CN118467243A - A database backup method, server and shared memory device - Google Patents

Abstract

The embodiment of the present application provides a database backup method, a server and a shared memory device, which relates to the technical field of servers and can ensure data synchronization between the database in the main computing node and the standby computing node. The method is applied to a shared memory device, and the shared memory device communicates with the main computing node and the standby computing node respectively based on the computing fast link CXL protocol; the method includes: receiving a target log sent by the main computing node; the target log is used to record the data changes in the database in the main computing node; receiving a data read request from the standby computing node, the data read request is used to request to read the target log; in response to the data read request, sending the target log to the standby computing node.

Description

A database backup method, server and shared memory device

Technical Field

The embodiments of the present application relate to the field of server technology, and in particular to a database backup method, a server, and a shared memory device.

Background Art

For HANA database, in order to ensure data security, the database system is generally set up to ensure high availability and disaster tolerance. The database system includes a primary computing node and a standby computing node, and a HANA database is deployed on each node. The two nodes perform system replication during the initial setup. During the subsequent operation of the primary computing node, if the data changes, the primary computing node can record the data changes through logs and send the logs to the standby computing node, so that the standby computing node can update its own data according to the logs to achieve data synchronization between the two nodes.

In order to avoid the need to consume network resources to send logs to the standby computing node every time the main computing node generates logs, the current main computing node is configured with a transmission buffer to store logs to be sent to the standby computing node. In certain circumstances, the logs in the transmission buffer are uniformly sent to the data standby node to save network resources. In this case, if the main computing node fails, there may be logs in the transmission buffer that have not been sent to the standby computing node, resulting in the data of the standby computing node being unable to keep pace with the data of the main computing node.

Summary of the invention

The embodiments of the present application provide a database backup method, a server, and a shared memory device, which can ensure data synchronization between the database in the main computing node and the backup computing node.

In a first aspect, an embodiment of the present application provides a database backup method, which is applied to a shared memory device, and the shared memory device communicates with a primary computing node and a standby computing node respectively based on a computing fast link CXL protocol; the method includes: receiving a target log sent by the primary computing node; the target log is used to record data changes in the database in the primary computing node; receiving a data read request from the standby computing node, the data read request is used to request to read the target log; and in response to the data read request, sending the target log to the standby computing node.

The database backup method provided by the embodiment of the present application introduces a shared memory device between the main computing node and the standby computing node, and the main and standby computing nodes communicate with the shared memory device respectively through the CXL protocol. After the main computing node commits a transaction and generates a target log in the log buffer, the main computing node can send the target log to the shared memory device, so that after the subsequent standby computing node sends a data read request to the shared memory device, the shared memory device can send the target log to the standby computing node. In this way, the standby computing node can update data according to the target log to realize the database backup process. This method of storing the target log to be sent to the standby computing node through an independent shared memory device of the device can ensure that the failure of the main computing node will not affect the log that has not been sent to the standby computing node, avoid the situation where the data between the main and standby computing nodes is not synchronized due to the loss of the log, and ensure the reliability of the database system. In addition, the CXL protocol is a high-speed interconnection protocol. Communication through the CXL protocol can ensure the transmission efficiency of the log and improve the operating performance of the database system.

In one possible implementation, the shared memory device is configured with a transmission buffer, which is used to store logs to be sent to the standby computing node; after receiving the target log sent by the main computing node, the method also includes: saving the target log to the transmission buffer; responding to the data read request, sending the target log to the standby computing node, including: responding to the data read request, obtaining the target log from the transmission buffer, and sending the target log to the standby computing node.

In another possible implementation, the target log is stored in a target area in the transmission buffer; after the target log is saved in the transmission buffer, the method further includes: configuring an occupation flag for the target area; the occupation flag is used to indicate that the content in the target area cannot be overwritten.

In yet another possible implementation, after acquiring the target log from the transmission buffer, the method further includes: deleting an occupation flag configured for the target area.

In another possible implementation, the shared memory device is configured with a server of a shared file system; receiving a target log sent by the primary computing node includes: receiving the target log sent by a first client through the server; the first client is configured on the primary computing node. Receiving a data read request from a standby computing node includes: receiving a data read request sent by a second client through the server; the second client is configured on the standby computing node.

In the second aspect, an embodiment of the present application provides a database backup method, which is applied to a main computing node, and the main computing node communicates with a shared memory device based on a computing fast link CXL protocol. The method includes: the main computing node generates a target log based on a submitted transaction; wherein the target log is stored in a log buffer; the log buffer is deployed on the main computing node, and is used to cache logs generated during the operation of the main computing node; the target log is used to record data changes in the database of the main computing node due to transactions; the target log is obtained from the log buffer; the target log is sent to the shared memory device, so that the standby computing node obtains the target log from the shared memory device; and the target log in the log buffer is stored in the hard disk of the main computing node.

In a third aspect, an embodiment of the present application provides a database backup method, which is applied to a standby computing node, and the standby computing node communicates with a shared memory device based on a computing fast link CXL protocol, and the method includes: sending a data read request to the shared memory device; the data read request is used to request to read a target log, and the target log is used to record data changes in the database in the main computing node; receiving the target log sent by the shared memory device; according to the target log, updating the data in the database in the standby computing node so that the data in the database in the standby computing node is synchronized with the data in the database in the main computing node; storing the target log in the hard disk of the standby computing node.

In a fourth aspect, an embodiment of the present application provides a database backup method, which is applied to a database system, wherein the database system includes a main computing node, a shared memory device and a standby computing node; the shared memory device communicates with the main computing node and the standby computing node respectively based on the computing fast link CXL protocol; the main computing node generates a target log based on the submitted transaction; wherein the target log is stored in a log buffer, and the log buffer is deployed on the main computing node for caching logs generated during the operation of the main computing node; the target log is used to record data changes in the database of the main computing node due to transactions; the main computing node obtains the target log from the log buffer and sends the target log to the shared memory device; the main computing node stores the target log in the log buffer in the hard disk of the main computing node; the standby computing node sends a data read request to the shared memory device, and the data read request is used to request to read the target log; the shared memory device sends the target log to the standby computing node in response to the data read request; the standby computing node updates the data of the database in the standby computing node according to the target log, so that the data of the database in the standby computing node is synchronized with the data of the database in the main computing node; the standby computing node stores the target log in the hard disk of the standby computing node.

In a fifth aspect, an embodiment of the present application provides a database system, including a main computing node, a standby computing node and a shared memory device; the shared memory device communicates with the main computing node and the standby computing node respectively based on the computing fast link CXL protocol; the main computing node is used to generate a target log based on the submitted transaction; wherein the target log is stored in a log buffer, the log buffer is deployed on the main computing node, and is used to cache the logs generated during the operation of the main computing node; the target log is used to record the data changes in the database in the main computing node due to the transaction; the main computing node is also used to obtain the target log from the log buffer and send the target log to the shared memory device; the main computing node is also used to store the target log in the log buffer in the hard disk of the main computing node; the standby computing node is used to send a data read request to the shared memory device, the data read request is used to request to read the target log; the shared memory device is used to send the target log to the standby computing node in response to the data read request; the standby computing node is also used to update the data of the database in the standby computing node according to the target log, so that the data of the database in the standby computing node is synchronized with the data of the database in the main computing node; the standby computing node is also used to store the target log in the hard disk of the main computing node.

In a sixth aspect, an embodiment of the present application provides a database backup device, which includes: a receiving module and a sending module; the receiving module is used to receive a target log sent by a primary computing node; the target log is used to record data changes in the database in the primary computing node; the receiving module is also used to receive a data read request from a standby computing node, the data read request is used to request to read the target log; the sending module is used to send the target log to the standby computing node in response to the data read request.

In the seventh aspect, an embodiment of the present application provides a database backup device, which includes: a generation module, an acquisition module, a sending module and a storage module; the generation module is used to generate a target log based on a submitted transaction; wherein the target log is stored in a log buffer, and the log buffer is deployed on the main computing node to cache the logs generated during the operation of the main computing node; the target log is used to record the data changes in the database of the main computing node due to transactions; the acquisition module is used to obtain the target log from the log buffer; the sending module is used to send the target log to a shared memory device, so that the standby computing node obtains the target log from the shared memory device; the storage module is used to store the target log in the log buffer in the hard disk of the main computing node.

In an eighth aspect, an embodiment of the present application provides a database backup device, which includes: a sending module, a receiving module, an update module and a storage module; the sending module is used to send a data read request to a shared memory device; the data read request is used to request to read a target log, and the target log is used to record data changes in the database in the main computing node; the receiving module is used to receive the target log sent by the shared memory device; the update module is used to update the data in the database in the standby computing node according to the target log, so that the data in the database in the standby computing node is synchronized with the data in the database in the main computing node; the storage module is used to store the target log in the hard disk of the standby computing node.

In the ninth aspect, an embodiment of the present application provides a server, which includes a processor and a memory; the processor is coupled to the memory; the memory is used to store computer instructions, and the computer instructions are loaded and executed by the processor so that the server implements the methods of the first to fourth aspects above.

In a tenth aspect, an embodiment of the present application provides a shared memory device that executes the method of the first aspect above.

In the eleventh aspect, an embodiment of the present application provides a computer-readable storage medium, which includes: computer software instructions; when the computer software instructions are executed in a server, the server implements the methods of the first to third aspects above.

In a twelfth aspect, an embodiment of the present application provides a computer program product, which, when running on a server, enables the server to implement the methods of the first to third aspects above.

The beneficial effects of the fourth to twelfth aspects mentioned above can be referred to the corresponding description of the first aspect and will not be repeated here.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic diagram of the architecture of a current database system provided in an embodiment of the present application;

FIG2 is a schematic diagram of the architecture of a database system provided in an embodiment of the present application;

FIG3 is a schematic diagram of a connection method for connecting a shared memory device based on CXL according to an embodiment of the present application;

FIG4 is a schematic diagram of the architecture of a server provided in an embodiment of the present application;

FIG5 is a schematic diagram of the architecture of a shared file system provided in an embodiment of the present application;

FIG6 is a schematic diagram of a flow chart of a database backup method provided in an embodiment of the present application;

FIG7 is a schematic diagram of the composition of a database backup device provided in an embodiment of the present application;

FIG8 is a schematic diagram of the composition of a database backup device provided in an embodiment of the present application;

FIG9 is a schematic diagram of the composition of a database backup device provided in an embodiment of the present application;

FIG. 10 is a schematic diagram of the composition of a server provided in an embodiment of the present application.

DETAILED DESCRIPTION

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 only 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.

It should be noted that, in the embodiments of the present application, words such as "exemplarily" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "exemplarily" 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 "exemplarily" or "for example" is intended to present related concepts in a specific way.

In order to facilitate the clear description of the technical solutions of the embodiments of the present application, in the embodiments of the present application, words such as "first" and "second" are used to distinguish between identical items or similar items with basically the same functions and effects. Those skilled in the art can understand that words such as "first" and "second" do not limit the quantity and execution order.

The following is a brief introduction to the technical terms involved in the embodiments of the present application:

1. High-performance analytical application (HANA): It is a memory-based relational database management system that is row-oriented and column-oriented. System applications and products (SAP) HANA is a combined application that integrates HANA database, data modeling, HANA management and data supply. SAP HANA is widely used in memory computing engines to process and analyze large amounts of real-time data, providing enterprises with real-time data analysis and management capabilities.

2. Transaction: In the database field, a transaction is a series of operations performed to implement a single logical function. These operations are either all executed or not executed, and are an indivisible whole. The main purpose of a transaction is to ensure the integrity and consistency of the database, and to maintain data consistency even in the event of a failure. For example, through a database operation, user A transfers 100 yuan to user B. The transaction corresponding to this database operation includes the following steps: (1) Start the transaction. (2) Execute the operation of deducting 100 yuan from user A's account. (3) Execute the operation of adding 100 yuan to user B's account. (4) Execute the operation of inserting a transfer record into the transaction record table. (5) Commit the transaction (save the data changes involved in the above steps (1)-(4) in the database). If there is an error in any link during the execution of the transaction (for example, the balance of user A's account is insufficient), all operations will not be saved in the database (for example, user B's account will not increase by 100 yuan, and the transaction record table will not record the transfer), ensuring that the database data will not be affected.

3. Compute Express Link (CXL): CXL is an open interconnection protocol standard that aims to provide an efficient, high-speed, and low-latency interface between processors and dedicated accelerators and high-performance storage systems to meet the needs of resource sharing, memory pooling, and efficient computing scheduling. CXL is characterized by high efficiency and low latency.

For systems such as SAP S4/HANA and BW4/HANA (S4/HANA is an enterprise resource planning system based on in-memory computing and running in real time. BW4/HANA is a solution in the field of business intelligence and data warehouse), high availability and disaster recovery are important factors that need to be considered when building them. Since both SAP S4/HANA and BW4/HANA run on the HANA database, it is crucial to ensure the high availability and disaster recovery of the HANA database. The high availability and disaster recovery functions mentioned here are generally achieved through backup, storage replication, system replication and other functions. Backup refers to creating a copy of the data and saving it to a safe place. When the original data is lost or damaged, the backup can be used for recovery. Storage replication refers to copying data from one storage system to another.

System replication is a solution provided by HANA database to ensure high availability and disaster recovery. As shown in Figure 1, it is applied to the disaster recovery system composed of primary and secondary computing nodes. The primary computing node is deployed with HANA database, and the same HANA database is backed up in the secondary computing node through system replication. In this way, the services (specific functions or components provided by HANA database, used to perform operations such as data storage and query) and instances (including related processes and memory structures to support the operation of services) in the primary computing node correspond to the services in the secondary computing node, and the corresponding services in the two nodes communicate in pairs. This communication method can ensure that the primary computing node and the secondary computing node remain synchronized (for example, if service A in the primary computing node performs a data write operation, then service A in the secondary computing node also performs the corresponding data write operation).

During the operation of the main computing node, if the service A in the main computing node executes a transaction that causes its own data to change, the main computing node can generate a log to record the data change, and the generated log is temporarily stored in the log buffer. The log can be sent to the standby computing node later, so that the service A in the standby computing node can update its own data according to the log to achieve data synchronization between the two nodes. Since the log only records the information of data changes, rather than the complete data set, the main computing node only needs to transmit the log instead of the entire data set, which can effectively reduce the amount and time of data transmission. The user end or application server can use the business services provided by the main computing node. When the main computing node fails, the standby computing node can take over the business services of the main computing node in time, that is, the user end or application server continues to use the business by accessing the standby computing node, ensuring the continuity of the business, the duration of business recovery and the amount of data recovery.

In current database systems, log backup is generally performed between the primary computing node and the secondary computing node using synchronous data replication or asynchronous data replication.

Synchronous data replication means that when the primary computing node executes a transaction to back up the log, the primary computing node first obtains the log and sends the log to the standby computing node through the network. After the standby computing node receives the log and stores it persistently on the hard disk, it returns a confirmation message to the primary computing node. After receiving the confirmation message, the primary computing node stores the log persistently on the hard disk.

It can be seen that in the synchronous data replication method, the primary computing node needs to wait for the confirmation of the backup computing node before executing the subsequent "local disk" (that is, storing logs on the hard disk) process. In order to reduce the impact of the information transmission process between the primary and backup nodes of the database on the primary computing node, this method has high requirements for low latency. Therefore, the primary and backup computing nodes are generally deployed in the same computer room or in the same city, resulting in limited usage scenarios and poor disaster recovery capabilities.

Asynchronous replication means that when the main computing node executes a transaction to back up the log, the main computing node first obtains the log and writes the log to the transmission buffer configured by itself (also called asynchronous log transmission buffer, logshipping_async_buffer), and then the main computing node can store the log persistently on the hard disk. In addition, the node in the database will start a separate thread to send the log in the transmission buffer to the standby computing node through the network. It should be understood that if the generated log is sent to the standby computing node every time it is generated, a large amount of network resources will be consumed. Therefore, by setting a transmission buffer for temporarily storing the logs to be sent to the standby computing node, and then sending the logs in the transmission buffer to the data standby node in a unified or sequential manner, network resources can be effectively saved.

The current asynchronous data replication method has the following disadvantages: if the main computing node fails, logs in the transmission buffer that have not been sent to the backup computing node will be lost, resulting in the data in the database of the backup computing node being unable to keep pace with the data in the database of the main computing node.

In summary, how to ensure that data synchronization between the primary and backup nodes of the database is not affected by the failure of the primary computing node is an urgent problem to be solved.

Based on this, an embodiment of the present application provides a database backup method. The database backup method provided by the embodiment of the present application introduces a shared memory device between the main computing node and the standby computing node, and the main and standby computing nodes communicate with the shared memory device through the CXL protocol respectively. After the main computing node commits a transaction and generates a target log in the log buffer, the main computing node can send the target log to the shared memory device, so that after the subsequent standby computing node sends a data read request to the shared memory device, the shared memory device can send the target log to the standby computing node. In this way, the standby computing node can update data according to the target log to realize the database backup process. This method of storing the target log to be sent to the standby computing node through an independent shared memory device of the device can ensure that the failure of the main computing node will not affect the log that has not been sent to the standby computing node, avoid the situation where the data between the main and standby computing nodes is not synchronized due to the loss of the log, and ensure the reliability of the database system. In addition, the CXL protocol is a high-speed interconnection protocol. Communication through the CXL protocol can ensure the transmission efficiency of the log and improve the operating performance of the database system.

The embodiments provided in this application are described in detail below in conjunction with the accompanying drawings.

Fig. 2 is a schematic diagram of the architecture of a database system provided by an embodiment of the present application. As shown in Fig. 2, it includes a main computing node, a shared memory device and a standby computing node.

The master computing node includes a data area, a log buffer, a data file, and a log file. The data area is a memory area in the master computing node, which is used to process data. The data in the data area can be persistently stored in the hard disk of the master computing node in the form of a data file. The log buffer is a memory area of the master computing node, which is used to cache logs generated during the operation of the master computing node. The master computing node can store the logs in the log buffer in the form of a log file in the hard disk of the master computing node.

The shared memory device includes a transmission buffer, which is a memory area in the shared memory device for storing logs to be sent to the standby computing node.

The standby computing node includes a data area, a log receiving area, data files, and log files. The functions of the data area, data files, and log files are the same as those of the main computing node, and will not be repeated here. The log receiving area is a memory area in the standby computing node, which is used to cache logs obtained from the main computing node.

In an embodiment of the present application, during the initialization phase of the database system startup, the main computing node can back up the original data in the main computing node to the standby computing node by means of the aforementioned system replication, so that the original data in the main and standby computing nodes remain consistent. During the operation in the main computing node, after the transaction is submitted, the main computing node can copy the target log generated in the log buffer to the transmission buffer in the shared memory device, and store the log in the log buffer in a log file. Further, the standby computing node reads the log from the log transmission buffer, and temporarily stores it in its own log receiving area for further processing, such as updating its own data according to the log, or storing the log in a log file. In this embodiment, the log records the information of data changes, rather than a complete data snapshot, so when performing a database backup, only these logs need to be transmitted, and the entire data set does not need to be transmitted. This effectively reduces the amount and time of data transmission and improves the efficiency of backup.

The shared memory device is introduced below.

The shared memory device may include a processor, a CXL controller, and a memory. As shown in FIG3 , the processor of the main computing node may be connected to the CXL controller in the shared memory device via the CXL protocol to access the memory in the shared memory device. Similarly, the processor of the standby computing node may be connected to the CXL controller in the shared memory device via the CXL protocol to access the memory in the shared memory device. The processors mentioned here may all be central processing units (CPUs).

Among them, the processor in the shared memory device is responsible for executing instructions, performing calculations and controlling other components of the shared memory device. In this embodiment, the processor in the shared memory device and the CXL controller work together to achieve efficient data transmission between the shared memory device and other components (such as the processors in the primary/standby computing nodes). In this embodiment, an operating system is configured in the shared memory device, and a server of a shared file system runs in the operating system. The processor of the shared memory device receives and processes data write requests from the primary computing node or data read requests from the standby computing node, as well as operations such as saving logs or reading logs by running the server. For the specific process, see the relevant description of the corresponding embodiment of Figure 6 below.

The CXL controller is a key component that connects the processor and memory. It is responsible for managing and controlling the transmission of data between these components, determining the consistency and integrity of the data, and ensuring that the processor and other components (such as memory) work together more efficiently. The CXL controller is one of the device types defined by the CXL protocol.

The CXL controller follows two sub-protocols included in the CXL protocol, namely CXL.io and CXL.mem. CXL.io is mainly used for initialization, register access, interrupt processing, and memory-mapped I/O transactions. This enables the CPU to effectively access the configuration space of external devices and is a key link in achieving efficient communication between the CPU and external devices. CXL.mem allows the CPU to access the memory of external devices using load and store commands. In addition, the CXL protocol can be used as a memory extension, allowing the CPU to directly access the memory of external devices, thereby realizing the pooling of memory resources and effectively improving the system's memory utilization efficiency and performance.

Exemplarily, the CXL controller may be a control chip integrated in a shared memory device, and the control chip may also be referred to as a memory expansion (CXL memory expander) chip, a CXL memory controller (CXL memory expandercontroller) or a CXL memory pooling (CXL memory expander pooling) chip.

The memory in the shared memory device may include a dual inline memory module (DIMM), a dynamic random access memory (DRAM), a synchronous dynamic random access memory (SDRAM), or a persistent memory (PMEM). This embodiment is described by taking the memory including DIMM or DRAM as an example.

In an embodiment of the present application, a shared memory area is configured in a shared memory device, and the above-mentioned transmission buffer is an area in the shared memory area. Among them, the shared memory area is a large piece of memory applied to the shared memory device. For a server, during the operation of an application in the server, the application often applies to or releases a piece of memory from the operating system of the server, and the frequent application to or release of memory from the operating system requires a frequent system call process, which will reduce the operating efficiency of the server. Therefore, the embodiment of the present application adopts the method of a shared memory area, and applies to the operating system of the shared memory device for a large piece of memory at one time as a shared memory area, and then directly uses the space in the shared memory area when reading and writing data, without applying to the operating system. By using the shared memory area as the intermediate storage medium of the target log, through the CXL protocol, the CPU of the primary and standby computing nodes can directly access the memory control of the shared memory area, which can effectively improve the operating efficiency of the database system and promote the rapid transmission and sharing of data in the primary and standby computing nodes.

In the embodiment of the present application, both the main computing node and the backup computing node can be servers. Specifically, in terms of form, the server can be a blade server, a high-density server, a rack server or a whole cabinet server; in terms of function, the server can be a general-purpose server, a graphics processing unit (GPU) server, an artificial intelligence (artificial intelligence) server, etc. FIG. 4 is a schematic diagram of the system architecture of the server. As shown in FIG. 4 , the hardware of the server includes a processor, an out-of-band controller, a hard disk and a memory. The software includes an out-of-band management module and an operating system (OS).

The out-of-band management module runs in the out-of-band controller, and the OS runs on the processor (as shown in FIG. 4 ).

The out-of-band management module may be a management unit of a non-business module. For example, the out-of-band management module may remotely maintain and manage the server through a dedicated data channel. The out-of-band management module is completely independent of the server's operating system and may communicate with the OS through the server's out-of-band management interface.

Exemplarily, the out-of-band management module may include a management unit for the server operation status, a management system in a management chip outside the processor, a server baseboard management controller (BMC), a system management module (SMM), etc. It should be noted that the embodiments of the present application do not limit the specific form of the out-of-band management module, and the above is only an exemplary description.

The memory, also called internal memory or main memory, is installed in the memory slot on the motherboard of the server. The memory communicates with the memory controller through a memory channel. The memory has at least one memory column, each memory column is located on one side of the memory, each memory column includes at least one sub-memory column, and the memory column or sub-memory column includes multiple memory chips (devices). Each memory chip is divided into multiple storage array groups (bankgroups), each storage array group includes multiple storage arrays (banks), each storage array is divided into multiple storage cells (cells), each storage cell has a row address and a column address, and each storage cell includes one or more bits. In one division method, the memory can be divided into memory chips, storage array groups, storage arrays, storage rows/storage columns, storage cells, and bits from the upper level to the lower level, wherein the addresses of memory particles, storage array groups, storage arrays, storage rows, storage columns, storage cells, and bits on the memory are real physical addresses. In another division method, the CPU divides the memory chip into multiple memory pages based on the paging mechanism, where the address of the memory page is a virtual address that needs to be converted before it becomes a real physical address.

The hard disk may be a plug-in hard disk equipped on the server, a smart media card (SMC), a secure digital (SD) card, a flash card, etc., or an external storage device such as a USB flash drive.

It should be noted that the system architecture and application scenarios described in the embodiments of the present application are intended to more clearly illustrate the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided in the embodiments of the present application. Ordinary technicians in this field can know that with the evolution of the system architecture and the emergence of new business scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

Fig. 5 is a schematic diagram of the architecture of a database backup system provided by an embodiment of the present application, and the database backup system can also be referred to as a shared file system. As shown in Fig. 5, the system includes two parts, a client and a server. The client can interact with the user process to access the client's local hard disk through the local file access module according to the instructions of the user process. Alternatively, the client can interact with the server through the communication module according to the instructions of the user process. For example, the client can send a read data request, the client's communication module converts the read data request into a function, sends the function to the server based on the remote procedure call protocol, and the server's communication module converts the function into a locally recognizable command, so that the server's local file access module accesses the server's local hard disk according to the locally recognizable command to read data, and returns the read data to the client through the communication module.

In addition, the server also deploys an authentication module (Rpc.nfsd) and an authorization module (Rpc.mount). The authentication module is used to determine whether the client is legitimate and can connect to the server. The authorization module is used to determine the client's operating authority on the server, such as whether data reading and writing are allowed.

In the example of the present application, a shared file system server is configured in the shared memory device (specifically, the processor in the shared memory device is responsible for the operation of the server), and clients are respectively run in the main computing node and the standby computing node, so that both can access the data in the shared memory device together. For the sake of distinction, the client running in the main computing node is referred to as the first client in the embodiment of the present application, and the client running in the standby computing node is referred to as the second client.

As an example, the shared file system can be a network file system (NFS). NFS is a file system based on the User Datagram Protocol (UDP). Its implementation mainly uses the Remote Procedure Call protocol to connect the client and the server. The Remote Procedure Call protocol provides a set of operations for accessing remote files that are independent of the machine, operating system, and underlying transmission protocol. NFS allows remote clients to access the data on the server through the network in a similar way to accessing the local file system.

Fig. 6 is a flow chart of a database backup method provided in an embodiment of the present application. Exemplarily, the database backup method provided in an embodiment of the present application can be applied to the system architecture shown in Fig. 2. The shared memory device communicates with the main computing node and the standby computing node respectively based on the CXL protocol.

As shown in FIG6 , the database backup method provided in the embodiment of the present application may specifically include the following steps:

S601. The main computing node generates a target log based on the submitted transaction.

S602: The main computing node obtains the target log from the log buffer.

The target log is stored in the log buffer, which is deployed in the main computing node and used to cache the logs generated during the operation of the main computing node. The target log is used to record the data changes caused by transactions in the database of the main computing node.

With respect to the above S601-S602, as mentioned above, the main computing node can provide services to the outside. In the process of providing services to the outside, the main computing node will submit various transactions, which are accompanied by a large number of data operations (such as data addition, deletion, modification, and query operations). These data operations will cause the data in the database to change. Therefore, the main computing node will generate a target log to record these data changes. These target logs are temporarily stored in the log buffer of the main computing node. When performing log backup later, the main computing node can obtain these target logs from the log buffer.

It should be understood that by recording data changes caused by transactions in logs, on the one hand, the primary computing node can recover data based on the logs when data is lost, and on the other hand, the backup computing node can keep data synchronized with the primary computing node based on the logs.

S603. The main computing node sends a target log to the shared memory device; correspondingly, the shared memory device receives the target log sent by the main computing node.

In an embodiment of the present application, when target logs are generated in the log buffer, the primary computing node may obtain these target logs and send the target logs to the shared memory device so that the standby computing node obtains the target logs from the shared memory device.

In one implementation, the main computing node and the shared memory device can achieve the transmission of the target log through the shared file system shown in Figure 5. Among them, the main computing node is configured with a first client, and the shared memory device is configured with a server of the shared file system. By running the client and the server, a communication link can be established between the main computing node and the shared memory device. Therefore, the above S603 can be implemented as follows: the main computing node runs the first client and sends the target log to the shared memory device. Correspondingly, the shared memory device receives the target log sent by the first client through the server.

In one implementation, a transmission buffer is configured in the shared memory device, and the transmission buffer is used to store logs to be sent to the standby computing node. In the above S603, after the shared memory device node sends the target log to the master computing node, the shared memory device also performs the following: save the target log to the transmission buffer. So that the subsequent standby computing node can read the target log from the transmission buffer.

It should be understood that setting the transmission buffer in a shared memory device can effectively avoid the situation in the related art where the main computing node fails and causes the loss of logs in the transmission buffer, thereby ensuring the disaster recovery performance of the database system. In addition, setting the transmission buffer in an independent shared memory device can ensure that the transmission buffer has sufficient capacity space. This avoids the main computing node being unable to write logs due to insufficient transmission buffer capacity when the main computing node is running under high load, thereby affecting the operating performance of the main computing node.

In one implementation, the target log is stored in the target area in the transmission buffer. After the shared memory device saves the target log to the transmission buffer, the shared memory device further performs: configuring an occupation mark for the target area, the occupation mark is used to indicate that the content in the target area cannot be overwritten.

The processor in the shared memory device manages the occupation flag by running the operating system. The occupation flag can be stored in the kernel space of the operating system, ensuring that the occupation flag can be quickly accessed and modified by the operating system, ensuring efficient access and management of files.

S604: The main computing node stores the target log in the hard disk of the main computing node.

It should be understood that, as mentioned above, the log buffer is a memory area, and the memory area is generally used to temporarily store data. Therefore, in order to ensure that the target log is not lost, the main computing node can store the target log persistently in its own hard disk to ensure the persistence of the target log, so as to facilitate data recovery based on the target log when needed later.

It should be noted that the embodiment of the present application does not limit the specific execution sequence of the above S603 and S604. The two steps are independent of each other and can be executed successively or simultaneously.

S605: The standby computing node sends a data read request to the shared memory device; correspondingly, the shared memory device receives the data read request from the standby computing node.

The data read request is used to request to read the target log.

In the embodiment of the present application, the standby computing node can send a data read request to the shared memory device in real time or under preset conditions to request to read the target log. Correspondingly, the shared memory device can receive the data read request and respond.

As an example, the preset condition may be that the standby computing node receives an instruction on log backup, which may come from the main computing node, a terminal of a database administrator, etc. Alternatively, the preset condition may be that the number of logs in the transmission buffer reaches a certain threshold (the shared memory device may detect the number of logs in the transmission buffer or the size of the occupied space in the transmission buffer in real time, and instruct the standby computing node to read when the number of logs or the size of the occupied space exceeds a certain threshold).

In one implementation, the standby computing node and the shared memory device can read the target log through the shared file system shown in Figure 5. Among them, the standby computing node is configured with a second client, and the shared memory device is configured with a server of the shared file system. By running the client and the server, a communication link can be established between the standby computing node and the shared memory device. Therefore, the above S605 can be implemented as follows: the standby computing node runs the second client and sends a data read request to the shared memory device. Correspondingly, the shared memory device receives the data read request sent by the second client through the server.

S606: The shared memory device sends the target log to the standby computing node in response to the data read request; correspondingly, the standby computing node receives the target log sent by the shared memory device.

In an embodiment of the present application, after receiving a data read request from a standby computing node, the shared memory device can respond to the request and send a target log to the standby computing node. Among them, the shared memory device can record the backup status of the target log (i.e., whether it has been sent to the standby computing node). After receiving the data read request from the standby computing node, the shared memory device can obtain the target log that has not been backed up and send it to the standby computing node. A log receiving area is configured in the standby computing node. After receiving the target log, the standby computing node can store the target log in the log receiving area to facilitate further processing of the target log later, such as executing S607 and S608 below.

In one implementation, a transmission buffer is configured in the shared memory device, and the transmission buffer is used to store logs to be sent to the standby computing node. Therefore, the above S606 can be implemented as: in response to the data read request, obtaining the target log from the transmission buffer and sending the target log to the standby computing node.

It can be seen that by setting a transmission buffer to store logs to be sent to the standby computing node, after receiving a data read request from the standby computing node, the target logs accumulated in the transmission buffer can be uniformly obtained and sent, thereby effectively saving network resource consumption.

In an implementation manner, after the shared memory device obtains the target log from the transmission buffer, the shared memory device further executes: deleting an occupation identifier configured in the target area.

It should be understood that by configuring and deleting the occupation mark, the shared memory device can record the backup status of the target log. There is no need for the main computing node or the shared memory device to inform the standby computing node which logs are newly generated and which logs have been backed up through information interaction. After receiving a data read request, the shared memory device can obtain the target log from the target area configured with the occupation mark and send it to the standby computing node. Then, the occupation marks of these target areas are deleted to avoid repeatedly obtaining the previously backed-up logs when sending logs to the standby computing node next time.

Exemplarily, the transmission buffer of the shared memory device is divided into 10 areas (area 1 to area 10 respectively). The shared memory device receives the target log sent by the main computing node and stores the target log in area 1. Furthermore, the shared memory device can configure an occupation mark for area 1. In this way, if the main computing node sends a new target log again, the content in area 1 cannot be overwritten, thereby avoiding the loss of the log. Similarly, after the standby computing node sends a data read request, the shared memory device can query which areas have occupation marks. The shared memory device can determine that the target log in area 1 with the occupation mark is a target log that has not been backed up, and can be obtained and sent to the standby computing node.

In addition, when the memory space is tight, the shared memory device can overwrite and save the new target log in the area without the configuration occupation mark, so as to ensure the full utilization of the storage space in the shared memory device.

S607: The standby computing node updates the data in the database of the standby computing node according to the target log.

In the embodiment of the present application, as described above, the target log records the data changes in the database in the primary computing node. During the database backup process, the standby computing node can update the data in its own database according to the content of the target log, so that the data in the database in the standby computing node is synchronized with the data in the database in the primary computing node.

S608: Store the target log in the hard disk of the standby computing node.

For the same reason as in S604 above, in order to ensure that the target log is not lost, the standby computing node may persistently store the target log in its own hard disk to ensure the persistence of the target log and facilitate subsequent data recovery based on the target log when needed.

It should be noted that the embodiment of the present application does not limit the specific execution sequence of the above S607 and S608. The two steps are independent of each other and can be executed successively or simultaneously.

A database backup method provided in an embodiment of the present application is applied to the database system shown in FIG2 , and includes the following steps:

1. The main computing node generates a target log based on the submitted transaction. The target log is stored in the log buffer, which is deployed on the main computing node and used to cache the logs generated during the operation of the main computing node. The target log is used to record the data changes caused by the transaction in the database of the main computing node.

2. The master computing node obtains the target log from the log buffer and sends the target log to the shared memory device;

3. The main computing node stores the target log in the log buffer in the hard disk of the main computing node;

4. The standby computing node sends a data read request to the shared memory device. The data read request is used to request to read the target log.

5. The shared memory device responds to the data read request and sends the target log to the standby computing node;

6. The standby computing node updates the data in the database of the standby computing node according to the target log, so that the data in the database of the standby computing node is synchronized with the data in the database of the primary computing node;

7. The standby computing node stores the target log in the hard disk of the standby computing node.

It should be understood that the above steps 2 and 3 can be performed in a different order or simultaneously. Similarly, the above steps 6 and 7 can be performed in a different order or simultaneously.

The following is a summary of the database backup solution provided by the embodiment of the present application. At the beginning, the master and standby nodes are connected to the shared memory device through the CXL protocol, and the shared memory area is configured in the shared memory device in advance, and the server of the shared file system is configured in the shared memory area. Secondly, the access strategy of the shared file system is configured for the master and standby nodes of the database: the master computing node runs the first client, and accesses the shared memory device through the first client to write the target log. The standby computing node runs the second client, and accesses the shared memory device through the second client to read the target log. The read-write strategy is configured for the shared memory device: after the master computing node writes the target log in the target area, the shared memory device configures an occupation mark for the target area; after the standby computing node reads the data in the target area, the shared memory device deletes the occupation mark configured in the target area. Then, modify the configuration of the master computing node, configure the transmission buffer in the original master computing node to the shared memory device, and the standby computing node reads the target log from the shared memory device. During operation, the master computing node writes the target log from the transmission buffer, and the standby computing node reads the target log from the transmission buffer to cyclically use the shared file system.

The embodiment of the present application provides a database backup method, which introduces a shared memory device between the main computing node and the standby computing node, and the main and standby computing nodes communicate with the shared memory device respectively through the CXL protocol. After the main computing node commits a transaction and generates a target log in the log buffer, the main computing node can send the target log to the shared memory device, so that after the subsequent standby computing node sends a data read request to the shared memory device, the shared memory device can send the target log to the standby computing node. In this way, the standby computing node can update data according to the target log to realize the database backup process. This method of storing the target log to be sent to the standby computing node through an independent shared memory device of the device can ensure that the failure of the main computing node will not affect the log that has not been sent to the standby computing node, avoid the situation where the data between the main and standby computing nodes is not synchronized due to the loss of the log, and ensure the reliability of the database system. In addition, the CXL protocol is a high-speed interconnection protocol. Communication through the CXL protocol can ensure the transmission efficiency of the log and improve the operating performance of the database system.

The above mainly introduces the solution provided by the embodiment of the present application from the perspective of the method. In order to achieve the above functions, the embodiment of the present application provides a hardware structure and/or software module corresponding to each function. It should be easily appreciated by those skilled in the art that, in combination with the modules and algorithm steps of each example described in the embodiment disclosed herein, the embodiment of the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

In an exemplary embodiment, the present application also provides a database backup device. The database backup device may be the above-mentioned main computing node, shared memory device or standby computing node. The database backup device may include one or more functional modules for implementing the database backup method of the above method embodiment.

For example, Fig. 7 is a schematic diagram of the composition of a database backup device provided in an embodiment of the present application. As shown in Fig. 7, the database backup device includes: a receiving module 701 and a sending module 702;

The receiving module 701 is used to receive the target log sent by the main computing node; the target log is used to record the data changes in the database of the main computing node;

The receiving module 701 is further used to receive a data read request from a standby computing node, where the data read request is used to request to read a target log;

The sending module 702 is used to send the target log to the standby computing node in response to the data read request.

In a possible implementation, the shared memory device is configured with a transmission buffer, and the transmission buffer is used to store the log to be sent to the standby computing node; the above-mentioned device also includes: a saving module 703. The saving module 703 is used to save the target log to the transmission buffer; the sending module 702 is specifically used to obtain the target log from the transmission buffer in response to the data read request, and send the target log to the standby computing node.

In another possible implementation, the target log is stored in the target area in the transmission buffer. The above device also includes: a configuration module 704. The configuration module 704 is used to configure an occupation mark for the target area; the occupation mark is used to indicate that the content in the target area cannot be overwritten.

In yet another possible implementation, the configuration module 704 is further configured to delete an occupation flag configured in the target area.

In another possible implementation, the shared memory device is configured with a server of a shared file system; the receiving module is specifically used to receive the target log sent by the first client through the server; the first client is configured on the main computing node. The receiving module is specifically used to receive the data read request sent by the second client through the server; the second client is configured on the standby computing node.

For example, Figure 8 is a schematic diagram of the composition of a database backup device provided in an embodiment of the present application. As shown in Figure 8, the database backup device includes: the device includes: a generating module 801, an acquiring module 802, a sending module 803 and a storing module 804;

The generation module 801 is used to generate a target log based on the submitted transaction; wherein the target log is stored in a log buffer, which is deployed on the main computing node and is used to cache the logs generated during the operation of the main computing node; the target log is used to record the data changes caused by the transaction in the database of the main computing node;

The acquisition module 802 is used to acquire the target log from the log buffer;

The sending module 803 is used to send the target log to the shared memory device so that the standby computing node obtains the target log from the shared memory device;

The storage module 804 is used to store the target log in the log buffer in the hard disk of the main computing node.

For example, Figure 9 is a schematic diagram of the composition of a database backup device provided in an embodiment of the present application. As shown in Figure 9, the database backup device includes: a sending module 901, a receiving module 902, an updating module 903 and a storage module 904;

The sending module 901 is used to send a data read request to the shared memory device; the data read request is used to request to read the target log, and the target log is used to record the data changes in the database in the main computing node;

The receiving module 902 is used to receive the target log sent by the shared memory device;

The updating module 903 is used to update the data in the database of the standby computing node according to the target log, so that the data in the database of the standby computing node is synchronized with the data in the database of the main computing node;

The storage module 904 is used to store the target log in the hard disk of the standby computing node.

In an exemplary embodiment, the embodiment of the present application further provides a server. FIG10 is a schematic diagram of the composition of the server provided in the embodiment of the present application. As shown in FIG10 , the server may include: a processor 1001 and a memory 1002; the memory 1002 stores instructions executable by the processor 1001; when the processor 1001 is configured to execute the instructions, the server implements the method described in the aforementioned method embodiment.

The embodiment of the present application provides a database system, including a main computing node, a standby computing node and a shared memory device; the shared memory device communicates with the main computing node and the standby computing node respectively based on the CXL protocol; the main computing node is used to generate a target log based on a submitted transaction; wherein the target log is stored in a log buffer, the log buffer is deployed on the main computing node, and is used to cache logs generated during the operation of the main computing node; the target log is used to record data changes in the database of the main computing node due to transactions; the main computing node is also used to obtain the target log from the log buffer and send the target log to the shared memory device; the main computing node is also used to store the target log in the log buffer in the hard disk of the main computing node; the standby computing node is used to send a data read request to the shared memory device, the data read request is used to request to read the target log; the shared memory device is used to send the target log to the standby computing node in response to the data read request; the standby computing node is also used to update the data of the database in the standby computing node according to the target log, so that the data of the database in the standby computing node is synchronized with the data of the database in the main computing node; the standby computing node is also used to store the target log in the hard disk of the main computing node.

The embodiment of the present application also provides a computer-readable storage medium. All or part of the processes in the above method embodiments can be completed by computer instructions to instruct the relevant hardware, and the program can be stored in the above computer-readable storage medium. When the program is executed, it may include the processes of the above method embodiments. The computer-readable storage medium can be the memory or memory of any of the above embodiments. The above computer-readable storage medium can also be an external storage device of the above database backup device, such as a plug-in hard disk, a smart memory card (smart media card, SMC), a secure digital (secure digital, SD) card, a flash card (flash card), etc. equipped on the above database backup device. Further, the above computer-readable storage medium can also include both the internal storage unit of the above database backup device and an external storage device. The above computer-readable storage medium is used to store the above computer program and other programs and data required for the above database backup device. The above computer-readable storage medium can also be used to temporarily store data that has been output or is to be output.

An embodiment of the present application also provides a computer program product, which includes a computer program. When the computer program product is run on a computer, the computer is enabled to execute any one of the database backup methods provided in the above embodiments.

Although the present application is described herein in conjunction with various embodiments, in the process of implementing the present application for which protection is sought, those skilled in the art may understand and implement other variations of the disclosed embodiments by viewing the drawings, the disclosure, and the appended claims. In the claims, the term "comprising" does not exclude other components or steps, and "one" or "an" does not exclude multiple components. A single processor or other unit may implement several functions listed in a claim. Certain measures are recorded in mutually different dependent claims, but this does not mean that these measures cannot be combined to produce good results.

Although the present application has been described in conjunction with specific features and embodiments thereof, it is obvious that various modifications and combinations may be made thereto without departing from the spirit and scope of the present application. Accordingly, this specification and the drawings are merely exemplary illustrations of the present application as defined by the appended claims, and are deemed to have covered any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, those skilled in the art may make various modifications and variations to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

The above are only specific implementations of the present application, but the protection scope of the present application is not limited thereto, and any changes or substitutions within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims (10)

1. A database backup method, characterized in that it is applied to a shared memory device, wherein the shared memory device communicates with a primary computing node and a backup computing node respectively based on a computing fast link CXL protocol; the method comprises: Receive a target log sent by the master computing node; the target log is used to record data changes in the database in the master computing node; receiving a data read request from the standby computing node, where the data read request is used to request to read the target log; In response to the data read request, the target log is sent to the standby computing node. 2. The method according to claim 1, characterized in that the shared memory device is configured with a transmission buffer, and the transmission buffer is used to store logs to be sent to the standby computing node; After receiving the target log sent by the main computing node, the method further includes: Saving the target log to the transmission buffer; In response to the data read request, sending the target log to the standby computing node includes: In response to the data read request, the target log is obtained from the transmission buffer, and the target log is sent to the standby computing node. 3. The method according to claim 2, wherein the target log is stored in a target area in the transmission buffer; after the target log is saved in the transmission buffer, the method further comprises: An occupation flag is configured for the target area; the occupation flag is used to indicate that the content in the target area cannot be overwritten. 4. The method according to claim 3, characterized in that, after obtaining the target log from the transmission buffer, the method further comprises: The occupancy flag configured in the target area is deleted. 5. The method according to any one of claims 1 to 4, characterized in that the shared memory device is configured with a server of a shared file system; the receiving the target log sent by the main computing node comprises: Receiving, through the server, the target log sent by the first client; the first client is configured on the main computing node; The receiving a data read request from the standby computing node includes: The data read request sent by the second client is received through the server; the second client is configured on the standby computing node. 6. A database backup method, characterized in that it is applied to a main computing node, the main computing node communicates with a shared memory device based on a computing fast link CXL protocol, and the method comprises: The master computing node generates a target log based on the submitted transaction; wherein the target log is stored in a log buffer, which is deployed on the master computing node and is used to cache logs generated during the operation of the master computing node; the target log is used to record data changes in the database of the master computing node caused by the transaction; Acquire the target log from the log buffer; Sending the target log to the shared memory device so that the standby computing node obtains the target log from the shared memory device; The target log in the log buffer is stored in the hard disk of the master computing node. 7. A database backup method, characterized in that it is applied to a standby computing node, the standby computing node communicates with a shared memory device based on a computing fast link CXL protocol, and the method comprises: Sending a data read request to the shared memory device; the data read request is used to request to read a target log, and the target log is used to record data changes in the database in the main computing node; Receiving the target log sent by the shared memory device; According to the target log, update the data in the database of the standby computing node so that the data in the database of the standby computing node is synchronized with the data in the database of the primary computing node; The target log is stored in the hard disk of the standby computing node. 8. A database backup method, characterized in that it is applied to a database system, wherein the database system includes a main computing node, a shared memory device and a standby computing node; the shared memory device communicates with the main computing node and the standby computing node respectively based on a computing fast link CXL protocol; The master computing node generates a target log based on the submitted transaction; wherein the target log is stored in a log buffer, which is deployed on the master computing node and is used to cache logs generated during the operation of the master computing node; the target log is used to record data changes in the database of the master computing node caused by the transaction; The main computing node obtains the target log from the log buffer and sends the target log to the shared memory device; The main computing node stores the target log in the log buffer in the hard disk of the main computing node; The standby computing node sends a data read request to the shared memory device, where the data read request is used to request to read the target log; The shared memory device sends the target log to the standby computing node in response to the data read request; The standby computing node updates the data in the database of the standby computing node according to the target log, so that the data in the database of the standby computing node is synchronized with the data in the database of the primary computing node; The standby computing node stores the target log in a hard disk of the standby computing node. 9. A server, characterized in that the server comprises a processor and a memory; the processor is coupled to the memory; the memory is used to store computer instructions, and the computer instructions are loaded and executed by the processor to enable the server to implement the database backup method as described in any one of claims 1-7. 10. A shared memory device, characterized in that it is used to execute the database backup method according to any one of claims 1 to 5.