CN109144406B - Metadata storage method, system and storage medium in distributed storage system - Google Patents

Abstract

In the distributed storage system, under the scene that data reliability is realized by metadata stripes formed by an EC algorithm, other metadata blocks in the metadata stripes are backed up by a main data storage node, and because the metadata blocks on the data storage nodes are only required to be backed up on the main data storage node, compared with the prior art that all the metadata blocks are multiple copies, the storage space is reduced, and simultaneously, when a client accesses the metadata, all the metadata blocks are only required to be accessed from the main data storage node, so that the metadata access speed is improved.

Description

Metadata storage method, system and storage medium in distributed storage system

technical field

The present invention relates to the technical field of data storage, and in particular, to a metadata storage method, system and storage medium in a distributed storage system.

Background technique

In a distributed storage system, after the management node stores user data in the storage node, metadata such as the logical address and physical address of the data will be generated, and the metadata should also be stored in the storage node. A common metadata storage method is to scatter the blocks in the metadata stripe to each storage node. When reading the metadata, it is necessary to read the blocks in the metadata stripe from each storage node and piece them together into a metadata stripe. , but the amount of data forwarding between storage nodes is large, which affects performance. Another way is to store metadata in the form of multiple copies on the storage node, but it will increase the storage space overhead.

SUMMARY OF THE INVENTION

In a first aspect, an embodiment of the present invention provides a metadata storage solution in a distributed storage system. The distributed storage system includes a management node and (M+N) storage nodes, the management node and (M+N) storage nodes. +N) storage nodes all store the partition view of metadata stripe; the partition view of metadata stripe includes main data storage node DS _A , data storage node DS _i and check storage node CS _r ; wherein, N is a natural number not less than 2, M is a natural number not less than 1, A is one of the natural numbers 1 to N, i is each of the natural numbers 1 to N except A, and r is each of the natural numbers 1 to M In the storage scheme: the management node determines the primary data storage node DS _A , the data storage node DS _i and the check storage node CS _r for the metadata strip according to the partition view of the metadata strip; The metadata strip includes metadata blocks D _A , D _i and check blocks C _r , D _i is sent to the data storage node DS _i , D _A is sent to the main data storage node DS _A , and Cr is sent to the check storage node CS _r ; the check storage node CS _r receives and stores Cr _; the data storage node DS _i receives and stores D _i , and splits the partition view according to the metadata Send D _i to the main data storage node DS _A ; the main data storage node DS _A receives and stores D _A and D _i . In this solution, under the protection mechanism of Erasure Coding (EC) for metadata, the primary data storage node DS _A backs up other metadata blocks D _i in the metadata stripe, because only the data storage node DS _i needs to be The metadata block D _i on the device is backed up on the primary data storage node DS _A. Compared with the multiple copies of all metadata blocks in the prior art, there is no need to verify the copy of the block, which reduces the storage space. At the same time, when the client accesses the metadata , all metadata blocks can be accessed from the main data storage node DS _A , which improves the metadata access speed. The distributed storage system of this solution may be a distributed file system, a distributed object storage system or a distributed block device storage.

Optionally, the management node determines the primary data storage node DS _A , the data storage node DS _i and the check storage node CS _r for the metadata strip according to the partition view of the metadata strip, specifically including: The management node determines the partition corresponding to the metadata strip according to the write request for generating the metadata in the metadata strip; the management node queries the metadata strip according to the partition corresponding to the metadata strip The partition view of the primary data storage node DS _A , the data storage node DS _i and the check storage node CS _r are determined.

Optionally, the management node determines the partition corresponding to the metadata strip according to the address carried in the write request.

Optionally, the storage of Cr by the check storage node CS _r specifically includes: the check storage node CS _r allocates a slice S _r to the Cr, and establishes a relationship between the identifier of the Cr and the slice S _r . Mapping relationship; the storage of _{Di by the data storage node DS i specifically} includes: the data storage node DS _i allocates a fragment SD _i to the _Di , and establishes a relationship between the identifier of the _Di and the fragment SD _i A mapping relationship; the main data storage node DS _A stores D _A and D _i , specifically including: the main data storage node DS _A allocates a slice SD _A for the D _A , and establishes the identity of the D _A with the The mapping relationship of the segment SD _A , the segment SD _i is allocated to the D _i , and the mapping relationship between the identifier of the D _i and the segment SD _i is established.

Further, the management node establishes a mapping relationship between the identifier of D _i and the data storage node DS _i and the main data storage node DS _A. When garbage collection is performed on the metadata stripe, the management node can recycle the data of the metadata block in the data storage node and the main data storage node according to the mapping relationship between the identifier of the metadata block in the metadata stripe and the storage node , which improves the efficiency of metadata recovery.

In the second aspect, correspondingly, an embodiment of the present invention further provides a distributed storage system, the distributed storage system includes a management node and (M+N) storage nodes, the management node and (M+N) storage nodes N) storage nodes all store partition views of metadata stripes; the partition views of metadata stripes include main data storage node DS _A , data storage node DS _i and check storage node CS _r ; wherein, N is A natural number not less than 2, M is a natural number not less than 1, A is one of the natural numbers 1 to N, i is each of the natural numbers 1 to N except A, and r is each of the natural numbers 1 to M; The distributed storage system is used to implement various implementation solutions of the first aspect.

Correspondingly, the present invention also provides a non-volatile computer-readable storage medium and a computer program product, when the memory of the storage device provided by the embodiment of the present invention loads the non-volatile computer-readable storage medium and the computer program product with the Computer program instructions, the computer program instructions can run in a distributed storage system, the distributed storage system includes a management node and (M+N) storage nodes, and the management node and (M+N) storage nodes both store There is a partition view of metadata stripe; the partition view of the metadata stripe includes the main data storage node DS _A , the data storage node DS _i and the check storage node CS _r ; wherein, N is a natural number not less than 2, M is a natural number not less than 1, A is one of the natural numbers 1 to N, i is each of the natural numbers 1 to N except A, and r is each of the natural numbers 1 to M; when one or more computers execute The computer program instructions are respectively used as a management node, a main data storage node DS _A , a data storage node DS _i and a check storage node CS _r in the distributed storage system to implement various implementation schemes of the first aspect.

The metadata storage solutions in the various distributed storage systems disclosed in the first aspect may also be applicable to the storage of data corresponding to the metadata. Correspondingly, the distributed storage system of the second aspect and the non-volatile computer-readable storage medium and computer program product of the third aspect are also suitable for data storage.

Description of drawings

In order to illustrate the technical solutions in the embodiments of the present invention more clearly, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments.

1 is a schematic diagram of a distributed block device storage architecture according to an embodiment of the present invention;

2 is a schematic structural diagram of a server in a distributed block device according to an embodiment of the present invention;

3 is a schematic diagram of a relationship between a data stripe and a partition view according to an embodiment of the present invention;

4 is a schematic diagram of a data striping provided by an embodiment of the present invention;

5 is a schematic diagram of a partition view according to an embodiment of the present invention;

6 is a schematic diagram of the relationship between a metadata stripe and a partition view according to an embodiment of the present invention;

7 is a flowchart of metadata storage according to an embodiment of the present invention;

FIG. 8 is a schematic diagram of a metadata striping provided by an embodiment of the present invention;

FIG. 9 is a schematic diagram of metadata storage provided by an embodiment of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly described below with reference to the accompanying drawings in the embodiments of the present invention.

的

系列产品。本发明实施例以分布式块设备存储为例进行说明。示例性的如图1所示，分布式块设备存储包括多台服务器1、服务器2、服务器3、服务器4、服务器5和服务器6，服务器间互相通信。在实际应用当中，分布式块设备存储中服务器的数量可以根据实际需求增加，本发明实施例对此不作限定。分布式块设备存储的服务器中包含如图2所示的结构。Distributed storage systems mainly include distributed file system storage, distributed object storage, and distributed block device storage. For example, Huawei

of

series of products. The embodiments of the present invention take distributed block device storage as an example for description. Exemplarily as shown in FIG. 1 , the distributed block device storage includes multiple servers 1, 2, 3, 4, 5, and 6, and the servers communicate with each other. In practical applications, the number of servers in the distributed block device storage may be increased according to actual requirements, which is not limited in this embodiment of the present invention. The distributed block device storage server includes the structure shown in Figure 2.

As shown in FIG. 2, each server in the distributed block device storage includes a central processing unit (CPU) 201, a memory 202, a hard disk 1, a hard disk 2, and a hard disk 3. The memory 202 stores computer instructions, and the CPU 201 executes the memory. The program instructions in 202 perform corresponding operations. The hard disk may be at least one of a mechanical hard disk and a solid-state hard disk. In addition, in order to save the computing resources of the CPU 201, a Field Programmable Gate Array (FPGA) or other hardware can also be used for the above-mentioned corresponding operations of the CPU 201, or the FPGA or other hardware and the CPU 201 can jointly complete the above-mentioned corresponding operations. For convenience of description, the embodiments of the present invention are collectively described as a processor configured to implement the foregoing corresponding operations.

In the structure shown in FIG. 2 , the application program is loaded in the memory 202 , the CPU 201 executes the application program instructions in the memory 202 , and the server acts as a client. The application program may be a virtual machine (Virtual Machine, VM), or may be a specific application, such as office software. Clients write data to or read data from distributed block device storage. The storage management program is loaded in the memory 202, and the CPU 201 executes the storage management program instruction as the virtual block storage management program in the memory 202, then the server, as a management node, is responsible for the management of the volume metadata, and is used to provide the client with a block protocol access interface, Provides distributed storage access point services for clients, enabling clients to access storage resources stored by distributed block devices through management nodes. The storage object program is loaded in the memory 202, the CPU 201 executes the storage object program instruction in the memory 202, and the server acts as a storage node for performing specific input/output (I/O) operations. Each server can run multiple storage object program processes. Exemplarily, a hard disk corresponds to running one storage object program process by default. Each storage object program process will manage a hard disk, and the server runs each storage object program. A process acts as a storage node. For specific implementation, it is also possible that a process of running a stored object program on one server corresponds to all hard disks on the server. The embodiment of the present invention is described by taking a storage object program process managing a hard disk as an example. When the distributed block device storage is initialized, the process of each storage object program will manage the hard disk by 1MB as a unit, and record the allocation information of each 1MB fragment in the metadata management area of the hard disk. Storage resource pool. The storage management program communicates point-to-point with the processes of all storage object programs in the resource pool that it can access, that is, the management node communicates with all storage nodes in the resource pool that it can access, so that the management node can concurrently access all hard disks in the resource pool.

When the distributed block device storage is initialized, the hash space (such as 0-2^32,) is divided into N equal parts, each equal part is a partition (Partition), and the N equal parts are divided equally according to the number of hard disks . For example, in distributed block storage device storage, N is 3600 by default, that is, the partitions are P1, P2, P3...P3600. As shown in Figure 3, assuming that the current distributed block device stores 18 hard disks (storage nodes), each storage node carries 200 partitions. The above-mentioned correspondence between partitions and storage nodes, that is, the partition view, will be allocated when the distributed block device storage is initialized, and will be adjusted later as the number of hard disks in the distributed block device storage changes. The distributed block device storage server will save the partition view in the memory 202, and the management node uses the partition view for fast routing. Each storage node also stores all partition views of the distributed block device storage system, that is, the correspondence between each partition and the storage node. At the same time, according to the reliability requirements of distributed block device storage, the Erasure Coding (EC) algorithm can be used to improve data reliability. bar, as shown in Figure 4, the partition view is "partition-main data storage node-data storage node 1-data storage node 2-check storage node, exemplary, the partition view is shown in Figure 5. The partition view Indicates that the partition corresponds to the main data node and the data storage node 1 and data node 2 for storing other data blocks of data striping, and the check storage node for storing the check data, which is stored in the data storage node 1 and data storage node 2. The backup data storage node of the data block is the primary data storage node.

Distributed block device storage will logically slice each logical unit number (LUN) according to 1MB size. For example, a 1GB LUN will be sliced into 1024*1MB slices. As shown in Figure 3, when the client sends a write request to the LUN through the management node, the Small Computer System Interface (SCSI) command will carry the LUN identifier (Identifier, ID), the logical block address (Logical BlockAddress, LBA) ID and the data to be written, the management node where the client is located receives the write request, and forms a key according to the LUN ID and LBA ID. The key will contain the LBA ID to 1MB rounding calculation information. An integer (in the range of 0-2^32) is calculated by Distributed Hash Table (DHT) Hash and falls in the specified partition; the management node where the client is located is based on the partition view recorded in the memory 202 Determine the main data storage node, data storage node 1, data storage node 2, and check storage node, and the management node sends the data block 1, data block 2, data block 3, and check block 4 in the EC data stripe to the main data storage node. Data storage node 1 , data storage node 2 , data storage node 3 and check storage node 4 . The main data storage node stores data block 1, the data storage node 1 stores data block 2, the data storage node 2 stores data block 3, and the check storage node stores check block 1. Data storage nodes 1 and 2 respectively determine the main data storage node according to the partition view, data storage node 1 backs up data block 2 to the main data storage node, data storage node 2 backs up data block 3 to the main data storage node, and the main data storage node Data block 2 and data block 3 are stored respectively. In the specific implementation, the main data storage node allocates slice 1 for data block 1 from the hard disk it manages, and establishes a mapping relationship between the identifier of data block 1 and slice 1; the data storage node 1 is the data block from the hard disk managed by it. 2. Allocate fragment 2, establish a mapping relationship between the identifier of data block 2 and fragment 2; data storage node 2 allocates fragment 3 for data block 3 from the hard disk it manages, and establishes the identifier of data block 3 and fragment 3. Mapping relationship; the verification storage node allocates fragment 4 to verification block 1 from the hard disk it manages, and establishes a mapping relationship between the identification of verification block 1 and fragment 4. The main data storage node receives data block 2 sent by data storage node 1 and data block 3 sent by data storage node 2. The main data storage node allocates slices 5 and 6 from the hard disks it manages, and the main data storage node creates data The mapping relationship between the identifier of block 2 and slice 5, and the mapping relation between the identifier of data block 3 and slice 6. In the embodiment of the present invention, taking the mapping relationship between the identifier of the data block and the fragment as an example, when one process of the storage object program corresponds to one hard disk, that is, the storage node is the hard disk itself, the identifier of the data block and the partition The mapping relationship of the slice is the mapping relation between the identifier of the data block and the physical address of the slice; when one process of the storage object program corresponds to multiple hard disks, that is, the storage node manages multiple hard disks, the identifier of the data block is the same as that of the slice. The mapping relationship includes the mapping between the identifier of the data block and the hard disk storing the data block, and the mapping from the hard disk storing the data block to the slice. The mapping relationship of the physical address of the slice is when. Further, data block 2 is stored in fragment 2 and fragment 5 respectively, data block 3 is stored in fragment 3 and 6 respectively, and the management node establishes and saves the identification of data block 2 and the data storage node 1 and the main data storage node. Mapping, establishing and saving the mapping relationship between the identifier of the data block 3 and the data storage node 2 and the main data storage node. Further, the data storage node 1 saves the mapping between the identification of the data block 2 and the data storage node 1 and the main data storage node, and the data storage node 2 saves the mapping relationship between the identification of the data block 3 and the data storage node 2 and the main data storage node. . When garbage collection is performed on data stripes, the management node can recycle the data of the data blocks in both the data storage node and the main data storage node according to the mapping relationship between the identifiers of the data blocks in the data stripes and the storage nodes, which improves the data storage efficiency. recycling efficiency

In the embodiment of the present invention, when the client sends a write request to the distributed block device storage to write data, metadata is generated for recording the logical address and physical address of the data, and the like. In this embodiment of the present invention, the metadata storage corresponding to the data uses the same EC algorithm as the data storage. The metadata stripe composed based on the EC algorithm has the same partition view as the above-mentioned composition data stripe based on the EC algorithm, as shown in FIG. 6 .

Metadata is stored in a distributed storage system, wherein the distributed storage system includes a management node and (M+N) storage nodes, and both the management node and (M+N) storage nodes store partition views of metadata stripes; The partition view of data striping includes the main data storage node DS _A , the data storage node DS _i and the check storage node CS _r ; wherein, N is a natural number not less than 2, M is a natural number not less than 1, and A is a natural number 1 to One of N, i is each of the natural numbers 1 to N except A, and r is each of the natural numbers 1 to M; the distributed storage system stores and executes the process shown in Figure 7:

Step 701: The management node determines the primary data storage node DS _A , the data storage node DS _i and the check storage node CS _r for the metadata strip according to the partition view of the metadata strip; the metadata strip includes Metadata blocks D _A , D _i and check block Cr.

Specifically, the management node determines the primary data storage node DS _A , the data storage node DS _i and the check storage node CS _r for the metadata strip according to the partition view of the metadata strip, which specifically includes: the The management node determines the partition corresponding to the metadata strip according to the write request for generating the metadata in the metadata strip; the management node queries the metadata strip according to the partition corresponding to the metadata strip. The partition view determines the primary data storage node DS _A , the data storage node DS _i and the check storage node CS _r .

Specifically, the management node determines the partition corresponding to the metadata strip according to the address carried in the write request. For details, please refer to the solution when the distributed block device stores the write request sent by the client, which will not be repeated here.

Step 702: The management node sends _{Di to the data storage node DS i ,} sends DA to the primary data storage node DS _A , and _{sends Cr} to the check storage node CS _r .

Step 703: The verification storage node CS _r receives and stores C _r .

Step 704: The data storage node DS _i receives and stores D _i , and sends D _i to the primary data storage node DS _A according to the partition view of the metadata stripe.

Step 705: The primary data storage node DS _A receives and stores D _A and D _i .

Specifically, the storage of Cr by the check storage node CS _r specifically includes: the check storage node CS _r allocates a slice S _r to the Cr, and establishes a mapping between the identifier of the Cr and the slice S _r The storage of D _i by the data storage node DS _i specifically includes: the data storage node DS _i allocates a segment SD _i to the D _i , and establishes a mapping between the identifier of the D _i and the segment SD _i The main data storage node DS _A stores D _A and D _i , specifically including: the main data storage node DS _A allocates a slice SD _A for the D _A , and establishes the identity of the D _A with the For the mapping relationship of the slice SD _A , the slice SD _i is allocated to the D _i , and the mapping relation between the identifier of the D _i and the slice SD _i is established. Further, the management node establishes a mapping relationship between the identifier of D _i and the data storage node DS _i and the main data storage node DS _A. Further, further, the data storage node 1 saves the mapping relationship between the identifier of the storage D _i and the data storage node DS _i and the main data storage node DS _A. When garbage collection is performed on the metadata stripe, the management node can recycle the data of the metadata block in the data storage node and the main data storage node according to the mapping relationship between the identifier of the metadata block in the metadata stripe and the storage node , which improves the efficiency of metadata recovery.

In this embodiment of the present invention, in combination with the aforementioned distributed block device storage and data storage methods, as shown in FIG. 8 , the metadata blocks in the metadata striping using the EC algorithm are D ₁ , D ₂ and D ₃ . The test block is C ₁ . The management node where the client is located determines the main data storage node, data storage node 1, data storage node according to the partition view "partition-main data storage node-data storage node 1-data storage node 2-check storage node" recorded in the memory 202 Node 2 and the checksum storage node. The partition view indicates that the partition corresponds to the main data node and data storage node 1 and data node 2 for storing other data blocks of metadata stripes, and the check storage node for storing check data, which is stored in data storage node 1 and data storage node 1. The backup data storage node of the metadata block of storage node 2 is the primary data storage node. The management node sends D ₁ , D ₂ , D ₃ and C ₁ in the metadata striping based on the EC algorithm to the main data storage node, data storage node 1, data storage node 2 and check storage node 4, respectively. The main data storage node receives and stores D ₁ , the data storage node 1 receives and stores D ₂ , the data storage node 2 receives and stores D ₃ , and the check storage node receives and stores C ₁ . Data storage nodes 1 and 2 respectively determine the main data storage node according to the partition view, data storage node 1 backs up D2 to the main data storage node, data storage node ₂ backs up _D3 to the main data storage node, and the main data storage node receives the _D2 and _D3 are stored. In the specific implementation, as shown in FIG. 9 , the main data storage node allocates slice 7 for D ₁ from the hard disk managed by it, and establishes a mapping relationship between the identifier of D ₁ and slice 7; In the middle, allocate slice 8 for D ₂ , and establish the mapping relationship between the identifier of D ₂ and slice 8; the data storage node 2 allocates slice 9 for D ₃ from the hard disk it manages, and establishes the identifier of D ₃ and slice 9. Mapping relationship; the verification storage node allocates shard 10 to C ₁ from the hard disk it manages, and establishes a mapping relationship between the identifier of C ₁ and the shard 10. The main data storage node receives D2 sent by data storage node ₁ and D3 sent by data storage node ₂ , the main data storage node allocates slices 11 and ₁₂ from the hard disks it manages, and the main data storage node establishes D2's The mapping relationship between the ID and the shard 11, and the mapping relationship between the ID of _D3 and the shard 12. In the embodiment of the present invention, taking the mapping relationship between the identifier of the metadata block and the fragment as an example, when one process of the storage object program corresponds to one hard disk, that is, the storage node is the hard disk itself, the identifier of the metadata block is The mapping relationship with the shard is the mapping relationship between the identifier of the metadata block and the physical address of the shard; when one process of the storage object program corresponds to multiple hard disks, that is, the storage node manages multiple hard disks, the identifier of the metadata block is The mapping relationship with the shard includes the mapping between the identifier of the metadata block and the hard disk storing the metadata block, and the mapping from the hard disk storing the metadata block to the shard. Further, D2 is stored in slices 8 and ₁₁ respectively, _D3 is stored in slices 9 and 12 respectively, the management node establishes and saves the identification of D2 and the mapping of the data storage node ₁ and the main data storage node, establishes And save the mapping relationship between the identifier of D ₃ and the data storage node 2 and the main data storage node. Further, the data storage node 1 saves the mapping between the identifier of D ₂ and the data storage node 1 and the main data storage node, and the data storage node 2 saves the mapping relationship between the identifier of D ₃ and the data storage node 2 and the main data storage node. When garbage collection is performed on the metadata stripe, the management node can recycle the data of the metadata block in the data storage node and the main data storage node according to the mapping relationship between the identifier of the metadata block in the metadata stripe and the storage node , which improves the efficiency of metadata recovery.

Therefore, in the scenario where metadata striping composed of the EC algorithm is used to achieve data reliability, the primary data storage node backs up other metadata blocks in the metadata striping, because only the metadata blocks on the data storage node need to be stored in the primary data storage node. Compared with the multiple copies of all metadata blocks in the prior art, the backup on the storage node reduces the storage space. At the same time, when the client accesses the metadata, it only needs to access all the metadata blocks from the main data storage node, which improves the metadata access. speed.

Embodiments of the present invention also provide a non-volatile computer-readable storage medium and a computer program product, the computer program instructions contained in the non-volatile computer-readable storage medium and the computer program product, and the CPU executing the computer program loaded in the memory The program instructions are used to realize the functions corresponding to the management node and the storage node (main data storage node, data storage node and check storage node) in various implementations of the present invention.

Exemplary descriptions are given in the embodiments of the present invention. "Slice 1" and "Slice 2" in the embodiment of the present invention. . . "Shard 12" etc. are not used to strictly define the precedence relationship, but are only used to distinguish different shards. The slice in this embodiment of the present invention may be a physical block in a hard disk or the like. The hard disk in this embodiment of the present invention, as described above, may be at least one of a mechanical disk and a solid-state hard disk. In the embodiment of the present invention, the hard disk corresponding to the process storing the object program may also be a storage array, etc., which is not limited in the embodiment of the present invention.

In the several embodiments provided by the present invention, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the division of the units in the apparatus embodiments described above is only a logical function division, and other division methods may be used in actual implementation, for example, multiple units or components may be combined or integrated 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 invention 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.

Claims (10)

1. The method for storing the metadata in the distributed storage system is characterized in that the distributed storage system comprises a management node and (M + N) storage nodes, wherein the management node and the (M + N) storage nodes store partition views of metadata strips; the partitioned view of the metadata stripe contains a primary data storage node DS_AData storage node DS_iAnd checking the storage node CS_r(ii) a Wherein N is a natural number not less than 2, M is a natural number not less than 1, a is one of natural numbers 1 to N, i is each of natural numbers 1 to N except a, r is each of natural numbers 1 to M; the method comprises the following steps:

the management node determines a primary data storage node DS for the metadata stripe according to the partition view of the metadata stripe_AData storage node DS_iAnd checking the storage node CS_r(ii) a The metadata stripe contains a metadata block D_A、D_iAnd a check block Cr;

the management node is D_iTo said data storage node DS_iD is_ATo the primary data storage node DS_AMixing C with_rSent to the check storage node CS_r；

The check storage node CS_rReceive and store C_r；

The data storage node DS_iReceiving and storing D_iAnd partitioning D according to said metadata stripe_iTo the primary data storage node DS_A；

The primary data storage nodeDS_AReceiving and storing D_AAnd D_i。

2. The method of claim 1, wherein the management node determines a primary data storage node (DS) for the metadata stripe based on a partition view of the metadata stripe_AData storage node DS_iAnd checking the storage node CS_rThe method specifically comprises the following steps:

the management node determines a partition corresponding to the metadata stripe according to a write request for generating the metadata in the metadata stripe;

the management node inquires the partition view of the metadata strips according to the partitions corresponding to the metadata strips to determine the main data storage node DS_AThe data storage node DS_iAnd said check storage node CS_r。

3. The method according to claim 2, wherein the management node determines the partition corresponding to the metadata stripe according to an address carried by the write request.

4. The method of claim 1, wherein the check storage node CS_rThe storing of Cr specifically includes: the check storage node CS_rAllocating a slice S for the Cr_rAnd establishing the identifier of the Cr and the segment S_rThe mapping relationship of (2);

the data storage node DS_iStore D_iThe method specifically comprises the following steps: the data storage node DS_iIs said D_iDistribution shards SD_iAnd establishing said D_iAnd the slice SD_iThe mapping relationship of (2);

the main data storage node DS_AStore D_AAnd D_iThe method specifically comprises the following steps: the main data storage node DS_AIs said D_ADistribution shards SD_AAnd establishing said D_AAnd the slice SD_AIs the mapping relationship of D_iDistribution shards SD_iAnd establishing said D_iAnd the slice SD_iThe mapping relationship of (2).

5. A distributed storage system, comprising a management node and (M + N) storage nodes, wherein the management node and the (M + N) storage nodes each store a partition view of a metadata stripe; the partitioned view of the metadata stripe contains a primary data storage node DS_AData storage node DS_iAnd checking the storage node CS_r(ii) a Wherein N is a natural number not less than 2, M is a natural number not less than 1, a is one of natural numbers 1 to N, i is each of natural numbers 1 to N except a, r is each of natural numbers 1 to M;

the management node is configured to determine a primary data storage node DS for the metadata stripe according to the partitioned view of the metadata stripe_AData storage node DS_iAnd checking the storage node CS_r(ii) a The metadata stripe contains a metadata block D_A、D_iAnd check block Cr, D_iTo said data storage node DS_iD is_ATo the primary data storage node DS_AMixing C with_rSent to the check storage node CS_r；

The check storage node CS_rFor receiving and storing C_r；

The data storage node DS_iFor receiving and storing D_iAnd partitioning D according to said metadata stripe_iTo the primary data storage node DS_A；

The main data storage node DS_AFor receiving and storing D_AAnd D_i。

6. The system of claim 5, wherein the management node is specifically configured to determine the partition corresponding to the metadata stripe according to a write request for generating the metadata in the metadata stripe, and determine the partition corresponding to the metadata stripe according to the partition corresponding to the metadata stripeQuerying the partition view of the metadata stripe to determine the primary data storage node DS_AThe data storage node DS_iAnd said check storage node CS_r。

7. The system according to claim 6, wherein the management node is further configured to determine, according to an address carried by the write request, a partition corresponding to the metadata stripe.

8. The system of claim 5, wherein the check storage node CS_rIn particular for assigning a slice S to said Cr_rAnd establishing the identifier of the Cr and the segment S_rThe mapping relationship of (2);

the data storage node DS_iIn particular for the use of_iDistribution shards SD_iAnd establishing said D_iAnd the slice SD_iThe mapping relationship of (2);

the main data storage node DS_AIn particular for the use of_ADistribution shards SD_AAnd establishing said D_AAnd the slice SD_AIs the mapping relationship of D_iDistribution shards SD_iAnd establishing said D_iAnd the slice SD_iThe mapping relationship of (2).

9. A non-transitory readable storage medium containing computer program instructions executable in a distributed storage system, the distributed storage system comprising a management node and (M + N) storage nodes each storing a partitioned view of a metadata stripe; the partitioned view of the metadata stripe contains a primary data storage node DS_AData storage node DS_iAnd checking the storage node CS_r(ii) a Wherein N is a natural number not less than 2, M is a natural number not less than 1, A is one of natural numbers 1 to N, i is each of natural numbers 1 to N except A, r isEach of natural numbers 1 to M; when the computer program instructions are executed by one or more computers, the one or more computers are operable as the management node to determine a primary data storage node, DS, for the metadata stripe from a partitioned view of the metadata stripe_AData storage node DS_iAnd checking the storage node CS_r(ii) a The metadata stripe contains a metadata block D_A、D_iAnd check block Cr, D_iTo said data storage node DS_iD is_ATo the primary data storage node DS_AMixing C with_rSent to the check storage node CS_r(ii) a The one or more computers act as the check storage node CS_rFor receiving and storing C_r；

Said one or more computers being said data storage node DS_iFor receiving and storing D_iAnd partitioning D according to said metadata stripe_iTo the primary data storage node DS_A；

Said one or more computers being said primary data storage node DS_AFor receiving and storing D_AAnd D_i。

10. The storage medium of claim 9, further comprising computer program instructions to cause the one or more computers to, as the management node, determine a partition corresponding to the metadata stripe according to a write request to generate metadata in the metadata stripe, determine the primary data storage node (DS) according to a partition view of the metadata stripe queried by the partition corresponding to the metadata stripe_AThe data storage node DS_iAnd said check storage node CS_r。

Images