CN118051642A - Data storage method and system of database system - Google Patents

Abstract

The embodiment of the application provides a data storage method and a data storage system of a database system, which relate to the technical field of databases, and comprise the following steps: the database system can be provided with at least two data catalogues, and the data in the same table space of the database system are stored in different data catalogues, so that the data in the same table can be stored in different storage paths.

Description

Data storage method and system of database system

Technical Field

The embodiments of the present application relate to the field of database technology, and in particular, to a data storage method and system for a database system.

Background technique

The database can store data. The current storage solution of the database system can support storing data from multiple tables in different storage paths. For example, the tables created inside the database system can be stored in the specified storage path, and the database system can also use external tables.

However, the current data storage solution of the database system cannot implement storage of different storage paths for the data in a single table.

Summary of the invention

In order to solve the above technical problems, the present application provides a data storage method and system for a database system. In this method, data of the same table of the database system can be stored in different data directories to achieve storage of data in the same table in different storage paths.

In a possible implementation, the present application provides a data storage method for a database system. The data directory of the database system includes: a first data directory and at least one second data directory. The method includes: based on a first write request, writing first data to a second physical file in the second data directory; wherein the first write request includes a request to write first data to a first tablespace; the second physical file is associated with the first tablespace; the first tablespace is also associated with a first physical file in the first data directory; the first physical file includes part of the data of the first tablespace.

The data directory is the file system directory of the database system.

For example, the first data directory is a default data directory of the database system, and the second data directory is an extended data directory of the database system, which is not limited here. In addition, the number of the second data directories can be one or more.

Among them, the first table space can be a table space created before the second data directory is set for the database, so the first table space can be regarded as an "old table". If the table space is created after the second data directory is set for the database, it can be regarded as a "new table".

The first table space is an "old table", the first data directory includes a first physical file associated with the first table space, and the first physical file includes partial data of the first table space, and the partial data may include at least one data page.

After the database system is configured with a second data directory, when writing data to the "old table", specifically the first write request here, the first write request is a data write request of the database, and the first write request includes a request to write first data to the first table space. Then the method can write the first data to the second physical file associated with the first table space in the second data directory (for example, a second data directory).

In this way, the method of the embodiment of the present application can store data in the same table space in different data directories when storing data in the database, so as to achieve the storage of data in the same table in different storage paths.

In a possible implementation, writing the first data to the second physical file in the second data directory based on the first write request includes: based on the first write request, when it is determined that the second data directory does not include the second physical file associated with the first tablespace, creating a second physical file associated with the first tablespace in the second data directory; and writing the first data to the second physical file.

Among them, when the second data directory does not include physical files managed by the first tablespace, it means that the second data directory has not written the data of the first tablespace. Then, the method can create an empty second physical file in the second data directory based on the first write request, and associate the second physical file with the first tablespace. In this way, the first tablespace is associated not only with the first physical file in the first data directory, but also with the second physical file in the second data directory, so that the data of the first tablespace can be written to physical files in different data directories to realize distributed storage of data of the same tablespace in different data directories.

In a possible implementation, the first physical file includes data of at least one data page of the first table space, and the second physical file includes data of at least one data page of the first table space.

Among them, any table space can be logically composed of multiple data pages. Then, when the method of the present application stores data from the same table space in different data directories, the granularity of the stored data is at the data page level. That is to say, the present application can store data of at least one data page of the first table space in the first physical file in the first data directory, and can also store data of another at least one data page of the first table space in the second physical file of the second data directory, thereby realizing cross-directory storage of data page granularity of the same table space.

In a possible implementation, after writing the first data to the second physical file in the second data directory, the method further includes: based on a first access request to the first tablespace, according to the file size of the first physical file, the file size of the second physical file, and the order of creation of the first physical file and the second physical file, mapping the first access address of the first tablespace to the second access address in a third physical file, wherein the third physical file is the first physical file or the second physical file, and wherein the first access request includes the first access address; accessing the data of the first tablespace according to the second access address.

The first access request may include but is not limited to: a data deletion request, a data query request, and a data modification request.

Since the data of the first tablespace is stored in at least two data directories, namely the first data directory and the second data directory, when a business requests access to the first tablespace, in order to accurately locate which physical file in which data directory the accessed data is located in, and the corresponding access address in the physical file, when processing the first access request to the first tablespace, the first access address corresponding to the first access request can be mapped to the second access address in the physical file to be accessed (for example, the first physical file or the second physical file) based on the respective file sizes of the first physical file and the second physical file associated with the first tablespace, and the order in which the first physical file and the second physical file are created.

Among them, the data in the first table space is stored in the first physical file and the second physical file, and after the second physical file is created, the data written to the first table space can be stored in the second physical file, and the first physical file is created before the second physical file, and a data page in the table space can correspond to a disk block of the disk corresponding to the physical file, and the data in the table space is written in sequence according to the order in which the data pages are created. Therefore, the method can determine how many of the first data pages of the first table space are stored in the first physical file and how many of the last data pages are stored in the second physical file based on the sizes of the above two physical files and the order in which the two physical files are created, so as to determine the physical file to be accessed this time and the second access address (e.g., logical address) in the physical file to be accessed this time based on the data page to be accessed by the first access request (corresponding to the first access address), so as to access the data of the first table space to the corresponding physical address in the disk corresponding to the physical file according to the second access address.

Exemplarily, the first access request may be a request triggered after the first data is written to the second physical file and without restarting the database system. Then, the file sizes of the first physical file and the second physical file and the information on the creation order of the two files may be obtained immediately after receiving the first access request.

Exemplarily, the first access request may also be after the database system is restarted after the first data is written to the second physical file. Then, after each database restart, the method of the present application can obtain the file size and file creation order of at least two physical files corresponding to the table space associated with different data directories and record them, so as to use the recorded information to accurately locate the address of the data to be accessed when receiving an access request to the table space associated with at least two data directories.

In this way, the embodiment of the present application can accurately locate the address of the accessed data when accessing data in a table space associated with at least two physical files in at least two data directories, so as to realize address mapping of the table space for distributed storage of data pages in the database system.

In a possible implementation, the method further includes: when the database system is restarted, determining at least two fourth physical files associated with the same second table space in the respective physical files in the first data directory and in the at least one second data directory; and recording first information, wherein the first information includes: the file size of each physical file in the at least two fourth physical files associated with the second table space, and the creation order between different fourth physical files in the at least two fourth physical files.

The second table space may include the first table space. The second table space refers to a table space associated with at least two physical files in different data directories.

Exemplarily, when the second tablespace includes the first tablespace, the at least two fourth physical files may include the first physical file and the second physical file.

This implementation can ensure that after each database restart, the database can support the demand for effective data access to a table space associated with multiple physical files.

In a possible implementation, after recording the first information, the method further includes: based on a second access request to the second tablespace, accessing data of the second tablespace in at least one of the fourth physical files associated with the second tablespace according to the first information recorded for the second tablespace.

The specific implementation principle of this implementation mode is similar to the implementation principle and effect of the above-mentioned implementation mode of processing the first access request to the first table space, and will not be repeated here.

In a possible implementation, the method further includes: creating a third tablespace in the database system based on a second write request, wherein the second write request includes a request for second data to be written to the third tablespace; creating a fifth physical file associated with the third tablespace in the second data directory; and writing the second data to the fifth physical file.

The third table space here may be referred to as a "new table", and the third table space may be a newly created table space requested by a business after the second data directory is set for the database system, and a request is made to write data into the newly created table space.

Exemplarily, the third tablespace may be different from the second tablespace and the first tablespace.

Optionally, in this embodiment, the third table space may be associated only with the fifth physical file in the second data directory, but not with any physical file in the first data directory, so that the data in the third table space is only stored in the second data directory.

In a possible implementation, the method further includes: determining at least two sixth physical files associated with the same fourth table space in the respective physical files in the first data directory and the at least one second data directory; connecting the data in the at least two sixth physical files in order of creation time from earliest to latest based on the creation order between different sixth physical files in the at least two sixth physical files to obtain the table data of the fourth table space; and backing up the table data of the fourth table space.

Exemplarily, the fourth tablespace here refers to a tablespace associated with at least two physical files in different data directories. For example, the fourth tablespace may include the second tablespace, or be the same as the second tablespace, which is not limited here. Similarly, the sixth physical file may be the same as the fourth physical file.

When backing up data in a table space that stores data in different data directories, this embodiment can connect the data of at least two sixth physical files associated with the table space in different data directories according to the creation order (here from early to late) between the at least two sixth physical files to obtain the table data of the table space and back it up.

For example, the data directory of the database system includes data directory 1 and data directory 2, wherein data directory 1 includes physical file a associated with table 1, data directory 2 includes physical file b associated with table 1, and data directory 2 is created after data directory 1, so that physical file b is created after physical file a. Then, when backing up the data of table 1, the method can read all the table data from physical file a and all the table data from physical file b, and connect the data start position of the table data read from physical file b to the data end position of the table data read from physical file a, so as to obtain the complete data of table 1 and back up the obtained complete data of table 1.

In this way, for the same table space storing table data in different data directories, the present application can also implement complete and accurate backup of the data in the table space.

In a possible implementation manner, the first physical file is created before the second physical file.

In a possible implementation, the multiple data directories of the database system are located in different file systems respectively.

The multiple data directories here include a first data directory and at least one second data directory.

In this way, data in the same table of the database can be stored in different file systems to meet different data storage requirements.

In a possible implementation manner, the disks corresponding to the multiple file systems corresponding to the database system are of different disk types.

The multiple file systems include file systems corresponding to the first data directory and at least one second data directory.

In this way, the database system can use different types of disks to implement data storage in the same table space, so as to utilize the characteristics of different types of disks (such as high speed, large capacity, etc.) to implement data storage.

In a possible implementation, the present application provides a data storage system for a database system. The data directory of the database system includes: a first data directory and at least one second data directory, and the data storage system includes: a first writing module, which is used to write first data to a second physical file in the second data directory based on a first writing request; wherein the first writing request includes a request to write first data to a first table space; the second physical file is associated with the first table space; the first table space is also associated with a first physical file in the first data directory; the first physical file includes part of the data of the first table space.

In a possible implementation, the first write module is specifically used to: based on the first write request, when it is determined that the second data directory does not include the second physical file associated with the first table space, create a second physical file associated with the first table space in the second data directory; and write the first data to the second physical file.

In a possible implementation, the first physical file includes data of at least one data page of the first table space, and the second physical file includes data of at least one data page of the first table space.

In a possible implementation, the data storage system further includes: a mapping module, used to map a first access address to the first table space to a second access address in a third physical file based on a first access request to the first table space, according to the file size of the first physical file, the file size of the second physical file, and the order of creation of the first physical file and the second physical file, wherein the third physical file is the first physical file or the second physical file, and wherein the first access request includes the first access address; a first access module, used to access data in the first table space according to the second access address.

In a possible implementation, the data storage system further includes: a first determination module, used to determine, when the database system is restarted, at least two fourth physical files associated with the same second table space in the respective physical files in the first data directory and in the at least one second data directory; a recording module, used to record first information, wherein the first information includes: the file size of each physical file in the at least two fourth physical files associated with the second table space, and the creation order between different fourth physical files in the at least two fourth physical files.

In a possible implementation, the data storage system also includes: a second access module, used to access the data of the second table space in at least one of the fourth physical files associated with the second table space based on a second access request to the second table space and in accordance with the first information recorded in the second table space.

In a possible embodiment, the data storage system also includes: a first creation module, used to create a third table space in the database system based on a second write request, wherein the second write request includes a request to write second data to the third table space; a second creation module, used to create a fifth physical file associated with the third table space in the second data directory; and a second write module, used to write the second data to the fifth physical file.

In a possible embodiment, the data storage system also includes: a second determination module, used to determine at least two sixth physical files associated with the same fourth table space in the respective physical files in the first data directory and the at least one second data directory; a reading module, used to read data in each of the at least two sixth physical files; a splicing module, used to connect the data read respectively from the at least two sixth physical files in order from earliest to latest in creation time based on the creation order between different sixth physical files in the at least two sixth physical files, so as to obtain the table data of the fourth table space; a backup module, used to back up the table data of the fourth table space.

In a possible implementation manner, the first physical file is created before the second physical file.

In a possible implementation, the multiple data directories of the database system are located in different file systems respectively.

In a possible implementation manner, the disks corresponding to the multiple file systems corresponding to the database system are of different disk types.

The effects of the data storage system of the database system of each of the above-mentioned embodiments are similar to the effects of the data storage method of the database system of each of the above-mentioned embodiments, and will not be described in detail here.

In a possible implementation, the present application provides a data storage device of a database system. The data storage device of the database system includes one or more interface circuits and one or more processors; the interface circuit is used to receive a signal from a memory and send the signal to the processor, the signal including a computer instruction stored in the memory; when the processor executes the computer instruction, the processor can implement the method in any of the above implementations.

The effect of the data storage device of the database system of this embodiment is similar to the effect of the data storage method of the database system of the above-mentioned embodiments, and will not be described in detail here.

In a possible implementation, the present application provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program runs on a computer or a processor, the computer or the processor executes the method in any one of the above implementations.

The effect of the computer-readable storage medium of this embodiment is similar to the effect of the data storage method of the database system of the above-mentioned embodiments, and will not be described in detail here.

In a possible implementation, the present application provides a computer program product. The computer program product includes a software program, and when the software program is executed by a computer or a processor, the method in any one of the above implementations is executed.

The effect of the computer program product of this embodiment is similar to the effect of the data storage method of the database system of the above-mentioned embodiments, and will not be described in detail here.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments of the present application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

FIG1 is a diagram showing an exemplary data storage architecture of a database management system;

FIG2a is a diagram showing an exemplary data storage architecture of a database management system;

FIG2 b is a diagram showing an exemplary data storage architecture of a database management system;

FIG2c is a schematic diagram showing an exemplary backup structure of backup data of a database;

FIG3 is a schematic diagram of the structure of a device provided in an embodiment of the present application;

FIG4 is a schematic diagram of the structure of a chip provided in an embodiment of the present application.

Detailed ways

The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

The term "and/or" in this article is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and/or B can mean: A exists alone, A and B exist at the same time, and B exists alone.

The terms "first" and "second" in the description and claims of the embodiments of the present application are used to distinguish different objects rather than to describe a specific order of objects. For example, a first target object and a second target object are used to distinguish different target objects rather than to describe a specific order of target objects.

In the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

In the description of the embodiments of the present application, unless otherwise specified, the meaning of "multiple" refers to two or more than two. For example, multiple processing units refer to two or more processing units; multiple systems refer to two or more systems.

First, define the following names:

Ordinary table: refers to the table that has been created in the default storage directory of the database system.

External Tables: Some databases support treating external common data files as a table, so the external common data files can be called external tables.

Tablespace: It is a logical division unit of a relational database. A tablespace can only belong to one database. All database objects are stored in a specified tablespace. However, it mainly stores tables, so it is called a tablespace. Each tablespace has a unique identifier (TablespaceID), and one table corresponds to one tablespace.

Data directory: used to store database data.

The following briefly introduces the data storage process of the database system in the related technology:

Related technology 1:

MariaDB can support putting read-only tables stored locally into Simple Storage Service (S3), where S3 is an object storage service whose interface is a popular standard for current object storage. This allows different data tables to be stored locally and in S3.

However, MariaDB uses external storage of data tables for data archiving or old data sharing, so it only supports external storage of read-only tables. MariaDB also only supports external storage of data with the smallest storage unit being a table, and cannot store part of the data in a table externally. After MariaDB imports a table into S3, the operations supported by the table in S3 are not completely consistent with those supported by ordinary tables in MariaDB. For example, tables in S3 do not support partition adjustment.

Related technology 2:

Oracle database can use external tables to allow users to use ordinary files outside the database system just like using ordinary tables in the database. The ordinary file can be stored in any location as long as the database can read the data in the ordinary file. The database only records the metadata information of the ordinary file, and the data in the ordinary file will not be imported into the database.

However, Oracle database can only read external tables, but not write them. External tables only support some database functions (for example, encrypted columns are not supported), while ordinary tables support a wide range of functions. Oracle database uses external ordinary files, and can only operate data in external storage paths with tables as the smallest unit.

Therefore, when supporting the storage of data from different data tables in different storage paths, the storage solutions of the database systems in the related art mainly have the problem that the database operations supported by the tables in different storage paths are inconsistent, and the smallest unit of data stored in different storage paths is the table.

In response to the above-mentioned technical problems existing in the storage solutions of database systems in the above-mentioned related technologies, the present application provides a data storage method and device for a database system, which can store different data in a table in directories of different file systems, so that the database can use different disks for data storage, and the database operations supported by tables stored in different disks can be the same as the database operations supported by ordinary tables.

FIG. 1 is a diagram showing an exemplary data storage architecture of a database management system before improvement of the present application.

As shown in FIG. 1 , a host 100 may include a database management system (DBMS) 101 , a file system 104 a , and a disk 103 a (also referred to as disk 1 ).

The file system 104a includes a data directory 102a, wherein the data directory 102a is the default data directory /mnt/data of the DBMS 101 shown in Fig. 1. The data directory 102a is used to store database data.

DBMS 101 uses a table space to store data of a table, and one table corresponds to one table space.

As shown in FIG. 1 , DBMS 101 includes five table spaces, namely, table space 1 to table space 5. Each table space corresponds to a physical file in the data directory 102a in the file system 104a. The data of a table stored in each table space is stored in the corresponding physical file in the data directory 102a in the file system 104a.

For example, the data of a table stored in each of tablespaces 1 to 5 are respectively stored in physical files 1a to 5a in the data directory 102a.

The file system 104 a is a management system for the logical addresses of the disk 1 , so the data in the physical files 1 a to 5 a are actually stored in the disk 1 .

However, the DBMS 101 only supports one data directory, and a table space in the DBMS 101 only supports the storage of table data in one physical file. On the contrary, a table space cannot support multiple physical files for the storage of table data.

To this end, the present application improves the host 100 and DBMS 101 shown in FIG. 1 , and FIG. 2 a is an illustrative diagram of the data storage architecture of the improved database management system of the present application.

As shown in FIG. 2 a , the host 100 may include a DBMS 101 , a disk 103 a (eg, a local disk 1 of the host 101 ), and a file system 104 a of the disk 103 a .

In addition, the host 100 may also mount a disk 103 b , and a file system 104 b of the disk 103 b may be created in the host 100 .

The file system 104 a may include the data directory 102 a of the DBMS 101 , and the file system 104 b may include the data directory 102 b of the DBMS 101 .

Compared with FIG. 1 , in the embodiment of the present application, the database has not only a default data directory 102a (e.g., /mnt/data as shown in FIG. 2a ), but also an extended data directory 102b (e.g., /mnt/data_2 as shown in FIG. 2a ). Of course, in other embodiments, the database may also have more than two data directories (e.g., multiple extended data directories), which is not limited in the present application, and the implementation principle is similar, which will not be described here.

In addition, in the embodiment of FIG. 2a, different data directories of the database (data directory 102a and data directory 102b) correspond to different disks. In other embodiments, different data directories of the database of the present application may also correspond to the same disk, for example, in the file system of the same disk, not only the default data directory 102a of the database is created, but also the extended data directory 102b of the database is created. The present application does not limit whether the disks corresponding to the multiple data directories of the database are the same.

In addition, the disk corresponding to the data directory of the database can be a local disk or an external disk (such as a cloud disk). This application does not limit the disk type corresponding to the data directory of the database. For example, in Figure 2a, disk 103a is the local disk of the host, and disk 103b is a cloud disk.

Continuing with reference to FIG2a, by creating a table in the database, a table space corresponding to the table can be created in DBMS101. As shown in FIG1, before creating the data directory 102b, 6 tables have been created in the database, and the corresponding table spaces are table space 1 to table space 5. After the data directory 102b is also set as the data directory of the database, DBMS101 creates a new table based on the business request, and the table corresponds to table space 6.

Exemplarily, tablespaces 1 to 5 are tablespaces created before the method of the present application creates the data directory 102 b for the DBMS 101 (for example, when the DBMS 101 has only one data directory 102 a ).

Exemplarily, after the method of the present application creates the data directory 102 b for the DBMS 101 , when a new table is created in the database again (for example, table 6 is created), a table space 6 corresponding to table 6 is created in the DBMS 101 .

It should be understood that the present application does not limit the number of table spaces within DBMS101, and does not limit the number of table spaces created after creating the data directory 102b for the database. The implementation principles of the table spaces and related solutions created after creating the data directory 102b are similar to the principles of the example of table space 6 here, and will not be repeated here.

In the following, with reference to FIG. 2a and FIG. 2b, table space 201 (also referred to as table space 1) is used as an example to describe the process of accessing table data in table space 201 of the database. The process of accessing data in table space 2 to table space 5 in FIG. 2a is similar to the example of table space 1 in FIG. 2a, and will not be described in detail later.

As described above, tablespaces 1 to 5 are tablespaces created by the method of the present application before creating data directory 102b for DBMS 101 (for example, when DBMS 101 has only one data directory 102a). As shown in FIG. 1 and FIG. 2a , each tablespace in DBMS 101 corresponds to a respective physical file in data directory 102a. For tablespaces 1 to 5, the corresponding physical files are respectively physical files 1a to physical files 5a, and physical files 1a to physical files 5a store table data of tablespaces 1 to 5, respectively.

Taking the database RDS MySQL as an example, please refer to FIG2a in comparison with FIG1. The management software can mount the disk 103b to the host 100 (or virtual machine) where the RDS MySQL instance is located, and create a file system 104b for the disk 103b, and create a data directory 102b (for example, /mnt/data_2) in the file system 104b. At this time, the data directory 102b does not include any data and is an empty directory. As the business writes data to the database, the physical files corresponding to the tablespace where the data is written are continuously added in the data directory 102b, and the data in the tablespace that can be written in the physical file is continuously added.

Then, the database management system 101 of the present application or the database administrator (Database Administrator, DBA) sets the extended data directory 102b for the RDS MySQL database through a set command.

For example, in the scenario of FIG. 1 , the DBMS has stored data of five tables (corresponding to tablespaces 1 to 5) in disk 103a. At this time, if it is determined that the storage space of disk 103a corresponding to data directory 102a is insufficient, or the performance of disk 103a is poor, DBMS 101 or the DBA may set data directory 102b as shown in FIG. 2a as the secondary data directory (i.e., the extended data directory) of the RDS MySQL database to specify a new data directory for the database.

For example, the database management system 101 or the database administrator (DBA) can use the set command to write the address of the data directory 102b (here /mnt/data_2) as a parameter into the metadata file (a configuration file of the database) of the RDS MySQL database. Of course, other files of the database can also be written, and there is no limitation here.

In a possible implementation, after the data directory 102b is set as the data directory of the RDS MySQL database, the data directory of the RDS MySQL database includes not only the data directory 102a but also the data directory 102b. After the data directory 102b is set as the data directory of the RDS MySQL database, newly generated database data or newly created tables in the RDS MySQL database can be stored in the data directory 102b, so that the database data is stored in the disk 103b.

As shown in FIG2b, table space 201 is logically composed of multiple data pages. Before physical file 1b is created in data directory 102b, table space 201 includes 100 data pages, namely data page 4001 to data page 4100. A data page in a table space is the smallest data storage unit in a database. The size of a data page corresponds to a specific number of storage bytes (such as disk blocks) on a physical storage space, and different data pages in the same table space physically correspond to multiple disk blocks of equal size. For example, before data directory 102b shown in FIG2a is set as the data directory of the RDS MySQL database, the data of 100 data pages in table space 201 are all stored in physical file 202a in the default data directory 102a, and physical file 202a is associated with table space 1. As shown in FIG2b, the data of 100 data pages of table space 201 in physical file 202a are respectively stored in 100 disk blocks of disk 103a, which are disk blocks 3001a to disk blocks 3100a. For example, the data of data page 4001 is stored in disk block 3001a via physical file 202a.

It should be understood that different disk blocks in the disk 103a shown in FIG. 2b are not limited to continuous physical addresses or discontinuous physical addresses.

For example, after the data directory 102b shown in Figure 2a is set as the data directory of the RDS MySQL database, for example, according to business needs, data needs to continue to be written in the table space 1 corresponding to the table 1. For example, at this time, the data page 4100 is full, then DBMS101 can create a data page 4101 in the table space, and according to business needs, write the data to be inserted (i.e., written) into the table space 1 into the data page 4101. The specific process will be explained below.

In FIG. 1 , table space 1 corresponds to only one physical file 1a in data directory 102a . However, in FIG. 2a after improvement in the present application, the database is further provided with a data directory 102b .

In a possible implementation, when DBMS101 receives a service request indicating a request to write data to tablespace 1, DBMS101 may create a new empty physical file 1b in data directory 102b. DBMS101 may associate the newly created physical file 1b in data directory 102b with tablespace 1 and save the association, so that tablespace 1 is associated not only with physical file 1a in data directory 102a but also with physical file 1b in data directory 102b, thereby realizing the association between the same tablespace and physical files in different data directories, so that the table data in the tablespace can be stored and accessed in physical files in different data directories.

For example, physical file 1a and physical file 1b can be associated with tablespace 1 by making the file name of the newly created physical file 1b the same as the file name of the physical file 1a that is already associated with tablespace 1. For example, the names of physical file 1 and physical file 2 are both the name of table 1 corresponding to tablespace 1 (for example, table XX).

For another example, the present application does not restrict the file name of the newly created physical file 1b, but after the new physical file 1b is created, DBMS101 can associate the file name of the physical file 1b with the name of the tablespace 1, and store the association information as metadata of the physical file 1b in the database. Before creating the data directory 102b, the pre-created physical file 1a has been associated with the tablespace 1, so that the effect of associating both the physical file 1a and the physical file 1b with the tablespace 1 can also be achieved.

In addition, when realizing the association between a physical file and a table space, it is not limited to using a file name or a table name to realize the association, and other information that can identify a physical file or a table space can also be used to realize the establishment of the association relationship.

In this way, after setting the data directory 102b as the data directory of DBMS101, after receiving a data write request for any tablespace from tablespace 1 to tablespace 5, DBMS can create a physical file corresponding to the corresponding tablespace in the data directory 102b, where the physical files are respectively physical files 1b to physical files 5b corresponding to tablespace 1 to tablespace 5 as shown in FIG2a, and DBMS101 can establish an association relationship between the newly created physical files in the data directory 102b and the corresponding tablespace, so that the same tablespace can be associated with the corresponding physical files in the data directory 102a and the corresponding physical files in the data directory 102b. Before the data corresponding to the data write request is written into the corresponding tablespace, the physical files 1b to physical files 5b are all empty files, and the file size continues to grow as the data is written.

Continuing to take table space 1 as an example, as shown in FIG2b, after DBMS101 associates the physical file 1a in the data directory 102a and the newly created physical file 1b in the data directory 102b with table space 1 and saves the association, DBMS101 can respond to a business request to write data to table space 1 and write the data to be inserted into the table space 1 into data page 4101, wherein data page 4101 is only a logical storage address, DBMS101 can write the data to be written into data page 4101 into physical file 1b, and the data inserted into data page 4101 of the table space 1 is actually stored in disk block 3001b of disk 103b. In this way, in order to store the data of a table in different data directories, the present application uses the table space as a unit of one or more data pages, and one or more data pages can correspond to a physical file in a data directory, thereby splitting the physical file corresponding to the table space into multiple physical files, and these physical files can be in different data directories respectively. Compared to the related art which can only store data in different storage locations on a table basis, the present application can use one or more data pages within a single table as units to implement data storage in different storage locations for different data pages of the same table, and can specify the storage location of data at the granularity of data pages.

It should be understood that in Figure 2b, only the data to be inserted into tablespace 1 is written into data page 4101 as an example. In the scenario of inserting database data into tablespace 1, the number of data pages corresponding to physical file 1b is not limited to one or more. For example, if the amount of inserted data is large, the data pages corresponding to the amount of data written to physical file 1b at a single time can be multiple.

In addition, in the scenario of writing data to physical file 1b, the data page corresponding to the data written to physical file 1b in table space 1 can also be in a data page that currently exists but has not been filled with data. For example, the size of data page 4002 is 10k, but only 8k of data is written to data page 4002. Then, the data to be inserted into table space 1 this time can also have 2k of data written to data page 4002. In this way, data page 4002 can correspond to 8k of data written to physical file 1a, and 2k of data written to physical file 1b.

For example, FIG. 2b also shows disk blocks 3002 to 3100 of disk 103b, etc. Then, after data directory 102b is set as the data directory of the database, DBMS 101 can insert data into tablespaces 2 to 5 shown in FIG. 2a for the first time according to a business request, and can also create physical files 2b to 5b associated with the corresponding tablespaces in data directory 102b, and write the data to be inserted into tablespaces 2 to 5 into the physical files 2b to 5b according to the business request, and the data inserted into tablespaces 2 to 5 are actually stored in the disk blocks of disk 103b, such as disk blocks 3002 to 3100, and the number of disk blocks to which data is written is not limited, and can be determined according to business requirements. And here, there is no restriction on which disk block the data inserted into tablespaces 2 to 5 is written. After data directory 102b is set as the data directory of the database, the principle of implementing the process of DBMS101 inserting data into any table space from table space 2 to table space 5 according to the business request is similar to the principle of implementing the process of inserting data into table space 1 described above with respect to FIG. 2b, and will not be repeated here.

Thus, as shown in FIG. 2a , table data of table space 1 in DBMS 101 can be stored in physical file 1a in data directory 102a and physical file 1b in data directory 102b. Similarly, table data of table space 2 in DBMS 101 can be stored in physical file 2a in data directory 102a and physical file 2b in data directory 102b; table data of table space 3 in DBMS 101 can be stored in physical file 3a in data directory 102a and physical file 3b in data directory 102b; table data of table space 4 in DBMS 101 can be stored in physical file 4a in data directory 102a and physical file 4b in data directory 102b; table data of table space 5 in DBMS 101 can be stored in physical file 5a in data directory 102a and physical file 5b in data directory 102b. Thus, one table space can correspond to at least two physical files, and the at least two physical files can be associated with the same corresponding table space, so that the at least two physical files can form a logical table space. In this way, data of different data pages of the same table space can be stored in different file systems, so as to realize the storage of data of the same table space in different types of disks.

It should be understood that which physical files of the tablespaces in DBMS101 are specifically included in the data directory 102b depends on the access conditions of the tablespaces in the database. If, after setting the data directory 102b for the database, the business does not request to write data to the tablespaces created in DBMS101, the data directory 102b will not create physical files associated with the tablespace. For example, the data directory 102b will not include physical files with the same names as the physical files of the corresponding tablespaces in the data directory 102a.

The present application is different from the related art 1 in that any type of table (not limited to read-only tables) of the present application can be stored in an extended data directory outside the default data directory 102a.

In addition, in the present application, the number of extended data directories newly set for the database can be not only one as exemplified in the above embodiment, but also two or more. When the number of extended data directories is two or more, different physical files associated with the same table space in different data directories of the database (including the default data directory and the extended data directory) also have a file creation order, so when accessing the data of the table space, the table data in the corresponding physical file can be accessed according to the creation order between different physical files of the table space. The specific principle is similar to the principle of one extended data directory, which will not be repeated here.

It should be understood that when adding a data directory to a database, it is not limited to adding a data directory only when the disk space corresponding to the existing data directory of the database is full or the performance is poor. Also, when the number of extended data directories is 2 or more (for example, the extended data directory includes the data directory 102b shown in Figure 2a, and the newly added data directory 102c not shown), the present application is not limited to creating a physical file associated with the table space in the data directory 102c and writing the data to be written to the table space to the corresponding physical file in the data directory 102c only when the data of the same table space is filled on the disk 103b corresponding to the data directory 102b. In other words, the present application does not limit the creation order and data writing order of different physical files corresponding to the same table space between different data directories.

In some embodiments, the application scenario of the present application can be a database system in a cloud environment, and the host 100 shown in Figure 2a can also be a server in a computing cluster. In some embodiments, the host 100 shown in Figure 2a can also be a virtual machine in a computing cluster, and the present application does not impose any restrictions on this.

Exemplarily, in a cloud environment, the host machine (or virtual machine) of the database can often conveniently use various types of storage disks, such as local disks, Serial Advanced Technology Attachment (SATA) cloud disks, Solid State Disk (SSD) cloud disks, etc. Different types of storage disks often have different prices, speeds, storage capacities, etc. In a cloud environment, the capacity of local disks is often limited, while the upper limit of the capacity of cloud disks is often very large. In order to enable the database system to utilize both the high speed of local disks and the large capacity of cloud disks, in some scenarios, the present application can mount the cloud disk on the database host machine to use the data directory of the cloud disk as an extended data directory of the database. In this way, in this application scenario, the present application can store part of the data of the table space in the default data directory 102a, so as to use the disk 103a (such as the local disk 1) for high-speed access to data, and store another part of the data of the table space in the extended data directory 102b, so as to use the disk 103b (such as the cloud disk) for mass storage of data, so as to achieve high-speed access and mass storage of database data.

In a possible implementation, after setting the data directory 102b for the database, after the physical files 1b to 5b corresponding to tablespaces 1 to 5 as shown in FIG. 2a have been created, and the physical files 1b to 5b have been associated with tablespaces 1 to 5, respectively, when DBMS 101 receives a data access request (which may include any database operation such as data insertion, deletion, modification, and query) for any tablespace 1 to 5 (taking tablespace 1 as an example) again, DBMS 101 may determine the physical file to be accessed this time in the physical file 1a and/or the physical file 1b associated with the tablespace 1 based on the data access request, the file size of the physical file 1a, the file size of the physical file 1b, and the order in which the physical files 1a and 1b are created, and based on the data access request, map the data pages to be accessed in tablespace 1 to the disk blocks corresponding to the corresponding physical files for data access, so as to achieve database access.

Since DMBS101 responds to the data access request of the business with database objects (such as table spaces), and this application only changes the storage location of the data page corresponding to the table space, so that the physical files corresponding to the table space can be 2 or more, in this way, the business-oriented table space has not changed, so that the business cannot perceive the change of the physical file. The user who triggers the business does not need to create a new database table or modify the database table to use the extended data directory 102b. And for tables with data pages located in multiple data directories (such as tables corresponding to any table space from table space 1 to table space 5), all database functions of ordinary tables can be supported.

In a possible implementation, referring to FIG. 2a , after the data directory 102b is set for the database, DBMS 101 receives a data access request, for example, the data access request includes creating a new table (such as table 6) and writing data to the table, then DBMS 101 can create a new table space 6 in response to the data access request, at which point the table space 6 does not yet include any data pages, and as data is written to the table space 6, the number of data pages in the table space 6 gradually increases. DBMS 101 can also create a physical file 6b associated with the table space 6 in the data directory 102b and save the association.

Since tablespace 6 is a tablespace for a newly created table after data directory 102b is created, tablespace 6 can be understood as a new table, and therefore, tablespace 6 currently has no corresponding physical file in data directory 102a. On the contrary, tablespaces 1 to 5 are tablespaces that have been created before data directory 102b is set for the database, and tablespaces 1 to 5 are all old tables. Therefore, before data directory 102b is set for the database, physical files of the above five tablespaces have been created in data directory 102a.

Continuing to describe the data access request of DBMS101 to the above table space 6, after DBMS101 creates a physical file 6b associated with the table space 6 in the data directory 102b and saves the association relationship, DBMS101 can write the data to be inserted into the table space 6 into the physical file 6b according to the data access request, so as to write the data of the table space 6 into the disk 103b. The data storage principle and process are similar to the process shown in FIG. 2b, and the data is also written according to the data page and the disk block, which will not be repeated here. In this way, the table data of table space 6 can be stored in the disk 103b, and the table data of table space 1 to table space 5 can be stored in the disk 103a, and optionally further stored in the disk 103b. This achieves the effect of storing the table data of different table spaces in different storage locations.

In a possible implementation, as the database data stored in the disk 103a as shown in FIG. 2a is deleted, the free storage space in the disk 103a is greater than a preset threshold value, which can be flexibly configured according to the requirements, and will not be described in detail here. Then, after receiving a data insertion request for a new table (e.g., table space 6), DBMS101 can create an empty physical file 6a in the data directory 102a, and associate the physical file 6a with the table space 6. For example, DBMS101 can save the association relationship between the physical file 6a and the table space 6, so that the table space 6 is not only associated with the physical file 6b in the data directory 102b, but also associated with the physical file 6a in the data directory 102a, so that the same table space is associated with the physical files in different data directories. Then, DBMS101 can write the data into at least one data page of the table space 6 according to the data insertion request, and write the data of the at least one data page into the physical file 6a, so that the database data inserted this time is stored in the corresponding disk block in the disk 103a. In this way, for a new table, different data pages of the same table space can also be stored in different storage locations.

Thus, in the embodiment of the present application, the data directory 102a and the data directory 102b may respectively store different physical files of the same table space of the DBMS 101 and/or physical files of different table spaces.

The above embodiment mainly describes the process of writing data to an old table, creating a new table (e.g., tablespace 6) and writing data to the new table after the data directory 102b is set as the data directory of the database. The process involves creating a physical file corresponding to the tablespace of the corresponding accessed table in the data directory 102b, and saving the association relationship between the physical file and the tablespace. For example, the association relationship between the physical file and the tablespace can be saved in a configuration file (e.g., a metadata file) of the database.

Then, in order to ensure that the database can support the need for effective data access to tablespaces associated with multiple physical files after each database restart, in a possible implementation, each time the database is restarted, DBMS101 can read the address of the data directory of the database by reading the configuration file of the database, which is the address of the data directory 102a (for example, /mnt/data shown in FIG2a ) and the address of the data directory 102b (for example, /mnt/data_2 shown in FIG2a ); then, DBMS101 scans the information (for example, name, size, etc.) of the physical files in the data directory 102a according to the address of the data directory 102a, and DBMS101 scans the physical files in the data directory 102b according to the address of the data directory 102a. The address of the physical file in the data directory 102b is scanned (for example, name, size, etc.). If the same tablespace is associated with physical files in both the data directory 102a and the data directory 102b, DBMS101 can scan two physical files with the same name from the two data directories. For example, DBMS101 scans physical file 202a and physical file 202b both named "tablespace 1", thereby determining that the two physical files are associated with tablespace 1, which means that the data of a table corresponding to tablespace 1 is stored in physical file 202a and physical file 202b.

Alternatively, in a possible implementation, when DBMS101 associates physical file 202b with tablespace 1, the association is not by creating a physical file 202b with the same name as physical file 202a, but by associating the name of physical file 202b with the name of tablespace 1 and saving the association in the metadata of physical file 202b. In this implementation, after the database is restarted, DBMS101 can determine which tablespace each physical file is specifically associated with by reading the metadata information of each physical file in data directory 102b; and DBMS101 can also determine which tablespace each physical file in data directory 102a is associated with by reading the names of physical files in data directory 102a in the above implementation (of course, other implementations are also possible and are not limited here). In this way, DBMS101 can also determine which physical files in data directory 102a and data directory 102b are associated with the same tablespace and which tablespace they are associated with when the database is restarted.

After the database is restarted, DBMS 101 may determine the physical files associated with the same table space in data directories 102a and 102b. Furthermore, DBMS 101 may record the sizes of at least two physical files associated with the same table space and the creation order of at least two physical files.

Taking table space 1 as an example, the same is true for other table spaces, which will not be described here. Each time after the database is restarted, DBMS 101 can record the file size of physical file 202a and the file size of physical file 202b associated with table space 1, and record the information that physical file 202a was created earlier than physical file 202b. In this way, after the database is restarted this time, if the business needs to access table space 1, DBMS 101 can use the recorded file size information and the order information of the file creation to map the data page in table space 1 to be accessed this time to the disk block of the corresponding physical file, so as to realize the mapping of logical address to physical address, so as to realize the accurate access to the data of table spaces storing data in different data directories.

Take FIG. 2b as an example, for example, a data page size is 10k, physical file 202a stores 100 data pages, and its file size is 1000k, and physical file 202b stores one data page, and its file size is 10k. Physical file 202a is created earlier than physical file 202b. It should be understood that when DBMS101 writes data to table space 1, it writes in sequence, for example, after data page 4001 is full, data page 4002 will be created, and after data page 4002 is full, data page 4003 will be created, and so on, so that data pages 4004 to 4101 are filled in sequence, which will not be repeated here.

Since the size of each data page is known, the sizes of physical file 202a and physical file 202b are known, and the creation order of physical file 202a and physical file 202b is known, DBMS101 can write data to the physical files in this creation order. Therefore, DBMS101 can determine that physical file 202a stores the data of the first 100 data pages of table space 1, and physical file 202b stores the data of the 101st data page of table space 1. The file system 104a where the physical file 202a is located is the file system of the disk 103a. The file system 104a and the disk 103a may have a mapping relationship between addresses (which may be implemented by any existing technology without limitation), so that the 100 data pages stored in the physical file 202a may be mapped to the corresponding 100 disk blocks of the disk 103a; similarly, the file system 104b and the disk 103b may have a mapping relationship between addresses (which may be implemented by any existing technology without limitation), so that one data page stored in the physical file 202b may be mapped to one corresponding disk block of the disk 103b. In this way, the 100 disk blocks stored in the physical file 202a are automatically mapped to the 1st to 100th data pages of the table space 1, and one disk block stored in the physical file 202b is automatically mapped to the 101st data page of the table space 1.

In this way, each time after the database is restarted, DBMS101 can map the logical data pages in table space 1 accessed by this data access request to the disk blocks of the disk based on the data access request for table space 1, the recorded file sizes of physical files 202a and physical files 202b associated with table space 1, and the creation order of the two files, so as to realize the mapping of logical addresses to physical addresses, so as to realize the access to data in the table space associated with at least two physical files (which may include query, deletion, and modification).

For example, if the data access request is to access the second data page of table space 1, data access is performed at the corresponding physical address mapped in disk 103a corresponding to physical file 202a; for another example, if the data access request is to access the 101st data page of table space 1, data access is performed at the corresponding physical address mapped in disk 103b corresponding to physical file 202b.

If the data access request is to insert data, it can be inserted into the physical file 202b by default without relying on the recorded file size and file creation order information, or it can be inserted into the physical file 202a when the disk 103a is idle. There is no restriction here.

In the embodiment of the present application, after each database restart, the database can record the file sizes of at least two physical files associated with the same table space, as well as the information of the order in which the files are created. In this way, after the database is restarted, when the service requests to delete, modify, query, and perform other operations on the above table space of the database, the database can map the logical address of the accessed data page in the table space corresponding to at least two physical files with the physical address of the physical file 202a or the physical file 201b based on the above recorded information, thereby achieving accurate data access to the table spaces respectively associated with at least two physical files in different data directories.

The above process takes tablespace 1 as an example to illustrate the data access process after the database is restarted. When the accessed table is a new table, such as tablespace 6 shown in Figure 2a, if the data directory 102a includes physical file 6a, then the physical file 6b here is created earlier than the physical file 6a. The other processes are the same as the data access process described in the above tablespace 1, and will not be repeated here.

In various embodiments of the present application, multiple physical files in the same data directory correspond to different table spaces. In other words, data in a table space can be stored in multiple data directories, but when data in a table space is stored in any data directory, the data in the table space can only be stored in one physical file in the corresponding data directory.

In a possible implementation, when backing up the table data of the above-mentioned database, the present application may improve the backup tool. The backup tool of the present application may obtain the address information of the data directory of the database by reading the configuration file of the database, such as the address of the data directory 102a and the address of the data directory 102b shown in FIG. 2a; then, the backup tool may use the addresses of the above-mentioned two data directories to scan the information (name, size, etc.) of the physical files from the two data directories, so as to determine which table spaces are respectively associated with physical files in the above-mentioned two data directories, and the physical files associated with the table spaces. , and the order in which the physical files are created; then, the backup tool can read data from the physical files associated with the same tablespace in the two data directories respectively. Taking tablespace 1 as an example, the backup tool can read data from physical file 202a in data directory 102a and read data from physical file 202b in data directory 102b; and the backup tool connects the data read from physical file 202b to the end of the data read from physical file 202a according to the order in which the physical files are created (wherein physical file 202a is created earlier than physical file 202b) to obtain the backup data of tablespace 1.

Continuing to refer to FIG. 2b, for example, the data in physical file 202a is stored in order from first to last, i.e., the data stored in disk block 3001a, disk block 3002a, ..., disk block 3100a. Then, as shown in FIG. 2c, the backup data of table space 1 is stored in order from first to last, i.e., the data stored in disk block 3001a, disk block 3002a, ..., disk block 3100a, disk block 3001b, wherein the data stored in disk block 3001b is connected to the end of the data in disk block 3100a.

Exemplarily, the backup tool may be an improvement on an existing database backup tool, such as adding a tablespace file aggregation component to implement data backup of a tablespace associated with multiple physical files. The backup tool may use any existing database data backup method for data in a tablespace associated with only one physical file, and is not limited here. For example, the backup tool may be software.

In this way, the database management system of the present application can support full data backup for tables stored in multiple data directories, and the database object operated by the data backup operation is still the table space, which does not involve the storage location of specific data pages and can be transparent to the full backup operation.

In the embodiment of the present application, by adding a physical file for storing the data of the table space, different data in a table corresponding to the table space can be stored in different physical files respectively, so that only the storage location of the underlying table data of the table space is changed, and multiple physical files located at different storage locations are associated with the same table space, so that the multiple physical files (such as physical file 1a and physical file 1b) constitute a logical table space object of table space 1, thereby realizing the splitting of the data storage location of the same table space. Since the database is based on database objects (such as table spaces, not physical files), the various functions of the database cannot detect the change of the storage location of the table data of the table space, and the business will not perceive the change of the storage location of the data in the table. In this way, when the business accesses the database data, the method of the present application can realize the hot switching access to the data of the table space corresponding to different physical files, and the business does not perceive the hot switching. In this way, the database operation functions supported by the table split by the storage location (such as the table corresponding to table space 1 to table space 6 shown in Figure 2a) can be consistent with the database operation functions supported by the ordinary table in the database that has not been split by the storage location (such as the table corresponding to table space 1 to table space 6 in Figure 1).

FIG. 2a is a schematic diagram showing an exemplary system framework structure. It should be understood that the architecture diagram shown in FIG. 2a is only an example, and the system of the present application may have more or fewer components than those shown in the figure, may combine two or more components, or may have different component configurations. The various components shown in FIG. 2a may be implemented in hardware, software, or a combination of hardware and software including one or more signal processing and/or application specific integrated circuits.

In a possible implementation, the present application also provides a data storage system for a database system, wherein the data directory of the database system includes: a first data directory and at least one second data directory, and the data storage system includes: a first write module, used to write first data to a second physical file in the second data directory based on a first write request; wherein the first write request includes a request to write first data to a first table space; the second physical file is associated with the first table space; the first table space is also associated with a first physical file in the first data directory; the first physical file includes part of the data of the first table space.

In a possible implementation, the first write module is specifically used to: based on the first write request, when it is determined that the second data directory does not include the second physical file associated with the first table space, create a second physical file associated with the first table space in the second data directory; and write the first data to the second physical file.

In a possible implementation, the first physical file includes data of at least one data page of the first table space, and the second physical file includes data of at least one data page of the first table space.

In a possible implementation, the data storage system further includes: a mapping module, used to map a first access address to the first table space to a second access address in a third physical file based on a first access request to the first table space, according to the file size of the first physical file, the file size of the second physical file, and the order of creation of the first physical file and the second physical file, wherein the third physical file is the first physical file or the second physical file, and wherein the first access request includes the first access address; a first access module, used to access data in the first table space according to the second access address.

In a possible implementation, the data storage system further includes: a first determination module, used to determine, when the database system is restarted, at least two fourth physical files associated with the same second table space in the respective physical files in the first data directory and in the at least one second data directory; a recording module, used to record first information, wherein the first information includes: the file size of each physical file in the at least two fourth physical files associated with the second table space, and the creation order between different fourth physical files in the at least two fourth physical files.

In a possible implementation, the data storage system further includes: a second access module for accessing data of the second table space in at least one of the fourth physical files associated with the second table space based on a second access request to the second table space and in accordance with the first information recorded in the second table space.

In a possible embodiment, the data storage system also includes: a first creation module, used to create a third table space in the database system based on a second write request, wherein the second write request includes a request to write second data to the third table space; a second creation module, used to create a fifth physical file associated with the third table space in the second data directory; and a second write module, used to write the second data to the fifth physical file.

In a possible embodiment, the data storage system also includes: a second determination module, used to determine at least two sixth physical files associated with the same fourth table space in the respective physical files in the first data directory and the at least one second data directory; a reading module, used to read data in each of the at least two sixth physical files; a splicing module, used to connect the data read respectively from the at least two sixth physical files in order from earliest to latest in creation time based on the creation order between different sixth physical files in the at least two sixth physical files, so as to obtain the table data of the fourth table space; a backup module, used to back up the table data of the fourth table space.

In a possible implementation manner, the first physical file is created before the second physical file.

In a possible implementation, the multiple data directories of the database system are located in different file systems respectively.

In a possible implementation manner, the disks corresponding to the multiple file systems corresponding to the database system are of different disk types.

The following describes a device provided in an embodiment of the present application. As shown in FIG3:

Fig. 3 is a schematic diagram of the structure of a data storage device of a database system provided in an embodiment of the present application. As shown in Fig. 3 , the device 500 may include: a processor 501 , a transceiver 505 , and optionally a memory 502 .

The transceiver 505 may be referred to as a transceiver unit, a transceiver, or a transceiver circuit, etc., and is used to implement a transceiver function. The transceiver 505 may include a receiver and a transmitter, the receiver may be referred to as a receiver or a receiving circuit, etc., and is used to implement a receiving function; the transmitter may be referred to as a transmitter or a transmitting circuit, etc., and is used to implement a transmitting function.

The memory 502 may store a computer program or software code or instruction 504, which may also be referred to as firmware. The processor 501 may implement the data storage method of the database system provided in each embodiment of the present application by running the computer program or software code or instruction 503 therein, or by calling the computer program or software code or instruction 504 stored in the memory 502. The processor 501 may be a central processing unit (CPU), and the memory 502 may be, for example, a read-only memory (ROM) or a random access memory (RAM).

The processor 501 and transceiver 505 described in the present application can be implemented on an integrated circuit (IC), an analog IC, a radio frequency integrated circuit RFIC, a mixed signal IC, an application specific integrated circuit (ASIC), a printed circuit board (PCB), an electronic device, etc.

The above-mentioned device 500 may further include an antenna 506. The modules included in the device 500 are only for illustration and are not limited in this application.

For example, the structure of the data storage device of the database system may not be limited by FIG. 3. The data storage device of the database system may be an independent device or may be a part of a larger device. For example, the implementation form of the data storage device of the database system may be:

(1) An independent integrated circuit IC, or chip, or chip system or subsystem; (2) A collection of one or more ICs, optionally, the IC collection may also include a storage component for storing data and instructions; (3) A module that can be embedded in other devices; (4) In-vehicle equipment, etc.; (5) Others, etc.

In the case where the data storage device of the database system is implemented in the form of a chip or a chip system, reference may be made to the schematic diagram of the chip structure shown in FIG4 . The chip shown in FIG4 includes a processor 601 and an interface 602 . The number of processors 601 may be one or more, and the number of interfaces 602 may be multiple. Optionally, the chip or chip system may include a memory 603 .

Among them, all relevant contents of each step involved in the above method embodiment can be referred to the functional description of the corresponding functional module, and will not be repeated here.

Based on the same technical concept, an embodiment of the present application also provides a computer-readable storage medium, which stores a computer program. The computer program includes at least one code segment, and the at least one code segment can be executed by a computer to control the computer to implement the above method embodiment.

Based on the same technical concept, the embodiment of the present application also provides a computer program, which, when executed, is used to implement the above method embodiment.

The program may be stored in whole or in part on a storage medium packaged together with the processor, or may be stored in whole or in part on a memory not packaged together with the processor.

Based on the same technical concept, the embodiment of the present application further provides a chip, including a processor. The processor can implement the above method embodiment.

The steps of the method or algorithm described in conjunction with the disclosure of the embodiments of the present application can be implemented in a hardware manner, or can be implemented by a processor executing a software instruction. The software instruction can be composed of corresponding software modules, and the software module can be stored in a random access memory (Random Access Memory, RAM), a flash memory, a read-only memory (Read Only Memory, ROM), an erasable programmable read-only memory (Erasable Programmable ROM, EPROM), an electrically erasable programmable read-only memory (Electrically EPROM, EEPROM), a register, a hard disk, a mobile hard disk, a read-only compact disk (CD-ROM) or any other form of storage medium known in the art. An exemplary storage medium is coupled to a processor so that the processor can read information from the storage medium and can write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can be located in an ASIC.

Those skilled in the art should be aware that in one or more of the above examples, the functions described in the embodiments of the present application can be implemented with hardware, software, firmware, or any combination thereof. When implemented using software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein the communication media include any media that facilitates the transmission of a computer program from one place to another. The storage medium can be any available medium that a general or special-purpose computer can access.

The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms without departing from the purpose of the present application and the scope of protection of the claims, all of which are within the protection of the present application.

Claims (15)

1. A data storage method for a database system, characterized in that the data directory of the database system includes: a first data directory and at least one second data directory, and the method includes: Based on the first write request, write the first data to a second physical file in the second data directory; Wherein, the first write request includes a request to write first data to the first table space; The second physical file is associated with the first table space; The first table space is also associated with a first physical file in the first data directory; The first physical file includes part of the data of the first table space. 2. The method according to claim 1, wherein writing the first data to the second physical file in the second data directory based on the first write request comprises: Based on the first write request, when it is determined that the second data directory does not include a second physical file associated with the first table space, creating a second physical file associated with the first table space in the second data directory; The first data is written to the second physical file. 3. The method according to claim 1 or 2, characterized in that the first physical file includes data of at least one data page of the first table space, and the second physical file includes data of at least one data page of the first table space. 4. The method according to any one of claims 1 to 3, characterized in that after writing the first data to the second physical file in the second data directory, the method further comprises: Based on a first access request to the first tablespace, according to a file size of the first physical file, a file size of the second physical file, and a creation order of the first physical file and the second physical file, mapping a first access address to the first tablespace to a second access address in a third physical file, wherein the third physical file is the first physical file or the second physical file, and wherein the first access request includes the first access address; Access the data in the first table space according to the second access address. 5. The method according to any one of claims 1 to 4, characterized in that the method further comprises: When the database system is restarted, determining at least two fourth physical files associated with the same second table space in the respective physical files in the first data directory and the at least one second data directory; Record first information, wherein the first information includes: the file size of each physical file in the at least two fourth physical files associated with the second table space, and the creation order between different fourth physical files in the at least two fourth physical files. 6. The method according to claim 5, characterized in that after recording the first information, the method further comprises: Based on a second access request to the second table space, data of the second table space is accessed in at least one of the fourth physical files associated with the second table space according to the first information recorded for the second table space. 7. The method according to any one of claims 1 to 6, characterized in that the method further comprises: Based on the second write request, creating a third table space in the database system, wherein the second write request includes second data requested to be written to the third table space; Creating a fifth physical file associated with the third tablespace in the second data directory; The second data is written to the fifth physical file. 8. The method according to any one of claims 1 to 7, characterized in that the method further comprises: Determine at least two sixth physical files associated with the same fourth table space in the respective physical files in the first data directory and the at least one second data directory; Reading data in each sixth physical file in the at least two sixth physical files; Based on the creation sequence between different sixth physical files in the at least two sixth physical files, data read from the at least two sixth physical files are connected in order of creation time from earliest to latest, so as to obtain table data of the fourth table space; Back up the table data in the fourth table space. 9. The method according to any one of claims 1 to 8, characterized in that the first physical file is created before the second physical file. 10. The method according to any one of claims 1 to 9, characterized in that the multiple data directories of the database system are located in different file systems respectively. 11. The method according to claim 9, characterized in that the disks corresponding to the multiple file systems corresponding to the database system are of different disk types. 12. A data storage system of a database system, characterized in that the data directory of the database system comprises: a first data directory and at least one second data directory, and the data storage system comprises: A first writing module, configured to write the first data into a second physical file in the second data directory based on a first writing request; Wherein, the first write request includes a request to write first data to the first table space; The second physical file is associated with the first table space; The first table space is also associated with a first physical file in the first data directory; The first physical file includes part of the data of the first table space. 13. A computer-readable storage medium, characterized in that it comprises a computer program, and when the computer program runs on a computer or a processor, the computer or the processor executes the method according to any one of claims 1 to 11. 14. A data storage device for a database system, characterized in that it comprises one or more interface circuits and one or more processors; the interface circuit is used to receive a signal from a memory and send the signal to the processor, the signal comprising a computer instruction stored in the memory; when the processor executes the computer instruction, the processor is used to execute the method as described in any one of claims 1 to 11. 15. A computer program product, characterized in that the computer program product comprises a software program, and when the software program is executed by a computer or a processor, the steps of the method according to any one of claims 1 to 11 are executed.