CN106648959B - Data storage method and storage system - Google Patents

Abstract

Embodiments of the present application provide a data storage method and storage system. The method is applied to a storage system including a first storage device and at least one second storage device, wherein the second storage device is a mirrored storage device of the first storage device, and the method includes: the first storage device receives a write request message , the write request message carries target data; the first storage device generates a target log according to the target data, and the target log includes the target data and identification information for identifying the target data; the first storage device reports to the second The storage device sends the target log; the second storage device stores the target log, and the storage does not need to configure a memory index. Therefore, the data storage method and storage system of the embodiments of the present application can directly store the target log without configuring the memory index in the second storage device, thereby reducing the occupation of the CPU by the second storage device.

Description

Data storage method and storage system

technical field

The present application relates to the field of computer storage, and more particularly, to methods and storage systems for data storage.

Background technique

Mirroring is a type of redundancy. Data on one disk has an exact copy on another disk, which is mirroring. The common mirror operation on the server is to form a redundant array of independent disks (Redundant Array Of Independent Disks, RAID) 1. RAID 1 is a RAID disk array composed of two hard disks, and its capacity is only equal to the capacity of one hard disk, because the other is only as a data "mirror".

In a distributed storage system, in order to meet the requirements of reliability, a mirroring mechanism is chosen between nodes. A copy of the local data is stored on the mirror node, and services can be quickly restored in the event of a failure.

In the prior art, the data to be stored is sent to the mirror node, and the mirror node needs to configure a memory index for the data to be stored, and the memory index is used to facilitate reading the data stored by the mirror node; then the mirror node needs to apply for resource space and set resources The mapping relationship between the storage address of the space and the memory index; finally, the data to be stored is stored in the resource space corresponding to the storage address. In the prior art, the above operation of storing the data to be stored by the mirror node takes a relatively high occupation of a central processing unit (Central Processing Unit, CPU) of the mirror node.

SUMMARY OF THE INVENTION

Embodiments of the present application provide a data storage method and a storage system, which can reduce the occupation of a CPU for storing data by a mirrored storage device.

In a first aspect, an embodiment of the present application provides a data storage method, and the method is applied to a storage system including a first storage device and at least one second storage device, wherein the second storage device is the first storage device The mirror storage device, the method includes: a first storage device receives a write request message, the write request message carries target data; the first storage device generates a target log according to the target data, and the target log includes the target data and Identification information for identifying the target data; the first storage device sends the target log to the second storage device; the second storage device stores the target log, and does not need to configure a memory index for storing the target log.

The embodiments of the present application are applied to a storage system including a first storage device and at least one second storage device, and the second storage device is a mirror storage device of a type of first storage device. The first storage device receives the write request message, and the write request message carries the data to be written (represented as target data), generates a corresponding target log according to the target data, and sends the target log to the second storage device. It is directly stored in the second storage device, because the target log includes identification information used to represent the target data, so the second storage device does not need to create a memory index when storing the target log. Compared with the prior art, the second As far as the memory index needs to be configured on the storage device, the present application can directly store the target log, thereby saving the CPU occupation when the second storage device configures the memory index.

In some possible implementations, the method further includes: when the second storage device finishes storing the target log, sending a storage completion message to the first storage device; the first storage device stores the target log; the first storage device After completing the storage of the target log and receiving the storage completion message, the storage device sends a write request response message.

After the first storage device completes writing the target log and receives the storage completion message sent by the second storage device, it sends a write request response message to the IO device, so as to avoid the need to wait for the target data to be stored in the memory index corresponding to the prior art. Only after the resource space can return the write request response message to the IO device, the delay from issuing the write request to receiving the write request response message is reduced.

In some possible implementations, the second storage device includes a plurality of resource spaces; wherein, storing the target log in the second storage device includes: the second storage device determines, according to the identification information, a first storage device for storing the target log Resource space, the first resource space is any resource space in the plurality of resource spaces of the second storage device that meets the resource requirements of the target log; the second storage device stores the target log in the first resource space .

The second storage device selects an appropriate resource space (represented as a first resource space) according to the identification information, and stores the target log in the first resource space of the second storage device. In addition, the second storage device can identify the target data according to the identification information, and can regularly store the target data in order in the form of the target log, so that the second storage device does not need to configure a memory index for the target data, thereby saving CPU usage of the second storage device.

In some possible implementations, the second storage device includes multiple resource spaces; the method further includes: acquiring, by the first storage device, a storage address of the second resource space and a storage capacity of the second resource space, the second The resource space is any resource space in a plurality of resource spaces of the second storage device; the first storage device determines whether the second resource space meets the requirements of the target log according to the storage capacity of the second resource space and the identification information. Resource requirements; when the second resource space meets the resource requirements of the target log, the first storage device generates an RDMA message according to the RDMA encoding method, the target log and the storage address of the second resource space, and the RDMA message includes the The storage address of the second resource space and the target log; wherein, sending the target log by the first storage device to the second storage device includes: the first storage device sending the RDMA to the second resource space of the second storage device message; wherein, storing the target log by the second storage device includes: storing the target log in a second resource space of the second storage device according to the RDMA message.

After storing the target log in the second resource space, the second storage device may return a storage completion message to the first storage device. Compared with the second storage device receiving the target log and needing to select an appropriate resource space for the target log, the embodiment of the present application can further save the CPU occupation of the second storage device.

In some possible implementations, the obtaining, by the first storage device, the storage address of the second resource space and the storage capacity of the second resource space includes: the first storage device sends a resource request message to the second storage device, the resource The request message is used to request the second resource space for storing the target log; the first storage device receives a resource request response message sent by the second storage device according to the resource request message, and the resource request response message carries the second resource space and the storage capacity of the second resource space.

The first storage device may send a resource request message to the second storage device, and the second storage device selects any one resource space (represented as a second resource space) among multiple resource spaces, and sends a resource request response message to the first storage device , and the resource request response message carries the storage address of the second resource space and the remaining capacity of the second resource space, so that the first storage device can select an appropriate resource space for the target log. The first storage device may acquire the address and storage capacity of the resource space of the second storage device only once, until the resource space has no remaining resources, or the remaining resources cannot meet the size of the current data to be written, and then perform the next acquisition. 2. The address and storage capacity of the resource space of the storage device.

In some possible implementations, the method further includes: configuring, by the first storage device, a memory index for the target data; and determining, by the first storage device, a resource space of the first storage device storing the target data according to the memory index , wherein the storage address of the resource space of the first storage device corresponds to the memory index; the first storage device writes the target data into the resource space of the first storage device.

The first storage device configures a memory index for the target data, and applies for a storage resource (represented as the resource space of the first storage device) for the target data, and configures the mapping relationship between the storage address of the applied storage resource and the memory index, the mapping relationship It is used for the subsequent first storage device to be able to read the data stored in the corresponding storage address according to the memory index, and finally write the target data into the resource space corresponding to the storage address. In this way, the first storage device stores the target data in the resource space corresponding to the memory index, and does not prolong the time that affects the return of the write request response message.

In a second aspect, an embodiment of the present application provides a storage system, and the storage system can implement the functions performed by the first storage device and the second storage device in the methods involved in the above aspects. The functions can be implemented by hardware, or can be The corresponding software implementation is performed by hardware. The hardware or software includes one or more units or modules corresponding to the above functions.

In a third aspect, a storage system is provided, the structure of the storage system includes a memory, and the processor is configured to support the functions performed by the first storage device and the second storage device in the methods involved in the above aspects. The communication interface is used to support communication between the first storage device, the second storage device and other network elements. The memory is for coupling with the processor and holds necessary program instructions and data for the first storage device and the second storage device.

In a fourth aspect, a computer storage medium is provided, and program codes are stored in the computer storage medium, and the program codes are used to instruct computer software instructions for executing the above storage system, and the program codes include the programs designed for executing the above aspects.

Based on the above technical solution, the first storage device generates a target log including the target data and identification information for identifying the target data according to the target data in the write request message by receiving the write request message carrying the target data, and sends the target log to the target data. The second storage device sends the target log, the second storage device stores the target log, and does not need to configure a memory index to store the target log, so the first storage device generates a target log including identification information that identifies the target data, and the second storage device The target log can be stored without configuring a memory index, thereby reducing the occupation of the CPU of the second storage device.

Description of drawings

In order to illustrate the technical solutions of the embodiments of the present application more clearly, the following briefly introduces the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some of the drawings in the present application. In the embodiments, for those of ordinary skill in the art, other drawings can also be obtained according to these drawings without any creative effort.

FIG. 1 is a schematic structural diagram of a storage system according to an embodiment of the present application;

FIG. 2 is a schematic structural diagram of a storage node and a management node according to an embodiment of the present application;

FIG. 3 is a work flow diagram of a Cache module according to an embodiment of the present application;

Fig. 4a, Fig. 4b and Fig. 4c are schematic diagrams of log formats according to an embodiment of the present application;

5 is a schematic structural diagram of a resource space according to an embodiment of the present application;

6 is a schematic flow chart of a method for prior art data storage;

7 is a schematic flowchart of a method for data storage according to an embodiment of the present application;

8 is a schematic flowchart of a data storage method according to another embodiment of the present application;

9 is a schematic block diagram of a storage system according to an embodiment of the present application;

FIG. 10 is a schematic structural diagram of a storage system according to an embodiment of the present application.

Detailed ways

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

FIG. 1 shows a schematic structural diagram of a storage system according to an embodiment of the present application. As shown in Figure 1, the storage system includes nodes and a network. The nodes are divided into user nodes (there may be multiple user nodes: user node 200a, user node 200b, user node 200c and user node 200d), storage nodes (there may be multiple storage nodes: storage node 100a, storage node 100b, Storage node 100c and storage node 100d) and management node 300. The network is divided into a front-end network 400 for carrying user input/output (Input/Output, IO) data and a back-end network 500 for carrying service data. The service data in this embodiment of the present application may refer to mirror data supporting RDMA.

The user node is used to receive user IO data at the front end and forward it to the storage node through the front end network. The storage node is used to store user service data, and a cache memory (Cache) function is also on the storage node. The management node is responsible for maintaining the mirror relationship between storage nodes.

FIG. 2 shows a schematic structural diagram of a storage node and a management node. Among them, the storage node can distinguish the main storage node of IO (take 100a in FIG. 1 as an example) and the mirror storage node of IO (take 100b in FIG. 1 as an example). The IO master node 100a mainly includes a front-end IO module 110a, a back-end disk module 120a and a Cache module 130a. The front-end IO module 110a is used for user IO reception, IO flow control and IO error handling. The back-end disk module 120a is used to provide the read/write function of user service data through the disk. The Cache module 130a is divided into a log layer 131a, a resource layer 132a and a local storage 133a. The management node 300 includes a mirror management module 310, and the mirror management module 310 is used to maintain the mirror relationship between nodes on the management node, and notify the Cache module 130. The structure of the mirror storage node 100b is the same as that of the main storage node 100a, and details are not described herein again.

FIG. 3 shows the internal workflow of the Cache module 130a. The log layer 131a receives the IO data sent by the front-end IO module 110a, converts the IO data into log format (for example, data in a write-ahead logging (WAL log) format) and sends it to the resource layer 132a; the resource layer 132a is responsible for Space allocation of resources (eg, Non-Volatile Dual In-line Memory Module (NVDIMM) resources), and logs are sent to mirror node 110b for storage. The primary storage node 100a receives the storage complete message, and the local storage 133a stores the log locally.

To facilitate understanding of the embodiments of the present application, the following elements are first introduced before the embodiments of the present application are introduced here.

Remote Direct Memory Access (RDMA) directly transfers data to the storage area of the mirror node through the network without any impact on the operating system, thus reducing the overhead of the CPU on the mirror side. The specific implementation of RDMA is through the negotiation between the sender and the receiver on the location of the destination buffer. Different from the traditional TCP connection, some changes must be made to the traditional Transmission Control Protocol (TCP) header data structure to support RDMA. protocol, mainly by adding a field representing the destination address (ie, generating an RDMA message). RDMA read and write operations require both the source and destination to know the location and size of the destination buffer. If a piece of data is written from the host's memory to another host's memory through RDMA, the address of the destination buffer must be provided and ensure that this buffer is Large enough to receive this piece of data.

There are three types of WAL log log formats: simple log, compound log, and command log.

The simple log is only for the log of an object of metadata (that is, the information describing the attributes of the data, used to indicate the storage location, historical data, resource search, file recording, etc.), or the log of a file of data. When a write IO only contains the data of one Cache module, the simple log format is usually used. The simple log format is shown in Figure 4a, including magic word part, data part, log header part and log tail part. The magic word part stores a group of magic words to indicate the start of the log, metadata is stored in the log header, and Cyclic Redundancy Check (CRC) check information of the log is stored in the log tail.

The composite log format is shown in Figure 4b. When a write IO contains data from multiple Cache modules, it needs to be converted into multiple simple logs, and then multiple simple logs are encapsulated with a composite log header to form a composite log. To ensure the consistency of these multiple logs. The data part of this composite log stores multiple simple logs, and the specific log number and data length are stored in the log header.

The command type log format is shown in Figure 4c. When a write command or IO of other commands sends data, the data carried by the IO does not need to be stored in the Cache. In order for the mirror node to perceive it, a log needs to be written. , describing the location and size of the implementation data. The failure range information is described in the command-type log header.

Resource allocation at the resource layer refers to dividing the mirrored space into multiple segments (usually using 8M NVDIMM space) to manage. Each segment can store multiple WAL logs (as shown in Figure 5). The WAL logs are stored continuously, and each WAL log is connected end to end. The magic word part stores a set of magic words to indicate the start of a segment, and the log header records whether the segment status is available and the segment serial number and other description information.

Fig. 6 shows the method for data storage in the prior art, and the specific process is as follows:

101. The IO front-end module sends a write request message to the local cache module, where the write request message carries the target data to be written.

102. The local cache module receives the target data, and sends the target data to the mirror cache module, so that the mirror cache module stores the target data.

103. The mirror cache module receives the target data, and configures a memory index for the target data. The memory index is used to facilitate reading the target data. Specifically, the mirror cache module can first query an existing index table whether the target exists. The memory index corresponding to the data, if it does not exist, create a corresponding memory index for the target data;

104. The mirror cache module applies for storage resources for the target data according to the memory index (that is, configures a suitable resource space for storing the target data), and sets the mapping relationship between the memory index of the target data and the storage address of the applied resource space, so that according to The index and the storage address corresponding to the index can read data in the storage space.

105. The image storage module stores the target data in the requested resource space.

106. After the mirror cache module completes the storage of the target data, it sends a storage completion message to the local cache module.

107. After the local cache module receives the storage completion message sent by the mirror cache module, it can start to store the target data in the local cache, and firstly configure a memory index for the target data. The specific process is the same as that of the mirror node storing data (such as the above-mentioned steps 103 and 104 ), and details are not described here.

108. The local cache module applies for storage resources for the target data.

109. The local cache module stores the target data in the applied resource space.

110. After the local cache module completes the storage of the target data, it sends a write request response message to the IO front-end module.

Therefore, in the prior art, when the mirror cache module stores mirror data, it needs to configure a memory index for the target data and apply for storage resources (the mirror cache module needs to find a suitable resource space according to various attributes such as the size and type of the target data, and configure The mapping relationship between the storage address of the resource space and the memory index) and the steps of storing the target data in the mirror cache module all require the participation of the CPU, so that the CPU usage of the mirror storage node is relatively high.

The embodiments of the present application are applied to a scenario where the first storage device stores data and the second storage device stores data when the first storage device works normally.

It should be understood that the first storage device in this embodiment of the present application corresponds to the primary storage node 100a in FIG. 2, and the second storage device corresponds to the mirror storage node 100b in FIG. 2. In other words, the first storage device in the embodiment of the present application is The primary storage node 100a and the second storage device in the above-mentioned FIG. 2 are the mirror storage node 100b in the above-mentioned FIG. 2; or the first storage device and the second storage device are respectively a device having the function of the Cache module 130a and the function of the Cache module 130b. equipment, which is not limited in this application.

FIG. 7 shows a schematic flowchart of a method for data storage according to an embodiment of the present application. The first storage device in this embodiment of the present application may be any one of the storage nodes 100a, 100b, 100c, and 100d in FIG. 1 , and the second storage device may be the mirror storage of the primary storage node determined by the management node 300 node.

701. The first storage device receives a write request message, where the write request message carries target data.

The embodiments of the present application are applied to a storage system including a first storage device and at least one second storage device, and the second storage device is a mirror storage device of a type of first storage device. In addition, the first storage device may also be a mirror storage device of the second storage device, that is, the first storage device and the second storage device may be mirror storage devices of each other, which is not limited in this application.

The first storage device may be the Cache module 130a in FIG. 2. The Cache module 130a receives the write request message sent by the IO front-end module 110a, but the present application is not limited thereto. The first storage device receives a write request message, and the write request message carries data to be written (represented as target data), and the target data may be data that needs to be processed by the same cache, or data that needs to be processed by multiple cache units; Or the target data does not need to be kept in the cache.

It should be noted that the first storage device may include multiple memory areas, processors, controllers, and an operating system running on the controllers, etc. The second storage device may be the same as the first storage device, or may be different. This is not limited.

702. The first storage device generates a corresponding target log according to the target data.

The target log includes target data and identification information for identifying the target data, where the identification information is various information indicating attributes of the target data. For example, the log format is shown in Figure 4a, Figure 4b, and Figure 4c. Taking a simple log as an example, the simple log includes a magic word part, a data part, a log header part, and a log tail part. The storage device can identify the start of the log according to the magic word part; the metadata is stored in the log header, and the metadata is used to indicate data storage location, historical data, resource search, file recording and other functions; the log tail part stores the CRC check information of the log , to ensure the reliability of the data. Therefore, the identification information included in the target log in this embodiment of the present application may be all information stored in the log header, log tail, and magic word parts.

Log formats include simple log, compound log, and command log. If the target data requires one cache process, the first storage device converts the target data into a simple log; if the target data requires multiple cache processes, the first storage device converts the target data into a compound log. If the target data does not need to be stored in the cache, the first storage device converts the target data into a command-type log to describe the data failure and the location and size of the invalid data. In this way, the first storage device can select and generate an appropriate target log for the target data according to whether the target data needs one cache process, multiple cache processes, or no cache process.

It should be understood that the embodiments of the present application do not distinguish between "cache" and "memory", that is, both the cache and the memory can meet the data storage requirements of the embodiments of the present application, and therefore are not limited in the present application.

For example, the target data is block storage data that needs to be processed by a cache, and the block storage data mainly includes volume identity (volume Identity, volume ID), block start address (Logical block address, LBA), number of blocks (lengh), user data (buff). Since the target data needs to be processed by a cache, the first storage device converts the block storage data into a simple log, and converts the LBA into the number of the cache management unit, for example, dividing the LBA by 64k and rounding up to obtain the unit number (chunk ID) , Take the remainder to get the offset (offset) in the unit, write the volume ID, chunk ID, offset into the log header, write the buff into the log data part, and use the cyclic redundancy check code to ensure the reliability of the data (Cyclic Redundancy Check, CRC) is written at the end of the log.

It should be understood that "cache" and "memory" are not distinguished in the embodiments of the present application, that is, both the cache and the memory can meet the data storage requirements of the embodiments of the present application.

703. The first storage device sends the target log to the second storage device.

The second storage device receives the target log sent by the first storage device, and stores the target log in the memory of the second storage device.

704. The second storage device mirrors and stores the target log.

The second storage device receives the target log generated by the first storage device according to the target data carried in the write request message, and directly stores the target log in the second storage device, because the target log includes an identifier used to represent the target data In this way, the second storage device does not need to create a memory index when storing the target log. Compared with the prior art, in which the second storage device needs to configure a memory index, the present application can directly store the target log, thereby saving the Second, the CPU usage of the storage device when the memory index is configured.

Optionally, as an embodiment, the first storage device receives a storage completion message that the second storage device completes the mirrored storage of the target log, and ensures that the second storage device has successfully completed writing the target log before the first storage device starts to store the target log. The target log is written in its own cache, so that the failure of the second storage device can be avoided and the CPU usage of the first storage device is wasted. After the first storage device completes writing the target log and receives the storage completion message sent by the second storage device, it sends a write request response message to the IO device, so as to avoid the need to wait for the target data to be stored in the memory index corresponding to the prior art. Only after the resource space can return the write request response message to the IO device, the delay from issuing the write request to receiving the write request response message is reduced.

It should be understood that the device sending the write request message and receiving the write request response message may be an IO device or other devices. For the convenience of description, an IO device is used as an example for description, but the present application is not limited thereto.

Optionally, before the first storage device receives the storage completion message, the first storage device starts writing the target log, after the first storage device completes writing the target log, and receives the storage completion message sent by the second storage device. After the message, a write request response message is sent to the IO device, which can further save the time for returning the write request response message to the IO device.

Optionally, as an embodiment, the second storage device includes a plurality of resource spaces; wherein, storing the target log in a mirror image of the second storage device includes: the second storage device, according to the identification information, determines the storage space of the target log. A first resource space, the first resource space is any resource space in the plurality of resource spaces of the second storage device that meets the resource requirements of the target log; the second storage device stores the target log image in the first storage device in the resource space.

Specifically, the second storage device receives a target log, the target log includes identification information for identifying target data, the second storage device selects an appropriate resource space (represented as a first resource space) according to the identification information, and the target The log is stored in the first resource space of the second storage device. In addition, the second storage device can identify the target data according to the identification information, and can regularly store the target data in order in the form of the target log, so that the second storage device does not need to configure a memory index for the target data, thereby saving CPU usage of the second storage device.

It should be understood that the first resource space is the resource space mentioned above in units of segments, and each segment can store multiple logs.

Optionally, the second storage device includes a plurality of resource spaces; the method further includes: the first storage device obtains the storage address of the second resource space and the storage capacity of the second resource space, and the second resource space is the Any resource space in a plurality of resource spaces of the second storage device; the first storage device determines whether the second resource space meets the resource requirements of the target log according to the storage capacity of the second resource space and the identification information; the When the second resource space meets the resource requirement of the target log, the first storage device generates an RDMA message according to the RDMA encoding method, the target log and the storage address of the second resource space, and the RDMA message includes the second resource space The storage address and the target log; wherein, sending the target log by the first storage device to the second storage device includes: the first storage device sending the RDMA message to the second storage device; wherein, the second storage device Mirroring the storing of the target log includes: the second storage device storing the target log in the second resource space according to the RDMA message.

Specifically, the second storage device includes multiple resource spaces, and the second storage device can freely send the storage capacity and storage address of any resource space. The first storage device obtains the storage address and storage capacity of any one of the multiple resource spaces in the second storage device (represented as the second resource space), and according to the storage capacity of the second resource space and the identifier of the target log The information determines whether the second resource space meets the resource requirements of the target log. Correspondingly, the second resource space is also the above-mentioned Segment resource space. Since the identification information of the target log includes the resource size of the target data (that is, the resources required for storing the target log), the first storage device can determine the second storage device by comparing the resource size of the target data in the identification information with the storage capacity of the second resource space. Whether the resource space can meet the resource requirements of the target log.

When the second resource space meets the resource requirement of the target log, the first storage device generates an RDMA message according to the RDMA encoding method, the target log and the storage address of the second resource space, and the RDMA message includes the second resource The storage address of the space and the target log, the first storage device sends the RDMA message to the second storage device, and after the second resource space of the second storage device receives the RDMA message, it parses to obtain the target log, and directly stores the target log in in the second resource space. After storing the target log image in the second resource space, the second storage device may return an image storage completion message to the first storage device. Compared with the second storage device receiving the target log and needing to select an appropriate resource space for the target log, the embodiment of the present application can further save the CPU occupation of the second storage device.

For example, as shown in FIG. 2, the first storage device can directly send the target log from the resource layer 132a to the resource layer 132b through the RDMA technology.

Optionally, as an embodiment, the obtaining, by the first storage device, the storage address of the second resource space and the storage capacity of the second resource space includes: the first storage device sending a resource request message to the second storage device, the The resource request message is used to request the second resource space for storing the target log; the first storage device receives a resource request response message sent by the second storage device according to the resource request message, and the resource request response message carries the second resource The storage address of the space and the storage capacity of the second resource space.

Specifically, the first storage device may send a resource request to the second storage device to obtain the storage address of the resource space of the second storage device and the remaining resource size of the resource space (that is, how much can be written in the resource space). data). The first storage device may send a resource request message to the second storage device, and the second storage device selects any one resource space (represented as a second resource space) among multiple resource spaces, and sends a resource request response message to the first storage device , and the resource request response message carries the storage address of the second resource space and the remaining capacity of the second resource space, so that the first storage device can select an appropriate resource space for the target log.

It should be understood that, in this embodiment of the present application, the first storage device may acquire the address and storage capacity of the resource space of the second storage device only once, until the resource space has no remaining resources, or the remaining resources cannot satisfy the current data to be written. size, and then perform the next acquisition of the address and storage capacity of the resource space of the second storage device.

Optionally, as an embodiment, the method further includes: the first storage device configures a memory index for the target data; the first storage device determines, according to the memory index, a resource of the first storage device that stores the target data space, wherein the storage address of the resource space of the first storage device corresponds to the memory index; the first storage device writes the target data into the resource space of the first storage device.

The first storage device sends a write request response message to the IO device after completing the storage of the target log and receiving the storage completion message returned by the second storage device. Since it is faster to read the target data according to the memory index, after that, the first storage device may further configure a memory index for the target data, where the memory index is used to identify the target data (for example, the size and type of the target data, etc.), And apply for a storage resource (represented as the resource space of the first storage device, the first storage device may also include multiple storage spaces) for the target data according to the memory index, and configure the storage address of the applied storage resource and the memory index. The mapping relationship is used for the subsequent first storage device to be able to read the data stored in the corresponding storage address according to the memory index, and finally write the target data into the resource space corresponding to the storage address, so that the first storage device will The target data is stored in the resource space corresponding to the memory index, which will not prolong the time for returning the response message of the write request.

It should be understood that the target data may be the target data carried in the write request or the target data obtained by parsing the target log, which is not limited in this application.

It should also be understood that the first storage device may not store the target log, but directly write the target data into the resource space corresponding to the memory index; or, after the first storage device writes the target log, the target data is written into The write request response message is sent only after the resource space corresponding to the memory index is stored. This application does not limit this.

The second storage device stores the target log so that when the first storage device fails, the second storage device can take over the data storage currently served by the first storage device. The specific failure symptoms are as follows:

When the first storage device fails, after the second storage device detects the failure, the second storage device takes over data storage. At this time, the second storage device needs to perform target log playback. Log playback refers to parsing the target log to obtain valid target data (that is, reading the target data one by one), configuring a memory index for the target data, and storing the target data in the corresponding In the resource space (that is, indicating the end of log playback), the resource space is the storage address corresponding to the memory index, and the resource space corresponding to the storage address; after that, the second storage device can process the data storage service.

After the first storage device fails to recover quickly, the first storage device can directly restart the playback log, and then the first storage device can process the data storage service.

When both the first storage device and the second storage device are faulty, after restarting, the first storage device can compete with the second storage device for the primary storage device (ie, take over the current data storage), and after competing with the primary storage device, the log can be played back. . After that, the second storage device can process the data storage service, without causing the loss of memory data, and improving the reliability of data storage. .

It should be understood that, in various embodiments of the present application, the size of the sequence numbers of the above-mentioned processes does not mean the sequence of execution, and the execution sequence of each process should be determined by its functions and internal logic, and should not be dealt with in the embodiments of the present application. implementation constitutes any limitation.

Therefore, in the data storage method according to the embodiment of the present application, the first storage device generates, according to the target data in the write request message, the target data and the identification information for identifying the target data by receiving the write request message carrying the target data. The target log is sent to the second storage device, the second storage device mirrors the target log, and the mirrored storage of the target log does not need to configure a memory index, so the first storage device generates an identifier that includes identifying the target data. For the target log of information, the second storage device can store the target log without configuring a memory index, thereby reducing the occupation of the CPU of the second storage device.

FIG. 8 shows an interaction flowchart of a method for data storage according to an embodiment of the present application. The meanings of various terms in the embodiments of the present application are the same as those in the foregoing embodiments.

It should be noted that this is only to help those skilled in the art to better understand the embodiments of the present application, rather than limiting the scope of the embodiments of the present application.

801. A first storage device receives a write request message, where the write request message carries target data.

802. The first storage device generates a corresponding target log according to the target data, where the target log includes the target data and identification information for identifying the target data.

803. The first storage device acquires the storage address of the resource space (represented as the second resource space) and the storage capacity of the resource space of the second storage device.

804. The first storage device determines, according to the storage capacity and identification information of the second resource space, whether the acquired storage capacity of the second resource space can meet the resource requirements of the target log.

805. After determining that the acquired storage capacity of the second resource space can meet the resource requirements of the target log, the first storage device generates an RDMA message according to the target log, the RDMA encoding method, and the storage address of the second resource space.

806. The first storage device sends an RDMA message to the second storage device.

807. The second storage device directly stores the target log in the second resource space corresponding to the storage address according to the RDMA message.

Optionally, the first storage device may directly send the target log to the second storage device, and the second storage device determines a resource space capable of storing the target log, so steps 803 to 807 may not be required.

808. After completing the mirror storage of the target log, the second storage device sends a storage completion message to the first storage device.

809. The first storage device locally stores the target log.

Step 808 and step 809 may not have a sequential relationship, which is not limited in this application.

810. The first storage device completes the storage of the target log locally, and after receiving the storage completion message, returns a write request response message to the IO device.

811. The first storage device locally configures a corresponding memory index for the target data, where the memory index is used to identify the target data.

812. The first storage device determines, according to the memory index, the resource space of the first storage device that stores the target data.

813. The first storage device stores the target data in the determined resource space of the first storage device.

It should be understood that for the specific indication manner of the above-mentioned corresponding information, reference may be made to the foregoing embodiments, which are not repeated here for brevity.

Therefore, in the data storage method according to the embodiment of the present application, the first storage device generates, according to the target data in the write request message, the target data and the identification information for identifying the target data by receiving the write request message carrying the target data. The target log, obtain the second resource space of the second storage device, when the second resource space meets the resource requirements of the target log, generate an RDMA message according to the target log, the RDMA encoding method and the storage address of the second resource space, and send the message to the target log. The second storage device sends the target log, the second storage device directly stores the target log in the second resource space according to the RDMA message, and does not need to configure a memory index to store the target log, so that the first storage device generates the target log including the identification target data. The target log of the identification information, and the target log is directly stored in the storage space that meets the resource requirements by sending an RDMA message, and the second storage device does not need to configure the memory index to store the target log, thereby reducing the CPU of the second storage device. occupied.

It should be understood that, in various embodiments of the present application, the size of the sequence numbers of the above-mentioned processes does not mean the sequence of execution, and the execution sequence of each process should be determined by its functions and internal logic, and should not be dealt with in the embodiments of the present application. implementation constitutes any limitation.

The data storage method according to the embodiment of the present application is described in detail above, and the storage device according to the embodiment of the present application will be described below.

FIG. 9 shows a schematic block diagram of a storage system 900 according to an embodiment of the present application, where the storage system includes a first storage device 910 and at least one second storage device 920, where the second storage device 920 is the first storage device A mirrored storage device of the device 910, the storage system 900 is configured to execute the data storage method described in FIG. 7, and the storage system 900 includes:

The first storage device 910 is configured to receive a write request message, where the write request message carries target data, generate a target log according to the target data, and send the target log to the second storage device 920, where the target log includes the target data and identification information used to identify the target data;

The second storage device 920 is configured to store the target log, and the storage of the target log does not require configuring a memory index.

Therefore, in the storage system of the embodiment of the present application, the first storage device generates a target including the target data and identification information for identifying the target data according to the target data in the write request message by receiving the write request message carrying the target data. log, and send the target log to the second storage device, the second storage device mirrors the target log, and the mirrored storage of the target log does not need to configure a memory index, so that the first storage device generates the identification information that includes the target data. For the target log, the second storage device can store the target log without configuring a memory index, thereby reducing the occupation of the CPU of the second storage device.

Optionally, in an embodiment of the present application, the second storage device 920 is further configured to send a storage completion message to the first storage device 910 when the target log is stored;

The first storage device 910 is further configured to store the target log, and send a write request response message after completing the storage of the target log and receiving the storage completion message.

Optionally, in an embodiment of the present application, the second storage device 920 includes multiple resource spaces;

The second storage device 920 is specifically used for:

According to the identification information, determine the first resource space for storing the target log, where the first resource space is any resource space that satisfies the resource requirements of the target log in the plurality of resource spaces of the second storage device 920;

The target log is stored in the first resource space.

Optionally, in an embodiment of the present application, the second storage device 920 includes multiple resource spaces;

The first storage device 910 is further configured to obtain the storage address of the second resource space and the storage capacity of the second resource space, and determine whether the second resource space satisfies the requirements according to the storage capacity of the second resource space and the identification information. The resource requirement of the target log, and when the second resource space meets the resource requirement of the target log, an RDMA message is generated according to the remote direct data access RDMA encoding method, the target log and the storage address of the second resource space, and an RDMA message is sent to the target log. The second resource space of the second storage device 920 sends the RDMA message, the second resource space is any resource space among multiple resource spaces of the second storage device 920, and the RDMA message includes the storage address of the second resource space and the target log;

The second storage device 920 is further configured to store the target log according to the RDMA message through the second resource space.

Optionally, in an embodiment of the present application, the first storage device 910 is further configured to send a resource request message to the second storage device 920, and receive a resource sent by the second storage device 920 according to the resource request message A request response message, where the resource request message is used to request the second resource space for storing the target log, and the resource request response message carries the storage address of the second resource space and the storage capacity of the second resource space.

Optionally, in an embodiment of the present application, the first storage device 910 is further configured to configure a memory index for the target data, and determine the resource space of the first storage device that stores the target data according to the memory index, and writing the target data into the resource space of the first storage device, wherein the storage address of the resource space of the first storage device corresponds to the memory index.

Therefore, in the storage system of the embodiment of the present application, the first storage device generates a target including the target data and identification information for identifying the target data according to the target data in the write request message by receiving the write request message carrying the target data. log, and send the target log to the second storage device, the second storage device mirrors the target log, and the mirrored storage of the target log does not need to configure a memory index, so that the first storage device generates the identification information that includes the target data. For the target log, the second storage device can store the target log without configuring a memory index, thereby reducing the occupation of the CPU of the second storage device.

The storage system 900 according to the embodiment of the present application may correspond to the storage system according to the data storage method of FIG. 7 or FIG. 8 according to the embodiment of the present application, and the above-mentioned and other operations and/or functions of each module in the storage system 900 are for the purpose of For the sake of brevity, the corresponding processes for implementing the foregoing methods are not repeated here.

10 shows a structure of a storage system provided by an embodiment of the present application, including at least one processor 1002 (eg, CPU), at least one network interface 1005 or other communication interface, at least one memory 1006, and at least one communication bus 1003, Used to realize the connection communication between these devices. The processor 1002 is used to execute executable modules, such as computer programs, stored in the memory 1006 . The memory 1006 may include a high-speed random access memory (Random Access Memory, RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory. The communication connection with at least one other network element is realized through at least one network interface 1005 (which may be wired or wireless).

In some embodiments, the memory 1006 stores a program 10061, and the processor 1002 executes the program 10061 for performing the following operations:

Receive a write request message through the network interface 1005, the write request message carrying the target data;

According to the target data, a target log is generated, and the target log includes the target data and identification information for identifying the target data;

Send the target log to the second storage device through the network interface 1005;

The target log is stored, and the storage of the target log does not require configuring a memory index.

It should be noted that the storage system may be used to execute each step and/or process in the method embodiment shown in FIG. 7 or FIG. 8 above.

It can be seen from the above technical solutions provided by the embodiments of the present application that, by receiving the write request message carrying the target data, the first storage device generates, according to the target data in the write request message, a message including the target data and an identifier for identifying the target data. Identify the target log of the information, and send the target log to the second storage device. The second storage device mirrors and stores the target log, and the mirrored storage of the target log does not need to configure a memory index, so that the first storage device generates the target data including the identification target. The target log of the identification information, the second storage device can store the target log without configuring a memory index, thereby reducing the occupation of the CPU of the second storage device.

Embodiments of the present application further provide a computer storage medium, where the computer storage medium can store program instructions for instructing any of the foregoing methods.

Optionally, the storage medium may specifically be the memory 1106 .

It should be understood that the specific examples in the present application are only for helping those skilled in the art to better understand the embodiments of the present application, rather than limiting the scope of the embodiments of the present application.

It should be understood that the term "and/or" in this document is only an association relationship to describe associated objects, indicating that there can be three kinds of relationships, for example, A and/or B, which can mean that A exists alone, and A and B exist at the same time , there are three cases of B alone. In addition, the character "/" in this document generally indicates that the related objects are an "or" relationship.

It should be understood that, in various embodiments of the present application, the size of the sequence numbers of the above-mentioned processes does not mean the sequence of execution, and the execution sequence of each process should be determined by its functions and internal logic, and should not be dealt with in the embodiments of the present application. implementation constitutes any limitation.

Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Skilled artisans may implement the described functionality using different methods for each particular application, but such implementations should not be considered beyond the scope of this application.

Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described systems, devices and units may refer to the corresponding processes in the foregoing method embodiments, which will not be repeated here.

In the several embodiments provided in this application, it should be understood that the disclosed system, apparatus and method may be implemented in other manners. For example, the apparatus embodiments described above are only illustrative. For example, the division of the unit is only a logical function division. In actual implementation, there may be other division methods, for example, multiple units or components may be combined or Integration into another system, or some features can be ignored, or not implemented. On the other hand, the shown or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, indirect coupling or communication connection of devices or units, and may be in electrical, mechanical or other forms.

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

In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit. The above-mentioned integrated units may be implemented in the form of hardware, or may be implemented in the form of software functional units.

The integrated unit, if implemented in the form of a software functional unit and sold or used as an independent product, may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be embodied in the form of a software product in essence, or the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including Several instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, Read-Only Memory (ROM, Read-Only Memory), Random Access Memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program codes .

The above are only specific embodiments of the present application, but the protection scope of the present application is not limited to this. should be covered within the scope of protection of this application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims (10)

1. A method for storing data is applied to a storage system comprising a first storage device and at least one second storage device, wherein the second storage device is a mirror storage device of the first storage device, and the method comprises the following steps:

a first storage device receives a write request message, wherein the write request message carries target data;

the first storage device generates a target log according to the target data, wherein the target log comprises the target data and identification information for identifying the target data;

the first storage device sends the target log to the second storage device;

the second storage device stores the target log, and the target log is stored without configuring a memory index;

the method further comprises the following steps:

when the second storage device finishes storing the target log, sending a storage finishing message to the first storage device;

the first storage device stores the target log;

and the first storage equipment sends a write request response message after finishing the storage of the target log and receiving the storage finishing message.

2. The method of claim 1, wherein the second storage device comprises a plurality of resource spaces;

wherein the second storage device storing the target log comprises:

the second storage device determines a first resource space for storing the target log according to the identification information, wherein the first resource space is any one of a plurality of resource spaces of the second storage device which meets the resource requirement of the target log;

the second storage device stores the target log in the first resource space.

3. The method of claim 1, wherein the second storage device comprises a plurality of resource spaces;

the method further comprises the following steps:

the first storage device acquires a storage address of a second resource space and a storage capacity of the second resource space, wherein the second resource space is any one of a plurality of resource spaces of the second storage device;

the first storage device determines whether the second resource space meets the resource requirement of the target log according to the storage capacity of the second resource space and the identification information;

when the second resource space meets the resource requirement of the target log, the first storage device generates an RDMA message according to a remote direct data access (RDMA) coding mode, the target log and a storage address of the second resource space, wherein the RDMA message comprises the storage address of the second resource space and the target log;

wherein sending, by the first storage device, the target log to the second storage device comprises:

the first storage device sending the RDMA message to a second resource space of the second storage device;

wherein the second storage device storing the target log comprises:

a second resource space of the second storage device stores the target log according to the RDMA message.

4. The method of claim 3, wherein the first storage device obtaining the storage address of the second resource space and the storage capacity of the second resource space comprises:

the first storage device sends a resource request message to the second storage device, wherein the resource request message is used for requesting to store the second resource space of the target log;

and the first storage device receives a resource request response message sent by the second storage device according to the resource request message, wherein the resource request response message carries a storage address of the second resource space and a storage capacity of the second resource space.

5. The method according to any one of claims 1 to 4, further comprising:

the first storage device configures a memory index for the target data;

the first storage device determines a resource space of the first storage device for storing the target data according to the memory index, wherein a storage address of the resource space of the first storage device corresponds to the memory index;

the first storage device writes the target data into a resource space of the first storage device.

6. A storage system is characterized by comprising a first storage device and at least one second storage device, wherein the second storage device is a mirror image storage device of the first storage device;

the first storage device is configured to receive a write request message, where the write request message carries target data, generate a target log according to the target data, and send the target log to the second storage device, where the target log includes the target data and identification information used to identify the target data;

the second storage device is used for storing the target log, and the target log is stored without configuring a memory index;

the second storage device is further used for sending a storage completion message to the first storage device when the target log is stored;

the first storage device is further configured to store the target log, and send a write request response message after the storage of the target log is completed and the storage completion message is received.

7. The storage system of claim 6, wherein the second storage device comprises a plurality of resource spaces;

the second storage device is specifically configured to:

determining a first resource space for storing the target log according to the identification information, wherein the first resource space is any one of a plurality of resource spaces of the second storage device which meets the resource requirement of the target log;

storing the target log into the first resource space.

8. The storage system of claim 6, wherein the second storage device comprises a plurality of resource spaces;

the first storage device is further configured to obtain a storage address of a second resource space and a storage capacity of the second resource space, determine whether the second resource space meets the resource requirement of the target log according to the storage capacity of the second resource space and the identification information, generate an RDMA message according to an RDMA encoding manner, the target log and the storage address of the second resource space when the second resource space meets the resource requirement of the target log, and send the RDMA message to the second resource space of the second storage device, where the RDMA message is any one of the plurality of resource spaces of the second storage device, and the RDMA message includes the storage address of the second resource space and the target log;

the second storage device is further configured to store the target log according to the RDMA message through the second resource space.

9. The storage system according to claim 8, wherein the first storage device is further configured to send a resource request message to the second storage device, and receive a resource request response message sent by the second storage device according to the resource request message, where the resource request message is used to request to store the second resource space of the target log, and the resource request response message carries a storage address of the second resource space and a storage capacity of the second resource space.

10. The storage system according to any one of claims 6 to 9, wherein the first storage device is further configured to configure a memory index for the target data, determine a resource space of the first storage device storing the target data according to the memory index, and write the target data into the resource space of the first storage device, wherein a storage address of the resource space of the first storage device corresponds to the memory index.

Images