. The delete flag is used to delete the data having the key in the SMR hard disk, and the modification mark is used to indicate that the data having the same key and written to the SMR hard disk is modified by using the data;

The application module 602 is configured to apply for a space to the chunk being used according to the length of the data record, where the data record includes a key, a data, and a mark in the write data instruction;

a first writing module 603, configured to write a data record into the space of the application;

The save module 604 is configured to save the address of the key, the mark, and the data written in the chunk being used.

In summary, the device for writing data into the shingled magnetic recording SMR hard disk provided by the embodiment is written into the chunk being used by writing the data record containing the key, the data and the mark; The data record contains keys and data. Therefore, it solves the problem that the distributed storage system layout scheme cannot be directly transplanted to the SMR hard disk, resulting in low write performance of the SMR hard disk; since the chunk is obtained according to the band virtualization, Data can be randomly written between chunks, thus achieving the flexibility of writing data on the SMR hard disk.

Please refer to FIG. 7, which illustrates writing data into a tile-like magnetic record according to another embodiment of the present invention. A structural block diagram of a device for writing a SMR hard disk, the device for writing data to a stacked magnetic recording SMR hard disk can be used in a device including an SMR hard disk, and the device for writing data into the stacked magnetic recording SMR hard disk can include : receiving module 701, application module 702, first writing module 703 and saving module 704;

The receiving module 701 is configured to receive a write data instruction, where the write data instruction includes a key, a data and a mark, the key is used to identify the data, the mark is a write mark, a delete mark or a modify mark, and the write mark is used to indicate that the data is written into the SMR hard disk. The delete flag is used to delete the data having the key in the SMR hard disk, and the modification mark is used to indicate that the data having the same key and written to the SMR hard disk is modified by using the data;

The application module 702 is configured to apply for a space to the chunk being used according to the length of the data record, where the data record includes a key, a data, and a mark in the write data instruction;

a first writing module 703, configured to write the data record into the space of the application;

The save module 704 is configured to save the address of the key, the tag, and the data written in the chunk being used.

In the first possible implementation manner in this embodiment, the application module 702 is configured to apply for a space to a data area in a chunk being used according to the length of the data record.

In a second possible implementation manner in this embodiment, the saving module 704 may include: a component unit 704a and a saving unit 704b;

The component unit 704a is configured to form an index record by inserting a key, a tag, and an address in which the data is written into the chunk being used;

The saving unit 704b is configured to save the index record to the first storage area in the memory corresponding to the chunk being used; or, when the remaining space in the chunk being used is insufficient to allocate the data to be written in the write data instruction The index record in the first storage area in the memory is stored in the index area in the chunk being used.

In a third possible implementation manner in this embodiment, the apparatus may further include: a modification module 705, a storage module 706, and a clearing module 707.

The modifying module 705 is configured to modify the attribute of the chunk being used to be used, and apply an unused chunk as the used chunk for the next data to be written;

The storage module 706 is configured to store an index record in the first storage area into a second storage area in the memory corresponding to the used chunk;

The clearing module 707 is configured to clear an index record in the first storage area.

In a fourth possible implementation manner in this embodiment, the apparatus may further include: determining a module 708. The second write module 709 and the reconstruction module 710.

The determining module 708 is configured to determine, according to the attribute of the chunk, the used chunk and the chunk being used when the hard disk is restarted;

The second writing module 709 is configured to write the index record stored in the used chunk to the second storage area in the memory corresponding to the used chunk;

The reconstruction module 710 is configured to reconstruct an index record corresponding to the data record according to the data record stored in the chunk being used, and write the reconstructed index record into the first storage area in the memory corresponding to the chunk being used.

In a fifth possible implementation manner in this embodiment, the reconstructed index record includes: a key of data in the data record, a mark corresponding to the data, and an address in which the data is written in the chunk being used.

It should be noted that the device for writing data into the stacked magnetic recording SMR hard disk provided by the above embodiment is only illustrated by the division of the above functional modules when writing data into the stacked magnetic recording SMR hard disk. In the application, the above function assignment can be completed by different functional modules as needed, that is, the internal structure of the device in which the SMR is located is divided into different functional modules to complete the above description. Part or part of the function. In addition, the device for writing data to the stacked magnetic recording SMR hard disk provided by the above embodiment belongs to the same concept as the method for writing data into the stacked magnetic recording SMR hard disk. For the specific implementation process, refer to the method embodiment. I won't go into details here.

Referring to FIG. 8, which is a block diagram showing the structure of an apparatus including an SMR hard disk provided in an embodiment of the present invention, the apparatus may include a receiver 802 and a processor 804, wherein the processor 804 and the receiver 802 are coupled.

The receiver 802 is configured to receive a write data instruction, where the write data instruction includes a key, a data and a mark, the key is used to identify the data, the mark is a write mark, a delete mark or a modify mark, and the write mark is used to indicate that the data is written into the SMR hard disk. The delete flag is used to delete the data having the key in the SMR hard disk, and the modification mark is used to indicate that the data having the same key and written to the SMR hard disk is modified by using the data;

The processor 804 is configured to apply for a space to the chunk being used according to the length of the data record, where the data record includes a key, a data, and a mark in the write data instruction;

The processor 804 is further configured to write the data record into the space of the application;

The processor 804 is further configured to save the address of the key, the tag, and the data written in the chunk being used.

In summary, the device including the SMR hard disk provided in this embodiment writes data records including keys, data, and tags into the chunk being used; since the data record generated according to the write data instruction includes the key and The data, therefore, solves the problem that the distributed storage system layout solution cannot be directly transplanted to the SMR hard disk, resulting in low write performance of the SMR hard disk; since the chunk is obtained according to the band virtualization, it can be randomly written between the chunks. The data thus achieves the effect of increasing the flexibility of writing data on the SMR hard disk.

Referring to FIG. 9, which is a structural block diagram of an apparatus including an SMR hard disk provided in another embodiment of the present invention, the device may include: a receiver 902, a processor 904, and a memory 906, wherein the processor 904 respectively Receiver 902 is coupled to memory 906, which stores at least one type of computer software, and processor 904 can perform related operations in accordance with computer software stored in memory 906.

The receiver 902 is configured to receive a write data instruction, where the write data instruction includes a key, a data and a mark, the key is used to identify the data, the mark is a write mark, a delete mark or a modify mark, and the write mark is used to indicate that the data is written into the SMR hard disk. , delete mark is used to delete the data with the key in the SMR hard disk, modify the mark Used to indicate that data having the same key and having been written to the SMR hard disk is modified by using the data;

The processor 904 is configured to apply for a space to the chunk being used according to the length of the data record, where the data record includes a key, a data, and a mark in the write data instruction;

The processor 904 is further configured to write the data record into the space of the application;

The processor 904 is further configured to save the address of the key, the tag, and the data written in the chunk being used.

In the first possible implementation manner in this embodiment, the processor 904 is further configured to apply for a space to the data area in the chunk being used according to the length of the data record.

In a second possible implementation manner in this embodiment, the processor 904 is further configured to form an index record by using an address in which the key, the tag, and the data are written into the chunk being used;

The processor 904 is further configured to save the index record to the first storage area in the memory corresponding to the chunk being used; or, when the remaining space in the chunk being used is insufficient to allocate the data to be written in the write data instruction The index record in the first storage area in the memory is stored in the index area in the chunk being used.

In a third possible implementation manner in this embodiment, the processor 904 is further configured to modify the attribute of the chunk being used to be used, and apply for an unused chunk as the next data to be used. Chunk

The processor 904 is further configured to store the index record in the first storage area into a second storage area in the memory corresponding to the used chunk;

The processor 904 is further configured to clear an index record in the first storage area.

In a fourth possible implementation manner in this embodiment, the processor 904 is further configured to: when restarting the hard disk, determine the used chunk and the chunk being used according to the attribute of the chunk;

The processor 904 is further configured to write an index record stored in the used chunk to a second storage area in the memory corresponding to the used chunk;

The processor 904 is further configured to reconstruct an index record corresponding to the data record according to the data record stored in the chunk being used, and write the reconstructed index record into the first storage area in the memory corresponding to the chunk being used. .

In summary, the device including the SMR hard disk provided in this embodiment includes a key and a number. According to the marked data record written in the chunk being used; since the data record generated according to the write data instruction contains keys and data, it solves the problem that the distributed storage system layout scheme cannot be directly transplanted to the SMR hard disk, resulting in SMR The problem of low write performance of the hard disk; since the chunk is obtained according to the band virtualization, data can be randomly written between the chunks, thereby achieving the effect of improving the flexibility of writing data on the SMR hard disk.

Those of ordinary skill in the art will appreciate that the elements and algorithm steps of the various examples described in connection with the embodiments disclosed herein can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the solution. A person skilled in the art can use different methods for implementing the described functions for each particular application, but such implementation should not be considered to be beyond the scope of the present invention.

A person skilled in the art can clearly understand that for the convenience and brevity of the description, the specific working process of the system, the device and the unit described above can refer to the corresponding process in the foregoing method embodiment, and details are not described herein again.

In the several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods may be implemented in other manners. For example, the device embodiments described above are merely illustrative. For example, the division of the unit may be only a logical function division. In actual implementation, there may be another division manner, for example, multiple units or components may be combined. Or it can be integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be in an electrical, mechanical or other form.

The units described as separate components may or may not be physically separated as The components displayed by the unit may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of the embodiment.

In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

The functions may be stored in a computer readable storage medium if implemented in the form of a software functional unit and sold or used as a standalone product. Based on such understanding, the technical solution of the present invention, which is essential or contributes to the prior art, or a part of the technical solution, may be embodied in the form of a software product, which is stored in a storage medium, including The instructions are used to cause a computer device (which may be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, and the like, which can store program codes. .

The above is only a specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily think of changes or substitutions within the technical scope of the present invention. It should be covered by the scope of the present invention. Therefore, the scope of the invention should be determined by the scope of the claims.

Claims

A method for writing data into a stacked magnetic recording SMR hard disk, wherein the SMR hard disk includes a plurality of chunks obtained by virtualizing a track band, the method comprising:

Receiving a write data instruction, the write data instruction including a key, a data, and a mark, the key being used to identify the data, the mark being a write mark, a delete mark, or a modify mark, the write mark being used to indicate that the Data is written into the SMR hard disk, the deletion flag is used to indicate deletion of data having the key in the SMR hard disk, and the modification flag is used to indicate that the same key is modified with the data and has been written The data of the SMR hard disk;

Requiring a space to the chunk being used according to the length of the data record, the data record including the key, the data and the token in the write data instruction;

Writing the data record into the space of the application;

The address is saved, the tag and the data are written to the address in the chunk being used.
The method according to claim 1, wherein the chunk being used includes a data area for storing a data record, and the requesting space for the chunk being used according to the length of the data record comprises:

A space is requested from the data area in the chunk being used according to the length of the data record.
The method according to claim 2, wherein said chunk being used contains an index area for storing an index record, said saving said key, said mark and said data being written to said The address in the chunk used, including:

Forming an index record by the key, the mark, and an address in which the data is written into the chunk being used;

Saving the index record to a first storage area in memory corresponding to the chunk being used; or, when the remaining space in the chunk being used is insufficient to allocate data to be written in the write data instruction, The index record in the first storage area in the memory is stored in the index area in the chunk being used.
The method according to claim 3, wherein after said storing said index record in said first storage area in said memory to said index area in said chunk being used, Also includes:

Modify the attribute of the chunk being used to used, and apply for an unused chunk as the used chunk for the next write data;

And storing the index record in the first storage area into a second storage area in the memory corresponding to the used chunk;

Clearing the index record in the first storage area.
The method according to any one of claims 2 to 4, wherein the method further comprises:

When the hard disk is restarted, the used chunk and the chunk being used are determined according to the attributes of the chunk;

And storing an index record stored in the used chunk into a second storage area in the memory corresponding to the used chunk;

Reconstructing an index record corresponding to the data record according to the data record stored in the chunk being used, and writing the reconstructed index record into the first storage area in the memory corresponding to the chunk being used .
The method according to claim 5, wherein said index record obtained by said reconstruction comprises: a key of data in said data record, a mark corresponding to said data, and said data is written to said index The address in the chunk used.
An apparatus for writing data into a stacked magnetic recording SMR hard disk, wherein the SMR hard disk includes a plurality of chunks obtained by virtualizing a track band, the device comprising:

a receiving module, configured to receive a write data instruction, where the write data instruction includes a key, a data, and a mark, wherein the key is used to identify the data, and the mark is a write mark, a delete mark, or a modify mark, and the write mark is used by And indicating that the data is written into the SMR hard disk, the deletion mark is used to indicate deleting data having the key in the SMR hard disk, and the modification mark is used to indicate that the same key is modified by using the data and Data that has been written to the SMR hard disk;

An application module, configured to apply for a space to a chunk being used according to a length of the data record, The data record includes the key, the data, and the flag in the write data instruction;

a first writing module, configured to write the data record into the space of the application;

And a saving module, configured to save the key, the mark, and the address in which the data is written in the chunk being used.
The apparatus according to claim 7, wherein said chunk being used includes a data area for storing a data record, and said application module is configured to use said chunk according to a length of said data record The data area in the application space.
The device according to claim 8, wherein the chunk being used includes an index area for storing an index record, and the saving module comprises:

a component unit, configured to form an index record of the key, the mark, and an address in which the data is written into the chunk being used;

a saving unit, configured to save the index record to a first storage area in the memory corresponding to the chunk being used; or, the remaining space in the chunk being used is insufficient to allocate a write data instruction to be written And storing, in the data, the index record in the first storage area in the memory to the index area in the chunk being used.
The device according to claim 9, wherein the device further comprises:

a modifying module, configured to modify the attribute of the used chunk to be used, and apply an unused chunk as a used chunk for writing data next time;

a storage module, configured to store the index record in the first storage area to a second storage area in the memory corresponding to the used chunk;

And a clearing module, configured to clear the index record in the first storage area.
The device according to any one of claims 8 to 10, wherein the device further comprises:

a determining module, configured to determine, according to the attribute of the chunk, the used chunk and the chunk being used when the hard disk is restarted;

a second writing module, configured to write an index record stored in the used chunk into the a second storage area in memory corresponding to the used chunk;

a reconstruction module, configured to reconstruct an index record corresponding to the data record according to the data record stored in the chunk being used, and write the reconstructed index record into the memory to correspond to a chunk being used The first storage area.
The apparatus according to claim 11, wherein said index record obtained by said reconstructing comprises: a key of data in said data record, a mark corresponding to said data, and said data is written to said index The address in the chunk used.