CN109840051B - Data storage method and device for a storage system - Google Patents

Abstract

In the method, when the topological structure of a storage system is a first topological structure, M first storage partition groups corresponding to the first topological structure are generated, and each of the M first storage partition groups corresponds to K first storage devices; when the topological structure of the storage system is updated from a first topological structure to a second topological structure different from the first topological structure, N second storage partition groups corresponding to the second topological structure are generated, each of the N second storage partition groups corresponds to P second storage devices, and then the M first storage partition groups stored in the storage system are updated by using the N second storage partition groups, so that after receiving new data to be stored, the new data to be stored is stored in the second storage device corresponding to at least one second storage partition group in the N second storage partition groups.

Description

Data storage method and device for a storage system

technical field

The present application relates to the field of storage technologies, and in particular, to a data storage method and device of a storage system.

Background technique

With the proliferation of Internet users and the development of business diversity, more and more data (eg, user data, business configuration data, etc.) need to be stored in a storage system so as to be used for analysis and business guidance.

In order to improve the storage performance of the storage system, a method of partitioning the hard disk using virtualization technology is proposed. , establish several partition teams (partition teams, PTs), for example, create 2053 PTs, each PT includes 6 partitions (partitions, pts), and then, according to the principles of balance and/or position balance, the Several PTs are scattered on all the hard disks of the storage system, so that each PT is fixedly mapped to the partitions in the 6 hard disks of the storage system. In this way, when storing data on this storage system, the storage system only needs to Select a PT from the PTs, and then automatically store the stripe on the corresponding hard disk according to the mapping relationship between the PT and the hard disk of the storage system, which can simplify the steps of data storage in the storage system. Only the index number of the PT corresponding to the stripe needs to be stored, and then the data can be acquired from the hard disk according to the index number of the PT, which can save the overhead of the storage system.

However, since the storage system is a dynamic system, for example, some storage nodes in the storage system may fail, thereby causing the storage capacity of the storage system to decrease (capacity reduction), or new storage may be added to the storage system Then, when the storage system changes, for example, when the storage system is reduced in capacity, some PTs in the pre-created PT groups cannot be used, so the partitions in other hard disks corresponding to the PT cannot be used. use, thereby reducing the space utilization in the storage system.

SUMMARY OF THE INVENTION

The present application provides a data storage method and device for improving the space utilization of a storage system.

A first aspect provides a data storage method for a storage system. In the method, when the topology of the storage system is a first topology, M first storage partition groups corresponding to the first topology are generated, and the Each of the M first storage partition groups corresponds to K first storage devices; when the topology of the storage system is updated from the first topology to a second topology different from the first topology is generated, N second storage partition groups corresponding to the second topology structure are generated, and each second storage partition group in the N second storage partition groups corresponds to P second storage devices, and then the N second storage partition groups are used. two second storage partition groups update the M first storage partition groups stored in the storage system, so that after receiving new data to be stored, the new data to be stored is stored in the same N second storage partition groups In the second storage device corresponding to at least one second storage partition group of , N, P, M and K are positive integers.

In the above technical solution, the storage system supports dynamic storage partition groups. When the topology structure of the storage system changes, the storage partition group of the storage system also changes. For example, the storage system consists of three storage nodes and each storage node There are 3 hard disks. To avoid the problem that the entire storage system cannot be used due to a single node failure, each storage partition group in the storage system can include 4 data disks and 2 parity disks. In this case, The space utilization rate of the storage system is 4/(4+2)=66.6%; when the storage system expands to 5 storage nodes and each storage node has 3 hard disks, the storage system can generate a new storage partition group , each storage partition group can include 8 data disks and 2 check disks, so the space utilization rate is 8/(8+2)=80%, so that the space utilization rate of the storage system can be improved. When the storage system is reduced in capacity, due to the failure of one of the hard disks in the three storage nodes, the storage system can also generate a new storage partition group to adapt to the reduced storage system. All partition groups can be used, which can improve the space utilization of the storage system.

In a possible design, the value of P is the same as the data redundancy mode configured by the storage system under the second topology structure, and the value of K is the same as the data redundancy mode configured by the storage system under the first topology structure The rest mode is the same.

In the above technical solution, the number of storage devices included in the storage partition group in the storage system can be the same according to the data redundancy modes configured by the storage system under different topological structures, so that the space of the hard disk of the storage system can be reduced. When fully applied, the utilization rate of the hard disk of the storage system can be improved.

In a possible design, when P is greater than K, a part of the data stored in each of the L first storage devices corresponding to each first storage partition group is different from the M second storage devices. The data stored in one of the P second storage devices corresponding to the partition group is the same, and L is a positive integer smaller than P.

In the above technical solution, after the storage system is expanded, the L first storage devices included in each first storage partition group generated under the first topology may be in the same position as the corresponding second storage partition group, respectively. The second storage device in the same location, for example, the first storage device for storing the first data fragment and the first storage device for storing the third data fragment included in the first first storage partition group The device is the same as the second storage device for storing the first data fragment and the second storage device for storing the third data fragment included in the first second storage partition group, that is, each The P second storage devices included in each of the second storage partition groups may be associated with the K first storage devices included in each of the first storage partition groups. After the storage system is expanded, the amount of data stored in each storage device in the second storage partition group becomes smaller. Therefore, when the data already stored in the storage system needs to be stored according to M second storage partition groups, the L Part of the data stored in the first storage device is the same as the data stored in the second storage device at the same location in the corresponding second storage partition group, so this part of data does not need to be migrated, which can reduce the consumption of data migration. resource.

In a possible design, the storage system may store part of the data in the original data of the K storage devices corresponding to one storage partition group in the M first storage partition groups according to the N second storage partition groups A second storage partition group in the partition group performs data migration, and part of data in each original data is different from data stored in each of the L second storage devices.

In the above technical solution, if the L first storage devices included in each first storage partition group are respectively the same as the second storage devices in the same position in the corresponding second storage partition group, the storage system may only It is necessary to migrate other data in each first storage partition group except a part of the data of the L first storage devices, and the part of the data of the L first storage devices is L corresponding to the L first storage devices. data stored in the second storage device, thereby reducing the amount of data to be migrated.

In a possible design, when P is less than K, the data stored in each first storage device in the S first storage devices corresponding to each first storage partition group is different from the data stored in the M second storage partitions. A part of data stored in one of the P second storage devices corresponding to the group is the same, and S is a positive integer smaller than P.

In the above technical solution, after the storage system is reduced in capacity, the S first storage devices included in each first storage partition group generated under the first topology structure may be respectively in the same position as the corresponding second storage partition group. The second storage devices in the same location are the same, for example, the first storage device included in the first first storage partition group for storing the first data fragment is the same as the first storage device included in the first second storage partition group. The second storage devices used to store the first data fragment are the same, that is, the P second storage devices included in each second storage partition group may be the same as the K included in each first storage partition group. A first storage device is associated. After the storage system is reduced in capacity, the amount of data stored by each storage device in the second storage partition group becomes larger. Therefore, when the data already stored in the storage system needs to be stored according to M second storage partition groups, the S The data stored in the first storage devices is the same as part of the data stored in the second storage devices at the same position in the corresponding second storage partition group, so the data stored in the S first storage devices may not need to be migrated , which can reduce the resources consumed by data migration.

In a possible design, the storage system may store part of the data in the original data of the K storage devices corresponding to one storage partition group in the M first storage partition groups according to the N second storage partition groups A second storage partition group in the partition group performs data migration, and part of data in each original data is different from any part of data stored in each of the P second storage devices.

In the above technical solution, if the S first storage devices included in each first storage partition group are respectively the same as the second storage devices in the same position in the corresponding second storage partition group, the storage system may only Data on other storage devices different from the S first storage devices in each first storage partition group needs to be migrated, so that the amount of data to be migrated can be reduced.

In a possible design, the P second storage devices corresponding to any second storage partition group in the N second storage partition groups are associated with any first storage partition in the M first storage partition groups. The K first storage devices included in the group are not the same.

In the above technical solution, whether the storage system is capacity expansion or capacity reduction, the plurality of second storage devices included in each second storage partition group generated by the storage system under the second topology structure may be different from those under the first topology structure. The multiple first storage devices included in the corresponding first storage partition group are different, or, each second storage partition group generated by the storage system under the second topology structure is different from the corresponding first storage partition under the first topology structure. The storage devices in the same position in the group are different, that is, the P second storage devices included in each second storage partition group may not be associated with the K first storage devices included in each first storage partition group , in this way, the storage system can flexibly select a manner for generating the M second storage partition groups, which can improve the flexibility of the storage system.

In a possible design, the storage system may store the original data of at least one first storage partition group in the M first storage partition groups according to one second storage partition group in the N second storage partition groups Perform data migration.

In the above technical solution, after the storage system has updated the storage partition groups from M first storage partition groups to N second storage partition groups, the data previously stored according to the M first storage partition groups can be migrated, Therefore, the already stored data can also be stored according to at least one second storage partition group in the N second storage partition groups, which can improve the space utilization rate of the storage system.

In a second aspect, a data storage device of a storage system is provided. The device includes a processor for implementing the method described in the first aspect. The apparatus may also include a memory for storing program instructions and data. The memory is coupled to the processor, and the processor can invoke and execute program instructions stored in the memory, so as to implement the method described in the first aspect above. The apparatus may also include a communication interface for the apparatus to communicate with other devices. Illustratively, the other device is a storage node.

In one possible design, the apparatus includes a communication interface and a processor, wherein:

The processor is configured to generate N second storage partition groups corresponding to the second topology structure when the topology structure of the storage system is updated from the first topology structure to the second topology structure, and the N second topological structure Each second storage partition group in the storage partition group corresponds to P second storage devices, and N and P are positive integers; wherein, the first topology structure is different from the second topology structure;

The processor is further configured to use the N second storage partition groups to update the M first storage partition groups stored in the storage system, the M first storage partition groups and the first topology structure Correspondingly, each first storage partition group in the M first storage partition groups corresponds to K first storage devices, and M and K are positive integers;

The processor is further configured to, after receiving the new data to be stored through the communication interface, store the new data to be stored in at least one second storage partition with the N second storage partition groups in the second storage device corresponding to the group.

In a possible design, the value of P is the same as the data redundancy mode configured by the storage system under the second topology, and the value of K is the same as that of the storage system under the first topology. The configured data redundancy mode is the same.

In a possible design, when P is greater than K, a part of the data stored in each of the L first storage devices corresponding to each first storage partition group is different from the M second storage device. The data stored in one of the P second storage devices corresponding to the storage partition group is the same, and L is a positive integer smaller than P.

In one possible design, the processor is also used to:

Part of the data in each original data is migrated according to one second storage partition group in the N second storage partition groups, and each original data is stored in the M first storage partition groups. For K storage devices corresponding to one storage partition group, part of data in each of the original data is different from data stored in each of the L second storage devices.

In a possible design, when P is less than K, the data stored in each of the S first storage devices corresponding to each first storage partition group is different from the data stored in the M second storage devices. Part of the data stored in one of the P second storage devices corresponding to the partition group is the same, and S is a positive integer smaller than P.

In one possible design, the processor is also used to:

Part of the data in each original data is migrated according to one second storage partition group in the N second storage partition groups, and each original data is stored in the M first storage partition groups. For K storage devices corresponding to one storage partition group, part of data in each original data is different from any part of data stored in each of the P second storage devices.

In a possible design, the P second storage devices corresponding to any one of the N second storage partition groups are the same as any one of the M first storage partition groups. The K first storage devices included in the storage partition group are different.

In one possible design, the processor is also used to:

Data migration is performed for each original data according to one second storage partition group in the N second storage partition groups, and the each original data is that the storage system in the first topology structure, according to the At least one first storage partition group in the M first storage partition groups stores data in the storage system.

In a third aspect, a data storage device of a storage system is provided. The communication device may be a controller of the storage system or a device in the controller of the storage system. The communication device may include a processing module and a communication module. These modules The method performed in any one of the above design examples of the first aspect can be performed, specifically:

The processing module is configured to generate N second storage partition groups corresponding to the second topology when the topology of the storage system is updated from the first topology to the second topology, the N second Each second storage partition group in the storage partition group corresponds to P second storage devices, and N and P are positive integers; wherein, the first topology structure is different from the second topology structure;

The processing module is further configured to use the N second storage partition groups to update the M first storage partition groups stored in the storage system, the M first storage partition groups and the first topology structure Correspondingly, each first storage partition group in the M first storage partition groups corresponds to K first storage devices, and M and K are positive integers;

The processing module is further configured to, after receiving the new data to be stored through the communication module, store the new data to be stored in at least one second storage partition with the N second storage partition groups in the second storage device corresponding to the group.

In a possible design, the value of P is the same as the data redundancy mode configured by the storage system under the second topology, and the value of K is the same as that of the storage system under the first topology. The configured data redundancy mode is the same.

In a possible design, when P is greater than K, a part of the data stored in each of the L first storage devices corresponding to each first storage partition group is different from the M second storage device. The data stored in one of the P second storage devices corresponding to the storage partition group is the same, and L is a positive integer smaller than P.

In a possible design, the processing module is also used to:

Part of the data in each original data is migrated according to one second storage partition group in the N second storage partition groups, and each original data is stored in the M first storage partition groups. For K storage devices corresponding to one storage partition group, part of data in each of the original data is different from data stored in each of the L second storage devices.

In a possible design, when P is less than K, the data stored in each of the S first storage devices corresponding to each first storage partition group is different from the data stored in the M second storage devices. Part of the data stored in one of the P second storage devices corresponding to the partition group is the same, and S is a positive integer smaller than P.

In a possible design, the processing module is also used to:

Part of the data in each original data is migrated according to one second storage partition group in the N second storage partition groups, and each original data is stored in the M first storage partition groups. For K storage devices corresponding to one storage partition group, part of data in each original data is different from any part of data stored in each of the P second storage devices.

In a possible design, the P second storage devices corresponding to any one of the N second storage partition groups are the same as any one of the M first storage partition groups. The K first storage devices included in the storage partition group are different.

In a possible design, the processing module is also used to:

Data migration is performed for each original data according to one second storage partition group in the N second storage partition groups, and the each original data is that the storage system in the first topology structure, according to the At least one first storage partition group in the M first storage partition groups stores data in the storage system.

In a fourth aspect, a computer-readable storage medium is provided, the computer-readable storage medium stores a computer program, the computer program includes program instructions, and the program instructions, when executed by a computer, cause the computer to execute a first The method of any one of the aspects.

In a fifth aspect, a computer program product is provided, the computer program product stores a computer program, and the computer program includes program instructions that, when executed by a computer, cause the computer to execute any one of the first aspects. method described in item.

In a sixth aspect, a chip system is provided, the chip system includes a processor, and may further include a memory, for implementing the method described in the first aspect. The chip system can be composed of chips, and can also include chips and other discrete devices.

For the beneficial effects of the second to sixth aspects and implementations thereof, reference may be made to the description of the method of the first aspect and the beneficial effects of the implementations thereof.

Description of drawings

1 is a schematic diagram of the distribution of 10 PTs created by a storage system in a storage device in an embodiment of the present application;

2A is a schematic diagram of an example of a view corresponding to a group of PTs in an embodiment of the present application;

2B is a schematic diagram of another example of a view corresponding to a group of PTs in an embodiment of the present application;

3A is a schematic diagram of an example of a storage system provided in an embodiment of the present application;

3B is a schematic diagram of another example of a storage system provided in an embodiment of the present application;

3C is a schematic diagram of another example of a storage system provided in an embodiment of the present application;

4 is a flowchart of an example of a data storage method provided in an embodiment of the present application;

5 is a schematic diagram of an example of 6 PTs of the storage system in the embodiment of the application;

6 is a schematic diagram of an example of a first view of a storage system under a first topology structure in an embodiment of the present application;

7A is a schematic diagram of an example of a second view of a storage system under a second topology structure in an embodiment of the present application;

7B is a schematic diagram of another example of a second view of a storage system under a second topology structure in an embodiment of the present application;

7C is a schematic diagram of another example of a second view of the storage system under the second topology structure in the embodiment of the present application;

8 is a flowchart of another example of a data storage method in an embodiment of the present application;

9 is a schematic diagram of another example of a view corresponding to a group of PTs in an embodiment of the present application;

10 is a schematic diagram of another example of a first view of a storage system in an embodiment of the application;

11 is a schematic diagram of another example of a second view of a storage system in an embodiment of the application;

12 is a schematic structural diagram of an example of a data storage device of a storage system provided in an embodiment of the application;

FIG. 13 is a schematic structural diagram of another example of a data storage device of a storage system provided in an embodiment of the present application.

Detailed ways

In order to make the objectives, technical solutions and advantages of the embodiments of the present application more clear, the embodiments of the present application will be described in further detail below with reference to the accompanying drawings.

Hereinafter, the technical terms involved in the embodiments of the present application will be explained to facilitate the understanding of those skilled in the art.

1) A partition (partition, pt) is the basic storage unit of the hard disk of the storage system. The pts are grouped, each pt belongs to only one PT, and each PT contains multiple pts.

2) Data redundancy mode, also known as data redundancy level, is a protection mechanism adopted when storing data in order to prevent data loss or errors. Generally speaking, the realization of data redundancy can usually include the following two methods, namely, a copy method or an erasure coding (erasure coding, EC) method, wherein the copy method can be understood as copying multiple copies of the data to be stored, for example, Copy the data to be stored into three copies, and then store the copied data in different partitions. In this way, when a partition fails and the stored data is lost, the data can be obtained from other partitions. In this case, the number of replicas is the data redundancy mode. The EC method can be understood as dividing the data to be stored into multiple pieces, and then encoding each piece of data after division, so as to obtain multiple pieces of data, and generate the verification part according to the multiple pieces of data, Then, the data fragmentation and the verification part are stored in different partitions. In this way, if one partition fails and the stored data is lost, the data can be reconstructed according to the data fragmentation and verification part stored on other storage nodes. . In this case, the sum of the number of data shards and inspection parts is the data redundancy mode, or it can be called EC ratio, for example, EC4+2 mode, which means 4 partitions for storing data shards, 2 partitions for storing the parity part; EC8+2 mode means 8 partitions for storing data shards and 2 partitions for storing the parity part. Of course, when data is stored in other ways, the data redundancy mode can also be interpreted similarly, which is not limited here.

3) Storage devices, such as serial advanced technology attachment (SATA) hard disks, small computer system interface (SCSI) hard disks, serial attached SCSI interfaces (serial attached SCSI, SAS), Fibre Channel Interface (fibre channel, FC) hard disk, mechanical hard disk (hard disk drive, HDD) and solid state drive (solid state drive, SSD) and so on.

4) View, which is a mapping table from pt to storage device (eg, hard disk). Please refer to FIG. 1 , which is a schematic diagram of the distribution of the 10 PTs created for the storage system in the storage device. In Figure 1, each group of PTs includes 6 PTs as an example, the 6 PTs of the same color are the same group of PTs, and A to J respectively represent different storage devices. For example, the 6 PTs of the first group of PTs are located in the storage device. On A to storage device F, the 6 pts of the second group of PTs are located on storage device B to storage device G, and so on. Taking the data stored in EC mode and the data redundancy mode as EC4+2 as an example, for the first group of PTs, the view shown in FIG. 2A can be obtained, where D1 to D4 indicate that 4 of the first group of PTs are used for storage The pt, P and Q of the data fragment are respectively used to store the pt of the check part. The characters in brackets indicate the storage device to which the pt belongs. For the second group of PTs, the view shown in Figure 2B can be obtained. For other PTs, the view shown in Figure 2B can be obtained. The same is true, and will not be repeated here. In the views shown in FIG. 2A and FIG. 2B , the numbers of the hard disks are arranged in ascending order, but in practice, the views may also be out of order, which is not limited here.

5) In the embodiments of the present application, "plurality" refers to two or more than two. In view of this, "plurality" may also be understood as "at least two" in the embodiments of the present application. "At least one" can be understood as one or more, such as one, two or more. For example, including at least one refers to including one, two or more, and does not limit which ones are included. For example, including at least one of A, B, and C, then including A, B, C, A and B, A and C, B and C, or A and B and C. "And/or", which describes the association relationship of the associated objects, means that there can be three kinds of relationships, for example, A and/or B, which can mean that A exists alone, A and B exist at the same time, and B exists alone. In addition, the character "/", unless otherwise specified, generally indicates that the related objects are an "or" relationship.

Unless stated to the contrary, ordinal numbers such as "first" and "second" mentioned in the embodiments of the present application are used to distinguish multiple objects, and are not used to limit the order, sequence, priority, or importance of multiple objects.

Some concepts involved in the present application have been introduced above, and the technical background of the present application will be described below.

The storage system is initially redundant in units of storage devices (for example, HDDs). For example, if the data redundancy mode is EC8+2, then 8 hard disks in the storage system are fixed as data disks to store data slices and use storage for fixed use. The two hard disks in the system are used as check disks to store the check parts generated according to the data shards. Since the data disk and the verification disk are fixed, all the verification parts of the data are written in these two verification disks. In this way, when the data in any one of the data disks needs to be updated, the two verification disks need to be updated at the same time. , so that the two parity disks become the bottleneck of the storage system due to high access frequency or too many access times.

In order to prevent some fixed hard disks from becoming a bottleneck, the industry has introduced a technology to store data fragments and check parts after breaking them up. All hard disks can store data slices and check parts. However, since the storage locations of data slices and check parts are not fixed, for each PT, it is necessary to record information such as the number of its corresponding multiple hard disks. bigger. Due to the above problems, the aforementioned method of partitioning a hard disk through virtualization technology is proposed. After several PTs are created, views corresponding to the several PTs are generated, so that data read and write operations are performed according to the views.

Since the view has been generated when the PT is created, when the storage system is expanded, storage nodes are added or some storage nodes fail, resulting in low space utilization in the storage system.

In view of this, an embodiment of the present application provides a data storage method for a storage system. In the method, when the topology of the storage system is a first topology, M first storage partitions corresponding to the first topology are generated group, each first storage partition group in the M first storage partition groups corresponds to K first storage devices, that is, a first view corresponding to the first topology structure is obtained, and the topology structure of the storage system is determined by the first topology When the structure is updated to a second topology structure different from the first topology structure, N second storage partition groups corresponding to the second topology structure are generated, and each second storage partition group in the N second storage partition groups is generated. The storage partition group corresponds to the P second storage devices, that is, a second view corresponding to the second topology is obtained, and then the N second storage partition groups are used to update the M first storage partition groups stored in the storage system, In this way, after receiving the new data to be stored, the new data to be stored is stored in the second storage device corresponding to at least one second storage partition group in the N second storage partition groups, N, P, M and K are positive integers. That is to say, in the method in this embodiment of the present application, the storage system supports a dynamic storage partition group (or view), and when the topology structure of the storage system changes, the storage partition group (or view) of the storage system also changes For example, a storage system consists of 3 storage nodes and each storage node has 3 hard disks. To avoid the problem that the entire storage system cannot be used due to a single node failure, each storage partition group in the storage system can include 4 hard disks. Data disk and 2 parity disks. In this case, the space utilization of the storage system is 4/(4+2)=66.6%; when the storage system expands, it expands to 5 storage nodes and each storage node If there are 3 hard disks, the storage system can generate a new storage partition group, and each storage partition group can include 8 data disks and 2 test disks, then the space utilization rate is 8/(8+2)=80%, Thus, the space utilization of the storage system can be improved. When the storage system is reduced in capacity, due to the failure of one of the hard disks in the three storage nodes, the storage system can also generate a new storage partition group to adapt to the reduced storage system. All partition groups can be used, which can improve the space utilization of the storage system.

The technical solutions in the embodiments of the present application are applied to a storage system, and the storage system may be a file storage system, a block storage system, or an object storage system, or a combination of the above storage systems, which is not limited in the embodiments of the present application.

FIG. 3A shows an example of a storage system involved in this embodiment of the present application. As shown in FIG. 3A , the storage system includes a controller 301 and a plurality of storage nodes 302 for storing data. A collection of coupled nodes composed of nodes, which cooperate to provide external services, and storage nodes can also be called storage racks. As shown in FIG. 3A , the storage system includes storage racks 1 to 3 . Each storage rack includes a plurality of hard disks, and the hard disks may be HDD disks. The plurality of hard disks constitute a storage server. In FIG. 3A , three hard disks constitute a storage server. At least one storage server constitutes a storage rack, and in FIG. 3A , one storage server constitutes a storage rack. The controller 301 is configured to generate a storage partition group according to the topological structure of the storage system, and manage operation requests, such as processing a request for a write operation, write data into a corresponding storage medium according to the storage partition group, or process a request for a read operation. The controller 301 can be a central processing unit (CPU), an application-specific integrated circuit (ASIC) or a field programmable gate array to request data from a corresponding storage medium according to a storage partition group. (field-programmable gate array, FPGA) etc.

Different from the architecture shown in FIG. 3A, the architecture shown in FIG. 3B includes multiple controllers, such as controller A and controller B, and communication between controller A and controller B can be performed, so that when When a controller in the storage system fails, the cluster storage system can still provide services for other devices (eg, clients) interacting with the storage system through other controllers.

It should be noted that the storage system is not limited to the architecture shown in FIG. 3A and FIG. 3B. For example, the storage system may include more storage nodes. For example, as shown in FIG. 3C, the storage system may include five storage nodes. node. The storage system may also include other devices, such as an arbitration server, etc. The storage system described in the embodiments of the present application is to more clearly describe the technical solutions of the embodiments of the present application, and does not constitute a technical solution provided by the embodiments of the present application. Limitations, those of ordinary skill in the art know that with the evolution of storage technologies and storage system architectures, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

The technical solutions provided by the embodiments of the present application are described below with reference to the accompanying drawings.

An embodiment of the present application provides a data storage method. Please refer to FIG. 4 , which is a flowchart of the method. In the following description, the method is used in the storage system shown in FIG. 3A as an example, that is, the method is executed by the controller in the storage system shown in FIG. 3A .

S41. The controller of the storage system generates and stores M first storage partition groups corresponding to the first topology, and each first storage partition group in the M first storage partition groups corresponds to K first storage devices.

In this embodiment of the present application, the value of K is the same as the data redundancy mode configured by the storage system under the first topology structure, and M and K are positive integers.

As an example, the storage system shown in FIG. 3A includes 3 storage nodes, each storage node includes 4 hard disks, a total of 12 hard disks, and the hard disks are labeled as hard disk 1 to hard disk 12 in sequence. The data redundancy mode configured by the storage system is EC4+2 mode. Assuming that the storage system creates 6 PTs, each PT includes 6 partitions, of which the data shard occupies 4 partitions, and the inspection part occupies 2 partitions. Thus, the PT shown in FIG. 5 is obtained. Then, using a data arrangement algorithm, the 6 PTs are mapped to the partitions of the 6 hard disks. The data arrangement algorithm may be a round-robin algorithm, that is, starting from the first hard disk in the storage system, for example, D1 in the first PT is mapped to the partition in hard disk 1, and then the number of hard disks is increased in sequence. Number, select the corresponding hard disk to map D2 in the first PT, that is to say, map D2 of the first PT to the partition of hard disk 2, and so on, until the pt in each PT is mapped to the partition of each hard disk. Alternatively, the data arrangement algorithm may also be a hash algorithm, for example, according to the number of pt in the PT and the number of the hard disk, a hash algorithm is used, such as message-digest algorithm five (MD5) or mmh3 algorithm, etc. , calculate the number of the hard disk mapped by the pt in each PT. Of course, other data arrangement algorithms can also be used, which are not limited here.

It should be noted that the data arrangement algorithm needs to meet the following conditions:

1) Data security.

When mapping pts to hard disks, each pt in the same PT should be distributed to storage devices with different security levels as much as possible. The security levels of various storage devices can be configured by technicians. For example, the order of security levels from high to bottom is: storage rack>storage server>hard disk. That is to say, if there are storage servers and storage racks in the storage device, the PTs in each PT can be distributed on different storage servers or different storage racks, so that data security can be higher. For example, in the storage system shown in Figure 3A, including 3 storage racks, 6 pts in one PT can be evenly distributed in the 3 storage racks, that is, in each storage rack 2 pts can be mapped. And since one storage rack includes two storage servers, the two pts in one storage rack can be mapped to different storage servers respectively.

2) Data balance.

When mapping pts to hard disks, try to make the number of pts on each hard disk proportional to the capacity of the hard disk. 2 pts are mapped on hard disk 1, and 1 pt in one PT is mapped on hard disk 2, so that a hard disk can be prevented from becoming a hot spot due to excessive access times on a certain hard disk. If the capacity of each hard disk in the storage system is the same, the same number of pts can be distributed in each hard disk.

3) Data type balance.

When mapping pts to hard disks, the number of pts located in the same position among multiple PTs on each hard disk is proportional to the capacity of the hard disk. For example, among the 10 PTs shown in The first pt (the pt whose position is 1) is taken as an example. The first pt of the 10 PTs is evenly distributed among the 10 hard disks from hard disk A to hard disk J. That is to say, each hard disk has a position of 1 pt, and the pts for other locations, like the first pt, are evenly distributed in hard disk A to hard disk J, so as to avoid when the small file is not full of stripes, that is to say, the small file may only be stored using 2 of the 4 partitions used to store data slices, if the small file is read or written, only the hard disk corresponding to the first 2 partitions will be accessed, and the hard disk corresponding to the last 2 partitions will be accessed. will not be accessed, thus making the hard disk corresponding to the first 2 partitions a hot issue.

4) Data recovery balance.

That is to say, if the hard disk where a PT in a PT is located fails, it is necessary to read data from the hard disk where other PTs in the PT are located to restore the data on the failed hard disk. For the data redundancy mode, EC4 is used. For a storage system in +2 mode, each PT mapped to the faulty hard disk must read data from the other 5 hard disks, in this case, from each PT from every other non-faulty hard disk. The total amount of data read should be as equal as possible.

In the embodiment of the present application, taking the controller using a polling algorithm to map 6 PTs to 12 hard disks as an example, the first view shown in FIG. 6 is obtained. In Figure 6, the 6 pts corresponding to PT1, PT3 and PT5 are mapped in HDD 1 to HDD 6, among which, 4 data disks are HDD 1 to HDD 4, HDD 5 and HDD 6 are inspection disks, PT2, PT4 and The 6 pts corresponding to PT6 are mapped in the hard disks 7 to 12, wherein, the 4 data disks are the hard disks 7 to 10, and the hard disks 11 and 12 are test disks.

S42. After receiving the data to be stored, the controller of the storage system determines to store the data to be stored in at least one partition group of the M first storage partition groups.

As an example, assuming that each PT can store 128 megabytes (Megabyte, MB) of data, then when the controller of the storage system receives data with a size of 256 MB, the controller can, according to a preset rule, store data from Two PTs are selected from among the plurality of PTs for storing the data. The preset rule may be to select several required PTs in sequence from multiple unused PTs. For example, the data is the first data received by the controller of the storage system, that is to say, before this, the No data is stored in the storage system, then the controller can select PT1 and PT2 for storing the data. Alternatively, it may also be randomly selected from a plurality of unused PTs, for example, the controller randomly selects PT1 and PT3 for storing the data. Then, the controller divides the 256MB data into 2 pieces, each piece of data is divided into 4 pieces of data, and the size of each piece of data is 32MB, and then, according to the content of the 4 pieces of data in each piece of data Generate 2 check parts, and finally, according to the 4 data slices and 2 check parts of each piece of data, generate a data storage instruction corresponding to the data, the data storage instruction includes the data slices that the hard disk needs to store Or the inspection part.

As another example, the controller of the storage system may use a shredding mechanism to store data. For example, if the controller receives data with a size of 32K, the controller will use 8K as the granularity to break up the data, thereby dividing it into 4 pieces of data with a size of 8K, and then store the 4 pieces of data with a size of 8K separately. into a different PT.

Of course, the controller of the storage system can also use other methods to select the corresponding PT, which is not limited here. For the convenience of description, in the following, the controller of the storage system selects PT1 and PT3 for storing the data as an example.

S43. The controller of the storage system sends a data storage instruction to at least one storage node, the at least one storage node receives the data storage instruction, and the at least one storage node receives the data storage instruction and executes the data storage instruction.

It can be seen from FIG. 6 that the hard disks corresponding to PT1 are hard disk 1 to hard disk 6, and the hard disks corresponding to PT3 are also hard disk 1 to hard disk 6. Therefore, the controller of the storage system will be directed to the hard disk 1 to hard disk 6 corresponding to respectively. The storage nodes (that is, storage node 1 to storage node 3) respectively send data storage instructions. After receiving the data storage instruction, each storage node in storage node 1 to storage node 3 stores the data shard or check part carried in the data storage instruction in the corresponding partition, thereby completing the data storage process .

S44. Each storage node in the at least one storage node sends a heartbeat to the controller of the storage system, and the controller of the storage system receives the heartbeat.

In the embodiment of the present application, the heartbeat sent by each storage node to the controller of the storage system can be understood as a self-defined information, such as a heartbeat packet or a heartbeat frame, to let the receiver know that he is "online" to ensure that the connection is complete. effectiveness. The specific content of the heartbeat can be the content agreed between the DNS and the client, or it can include information about the status of each server and each hard disk in the node. Of course, it can also be an empty packet that only includes the header, which is not described here. limit. Each storage node can automatically send a heartbeat to the controller of the storage system at a certain time interval, or the controller of the storage system can first send query information to each storage node, and the storage node in the "online" state receives the After querying the information, it feeds back the heartbeat to the controller. In the embodiment of the present application, each storage node automatically sends a heartbeat to the controller at a certain time interval as an example.

Step S44 is an optional step, that is, it does not have to be executed. It is indicated by a dashed line in FIG. 4 .

S45. The controller of the storage system determines capacity expansion of the storage system.

In the embodiment of the present application, the expansion of the storage system can be understood as the topology structure of the storage system is updated from the first topology structure to the second topology structure, and the number of storage nodes in the second topology structure increases.

After the controller of the storage system receives the heartbeat of each storage node, it can determine the number of storage nodes in the storage system according to the number of heartbeats received within a preset time period. For example, in the storage system shown in FIG. 3A , the controller of the storage system may receive heartbeats sent by three storage nodes, thereby determining that there are three storage nodes in the storage system. If each received heartbeat also includes the status of the server and hard disk in the node, the controller can learn the number and status of storage devices in each storage node. After a hard disk) fails, the controller of the storage system can determine that the storage device is unavailable according to the heartbeat of the storage node, so that the PT corresponding to the storage device is not used during data storage.

Since the storage system is a dynamic system, the storage system may expand its capacity. For example, due to the increase in the amount of data that needs to be stored, storage racks, storage servers and hard disks are added, and there may also be capacity reduction, such as a reduction in the size of the storage system. Reduced due to small or damaged storage devices, etc. Taking the expansion of the storage system as an example, the storage system is changed from the topology shown in Figure 3A to the topology shown in Figure 3C, that is, two storage nodes are added, and each storage node includes four hard disks. After the storage system is expanded, the controller receives the heartbeats of 5 storage nodes within a preset time period, thereby determining that the topology of the storage system is expanded from the first topology with 3 storage nodes to 5 storage nodes. the second topology.

Of course, the controller of the storage system may also use other methods to determine whether the topology of the storage system has changed, which is not limited here. In this embodiment of the present application, the controller of the storage system determines that the topology of the storage system has changed through heartbeat. For example.

S46. The controller of the storage system generates N second storage partition groups corresponding to the second topology structure, and each second storage partition group in the N second storage partition groups corresponds to P second storage devices.

In this embodiment of the present application, the value of P is the same as the data redundancy mode configured by the storage system under the second topology structure, and N and P are positive integers.

As an example, please refer to Table 1 for the correspondence between various topology structures and data redundancy modes stored in the storage system. In Table 1, when the topology of the storage system is 3 storage nodes, the storage system can include two data redundancy modes, EC4+2 or EC3+3. When the topology of the storage system is 5 storage nodes, The storage system may include multiple data redundancy modes such as EC8+2, EC6+3, or EC6+2. When the storage system has other topologies, it corresponds to other data redundancy modes, which are not listed here.

Table 1

In this way, when the controller of the storage system determines that the topology of the storage system has changed, for example, the first topology of three storage nodes shown in FIG. 3A is changed to the second topology of five storage nodes shown in FIG. 3C . , the data redundancy mode configured by the storage system under the second topology structure can be determined according to the corresponding relationship of the data redundancy modes corresponding to various pre-stored topological structures, and then according to the determined data redundancy mode, the N second storage partition groups corresponding to the two topology structures.

It should be noted that, in the above description, the number of storage nodes is used as an example to describe the topology. In practical applications, the change of the topology may not only include the change of the number of storage nodes. Changes in the number of storage servers and hard disks, etc. In this case, it is also possible to define whether the topology structure changes according to the number of storage servers and hard disks, and the data redundancy mode configured under each topology structure. The specific process It is similar to the corresponding content when the topology structure is described above by taking the number of storage nodes as an example, and details are not repeated here.

In the embodiment of the present application, the storage system is changed from the first topology structure of 3 storage nodes shown in FIG. 3A to the second topology structure of 5 storage nodes shown in FIG. 3C , and the controller of the storage system determines that the The data redundancy mode configured in the second topology is EC8+2, that is to say, in the second topology, a PT includes 10 pts, of which 8 pts are used to store data slices, and 2 pts are used for In the storage verification part, that is, P is greater than K, the controller of the storage system may generate N second storage partition groups corresponding to the second topology, including but not limited to the following two ways.

In the first manner, N second storage partition groups are generated without considering the M first storage partition groups under the first topology.

As an example, the controller of the storage system first creates 6 PTs, and each PT includes 10 partitions, of which the data slice occupies 8 partitions, and the verification part occupies 2 partitions. A data arrangement algorithm, such as a polling algorithm or a hash algorithm, etc., maps the 6 PTs to 5 storage nodes (20 hard disks in total) of the storage system.

In this manner, the N second storage partition groups corresponding to the second topology structure may include the following two situations:

In case a, the first storage device at the L positions in each first storage partition group and the second storage device at the corresponding L positions in the corresponding second storage partition group are the same storage device, and L is positive. Integer.

Please refer to FIG. 7A , which is an example of the second view. In FIG. 7A , the value of N is 6. As shown in FIG. 7A , among the first 4 hard disks in each second storage partition group, the hard disk number at one position is the same as the hard disk number in the same position in the first view. For example, in the view shown in FIG. 7A , the hard disk where D3 of PT1 is located is hard disk 3, and in the view shown in FIG. 6, the hard disk where D3 of PT1 is located is also hard disk 3; in the view shown in FIG. 7A, the hard disk where D3 of PT2 is located is hard disk 9 , in the view shown in FIG. 6 , the hard disk where D3 of PT2 is located is also the hard disk 9; the hard disk where D1 of PT3 in the view shown in FIG. 7A is located is the same as that of D1 of PT3 in the view shown in FIG. 6 The hard disk is the same, and there are similar situations in other PTs, which are not listed here. In this case, the value of L is 1. The same hard disk in the first view and the second view is shown by hatching in FIG. 7A.

In case b, the P second storage devices corresponding to any one of the N second storage partition groups are different from the Kth storage devices included in any one of the M first storage partition groups. A storage device is not the same.

Please refer to FIG. 7B , which is an example of the second view. In FIG. 7B , the value of N is 6. As shown in FIG. 7B , the hard disk numbers corresponding to the 4 hard disks mapped by the first 4 pts in each second storage partition group and the 4 hard disk numbers mapped to the first 4 pts of the corresponding first storage partition group in the first view The hard disk numbers corresponding to the hard disks are completely different. For example, in the view shown in Figure 7B, the hard disk where D1 of PT1 is located is hard disk 14, the hard disk where D2 is located is hard disk 15, the hard disk where D3 is located is hard disk 4, and the hard disk where D4 is located is Hard disk 3, in the view shown in Figure 6, the hard disk where D1 of PT1 is located is hard disk 1, the hard disk where D2 is located is hard disk 2, the hard disk where D3 is located is hard disk 3, and the hard disk where D4 is located is hard disk 4; In the view shown, the hard disk where D1 of PT2 is located is hard disk 17, the hard disk where D2 is located is hard disk 19, the hard disk where D3 is located is hard disk 10, and the hard disk where D4 is located is hard disk 9. In the view shown in FIG. The hard disk where D1 is located is hard disk 7, the hard disk where D2 is located is hard disk 8, the hard disk where D3 is located is hard disk 9, and the hard disk where D4 is located is hard disk 10. Similar situations exist in other PTs, which are not listed here.

In this case, when the topology of the storage system changes, the storage system can support dynamic views. That is, when the topology of the storage system changes, the corresponding view can also change. For example, increase the number of views in each PT. The number of pts can improve the utilization of the storage space of the storage system.

In the second manner, considering the M first storage partition groups under the first topology, generate N second storage partition groups, that is, corresponding to the M first storage partition groups under the first topology. The first view is linked with the second views corresponding to the N second storage partition groups under the second topology, and the second view will consider the layout of the pts in the same position in the first view to ensure reliability, balance, and recovery. Under the premise of balance and other constraints, try to ensure that the pts in the same position in each PT fall into the same hard disk.

As an example, the controller of the storage system may use the same data arrangement algorithm as in S41 for many times, such as a polling algorithm or a hash algorithm, etc., to obtain multiple second views corresponding to the 6 PTs, and then According to the linkage between the plurality of second views and the first view, select a second view from the plurality of second views with the most pts at the same position in each PT falling into the same hard disk, as the second view in the second view A second view of the M second storage partition groups under the topology.

Please refer to FIG. 7C , which is an example of the second view. In FIG. 7C , the value of N is 6. As shown in FIG. 7C , among the first 4 hard disks in each second storage partition group, the hard disk numbers in two positions are the same as the hard disk numbers in the same position in the first view. For example, in the view shown in FIG. 7C , the hard disk where D3 of PT1 is located is hard disk 3, and the hard disk where D4 is located is hard disk 4. In the view shown in Figure 6, the hard disk where D3 of PT1 is located is also hard disk 3, and the hard disk where D4 is located is also hard disk 4; In the view shown in FIG. 7C, the hard disk where D3 of PT2 is located is hard disk 9, and the hard disk where D4 is located is hard disk 10. In the view shown in FIG. 6, the hard disk where D3 of PT2 is located is also hard disk 9, and the hard disk where D4 is located is also hard disk 9. It is also the hard disk 10; there are similar situations in other PTs, which are not listed here. The same hard disk in the first view and the second view is shown by hatching in FIG. 7C.

In this case, since the newly generated second view is extended based on the original first view, the data in the pts at the same position in each PT can not be migrated, thereby reducing the number of data migrations , to avoid occupying more bandwidth resources or input/output (I/O) resources, so that more bandwidth resources or I/O resources can be used for business services, which can improve the business service capability of the storage system.

S47. The controller of the storage system uses the N second storage partition groups to update the M first storage partition groups stored in the storage system.

After the controller of the storage system acquires the N second storage partition groups under the second topology, the N second storage partition groups can be used to replace the M first storage partition groups stored in the storage system, that is, In other words, only one type of storage partition group is stored in the storage system, and the storage partition group stored in the storage system corresponds to the current topology of the storage system.

Alternatively, the controller of the storage system may also store a variety of different storage partition groups. For example, the storage system stores two types of storage partition groups, which are the M first storage partition groups of the storage system under the first topology structure. and N second storage partition groups of the storage system under the second topology. In this way, for data that has been stored according to several first storage partition groups in the M first storage partition groups, data migration may not be performed, thereby reducing resources consumed by data migration.

S48. The controller of the storage system performs data migration for the data that has been stored according to at least one first storage partition group in the M first views.

In this embodiment of the present application, according to the different ways of generating the N second storage partition groups in S46, the ways of migrating data by the controller of the storage system are also different.

For the case a of the first manner in S46, since the first storage devices at L locations in each first storage partition group and the second storage devices at corresponding L locations in the corresponding second storage partition group The device is the same storage device, that is, a part of the data stored in each of the L first storage devices corresponding to each first storage partition group needs to correspond to M second storage partition groups The data stored in one of the P second storage devices are the same, so that the controller of the storage system stores the K corresponding to one storage partition group in each of the M first storage partition groups. Part of the data in the original data of the storage device is migrated according to one second storage partition group in the N second storage partition groups, and the part of the data in each original data is the same as that in each of the L second storage devices. The data stored in the second storage devices are all different.

As an example, it can be seen from FIG. 7A that the hard disk where one PT is located in each of the 6 PTs is the same as the hard disk where the PT in the corresponding PT in the first storage partition group is located. For example, PT1 shown in FIG. 7A The hard disk where D3 is located is hard disk 3. In the view shown in Figure 6, the hard disk where D3 of PT1 is located is also hard disk 3. Therefore, for the data stored in PT1 shown in Figure 6, only D1 and D2 can be migrated , D4 can be, and can not migrate D3. Each verification part needs to be regenerated based on the migrated multiple data shards. In this way, when the topology of the storage system is expanded from 3 storage nodes to 5 storage nodes, data migration can be reduced by 1/4=25%, and resources of the storage system can be saved.

For the case b of the first manner in S46, since the P second storage devices corresponding to any one of the N second storage partition groups are different from any one of the M first storage partition groups The K first storage devices included in a storage partition group are different, so the controller of the storage system will store each original data stored in at least one first storage partition group in the M first storage partition groups according to One second storage partition group among the N second storage partition groups performs data migration.

As an example, it can be seen from FIG. 7B that the hard disk where multiple PTs in each of the 6 PTs are located is different from the hard disk where multiple PTs in the corresponding PTs in the first storage partition group are located. For example, FIG. 7A The hard disks where D1 to D4 of PT1 shown are located are different from the hard disks where D1 to D4 of PT1 are located in the view shown in FIG. 6 . Therefore, for the data stored in PT1 shown in FIG. In the hard disk corresponding to PT1 shown in FIG. 7A , each verification part needs to be regenerated according to the multiple data fragments after migration.

For the second manner in S46, reference may be made to the description for the case a of the first manner in S46, which will not be repeated here. It should be noted that, since the M first storage partition groups are considered when generating the N second storage partition groups, that is, there are more pts in the same position in each PT that fall into the same hard disk, so , compared with the data volume of the data to be migrated in the case a of the first method in S46, in this case, the data volume of the data to be migrated is less, so that the data migration can be further reduced The resources consumed by the storage system.

It should be noted that step S48 is an optional step, that is, it is not required to be executed. For example, some log-type storage systems can also be expanded without migrating data. The newly added storage nodes and hard disks will write more data in the future to gradually equalize the capacity. For example, the second view is based on the original first view. Generated, before the view is changed, four partitions D5, D6, D7, and D8 are added, and the data stored in the storage system according to the first view is gradually adapted to the second view along with garbage collection. In FIG. 4 , step S48 is indicated as an optional step by a dotted line.

S49. After receiving the new data to be stored, the controller of the storage system stores the new data to be stored in a second storage device corresponding to at least one second storage partition group in the N second storage partition groups .

After the controller of the storage system generates N second storage partition groups, when receiving the data to be stored, data storage is performed according to the N second storage partition groups, and the specific data storage process is the same as that in step S42, It is not repeated here.

In the above technical solution, the storage system supports dynamic storage partition groups (or views). When the topology of the storage system changes, the storage partition groups (or views) of the storage system also change. For example, when the storage system expands Afterwards, the data in the storage system can be synchronized to a higher data redundancy mode for data storage, thereby improving the space utilization of the storage system.

In the embodiment shown in FIG. 4 , the processing procedure when the storage system is expanded is described. The following describes the process of reducing the capacity of the storage system. Please refer to FIG. 8 , which is a flowchart of another example of the data storage method in the embodiment of the present application. In the following description, the method is used in the storage system shown in FIG. 3C as an example, that is, the method is executed by the controller in the storage system shown in FIG. 3C .

S81. The controller of the storage system generates and stores M first storage partition groups corresponding to the first topology structure, and each first storage partition group in the M first storage partition groups corresponds to K first storage devices.

As an example, the storage system shown in FIG. 3C includes 5 storage nodes, each storage node includes 4 hard disks, a total of 20 hard disks, and the hard disks are labeled as hard disk 1 to hard disk 20 in sequence. The data redundancy mode configured by the storage system is EC8+2 mode. Assuming that the storage system creates 6 PTs, each PT includes 10 PTs, of which the data slice occupies 8 partitions, and the inspection part occupies 2 partitions. Thus, the PT shown in FIG. 9 is obtained.

In the embodiment of the present application, taking the controller using a polling algorithm to map 6 PTs to 20 hard disks as an example, the first view shown in FIG. 10 is obtained. In Figure 10, 10 pts corresponding to PT1, PT3 and PT5 are mapped in HDD 1 to HDD 10, among which 8 data disks are HDD 1 to HDD 8, HDD 9 and HDD 10 are inspection disks, PT2, PT4 and The 10 pts corresponding to the PT6 are mapped in the hard disks 11 to 20 , among which 8 data disks are the hard disks 11 to 18 , and the hard disks 19 and 20 are test disks.

S82. After receiving the data to be stored, the controller of the storage system determines to store the data to be stored in at least one partition group of the M first storage partition groups.

S83. The controller of the storage system sends a data storage instruction to at least one storage node, the at least one storage node receives the data storage instruction, and the at least one storage node receives the data storage instruction and executes the data storage instruction.

S84. Each storage node in the at least one storage node sends a heartbeat to the controller of the storage system, and the controller of the storage system receives the heartbeat.

S85. The controller of the storage system determines the capacity reduction of the storage system.

In this embodiment of the present application, the capacity reduction of the storage system can be understood as that the topology structure of the storage system is updated from the first topology structure to the second topology structure, and the number of storage nodes in the second topology structure becomes smaller.

Steps S81 to S85 are similar to steps S41 to S45 and will not be repeated here.

S86. The controller of the storage system generates N second storage partition groups corresponding to the second topology structure, and each second storage partition group in the N second storage partition groups corresponds to P second storage devices.

When the controller of the storage system determines that the topology of the storage system has changed, the controller of the storage system may use the correspondence between the topology and the data redundancy mode shown in Table 1 to determine the data redundancy corresponding to the second topology. model. This process is similar to the corresponding content in S46, and will not be repeated here.

In this embodiment of the present application, the storage system is changed from the first topology structure of 5 storage nodes shown in FIG. 3C to the second topology structure of 3 storage nodes shown in FIG. 3A , and the controller of the storage system determines that the The data redundancy mode configured in the second topology is EC4+2, that is to say, in the second topology, a PT includes 6 pts, of which 4 pts are used to store data slices, and 2 pts are used for In the storage verification part, that is, P is less than K, the controller of the storage system generates N second storage partition groups corresponding to the second topology, which may include, but are not limited to, two ways. M first storage partition groups under the topology structure generate N second storage partition groups; in the second method, consider M first storage partition groups under the first topology structure, and generate N second storage partition groups , these two manners are similar to the corresponding contents in step S46, and are not repeated here.

The second method is used as an example to illustrate. As an example, the controller of the storage system may use the same data arrangement algorithm as in S41 for many times, such as a polling algorithm or a hash algorithm, etc., to obtain the same data as the six data. A plurality of second views corresponding to the PT, and then according to the linkage between the plurality of second views and the first view, from the plurality of second views, select the pt at the same position in each PT that falls into the same hard disk at most. A second view of , as a second view of the M second storage partition groups under the second topology.

Please refer to FIG. 11 , which is an example of the second view. In FIG. 11 , the value of N is 6. As shown in FIG. 11 , among the 4 hard disks in each second storage partition group, the hard disk numbers in two positions are the same as the hard disk numbers in the same position in the first view. For example, in the view shown in FIG. 11 , the hard disk where D3 of PT1 is located is hard disk 3, and the hard disk where D4 is located is hard disk 4. In the view shown in Figure 10, the hard disk where D3 of PT1 is located is also hard disk 3, and the hard disk where D4 is located is also hard disk 4; In the view shown in Figure 11, the hard disk where D3 of PT2 is located is hard disk 9, and the hard disk where D4 is located is hard disk 10. In the view shown in Figure 10, the hard disk where D3 of PT2 is located is also hard disk 9, and the hard disk where D4 is located is also hard disk 9. It is the hard disk 10; there are similar situations in other PTs, which are not listed one by one here.

S87. The controller of the storage system uses the N second storage partition groups to update the M first storage partition groups stored in the storage system.

S88. The controller of the storage system performs data migration for the data that has been stored according to at least one first storage partition group in the M first views.

In this embodiment of the present application, according to different ways of generating the N second storage partition groups in S86, the ways of migrating data by the controller of the storage system are also different. Taking the manner shown in FIG. 11 as an example, since the first storage devices at S positions in each first storage partition group and the second storage devices at S corresponding positions in the corresponding second storage partition group are The same storage device, that is to say, the data stored in each of the S first storage devices corresponding to each first storage partition group, the data stored in each of the P first storage devices corresponding to the M second storage partition groups. A part of the data stored in a second storage device in the two storage devices is the same, and S is a positive integer smaller than P, so that the controller of the storage system assigns each storage partition group stored in the M first storage partition groups to one storage partition group. Part of the data in the original data of the corresponding K storage devices is migrated according to one second storage partition group in the N second storage partition groups, and the partial data in each original data is the same as the P second storage partition group. Any part of data stored in each of the second storage devices in the storage devices is different.

As an example, it can be seen from FIG. 11 that the hard disk where 2 pts are located in each of the 6 PTs is the same as the hard disk where the pts in the corresponding PTs in the first storage partition group are located. For example, as shown in FIG. 11 , In the view shown, the hard disk where D3 of PT1 is located is hard disk 3, and the hard disk where D4 is located is hard disk 4. In the view shown in Figure 10, the hard disk where D3 of PT1 is located is also hard disk 3, and the hard disk where D4 is located is also hard disk. 4. Therefore, for the data stored in PT1 shown in FIG. 10, only D1, D2, D5-D8 may be migrated, and D3 and D4 may not be migrated. Each verification part needs to be regenerated based on the migrated multiple data shards.

For the data migration process corresponding to other ways of generating N second storage partition groups, please refer to the content in step S48, and details are not repeated here.

S89. After receiving the new data to be stored, the controller of the storage system stores the new data to be stored in a second storage device corresponding to at least one second storage partition group in the N second storage partition groups .

In the above technical solution, the storage system supports dynamic storage partition groups (or views). When the topology of the storage system changes, the storage partition groups (or views) of the storage system also change. Capacity reduction, due to the failure of 2 storage nodes in the 5 storage nodes, the storage system can also generate a new storage partition group to adapt to the reduced storage system, each storage partition group in the new storage partition group It can be used to improve the space utilization of the storage system.

In the above embodiments provided by the present application, the methods provided by the embodiments of the present application are respectively introduced from the perspective of interaction between a controller of a storage system and a storage node. In order to implement the functions in the methods provided by the above embodiments of the present application, the controller of the storage system may include a hardware structure and/or a software module, and implement the above functions in the form of a hardware structure, a software module, or a hardware structure plus a software module . Whether one of the above functions is performed in the form of a hardware structure, a software module, or a hardware structure plus a software module depends on the specific application and design constraints of the technical solution.

FIG. 12 shows a schematic structural diagram of a data storage device 1200 of a storage system. The data storage device 1200 may be a controller of a storage system, which can implement the functions of the controller of the storage system in the methods provided in the embodiments of the present application; the data storage device 1200 may also be a controller capable of supporting the storage system to implement the implementation of the present application. The example provides an apparatus for storing the functions of a controller of a system in a method. The data storage device 1200 may be a hardware structure, a software module, or a hardware structure plus a software module. The data storage device 1200 may be implemented by a chip system. In this embodiment of the present application, the chip system may be composed of chips, or may include chips and other discrete devices.

The data storage device 1200 may include a communication module 1201 and a processing module 1202 .

The communication module 1201 can be used to perform steps S43 and S44 in the embodiment shown in FIG. 4 , or to perform steps S83 and S84 in the embodiment shown in FIG. 8 , and/or to support the descriptions herein other processes of the technology. The communication module 1201 is used for the data storage device 1200 to communicate with other modules, and it can be a circuit, a device, an interface, a bus, a software module, a transceiver or any other device that can implement communication.

The processing module 1202 may be used to perform steps S41 to S42 and steps S45 to S49 in the embodiment shown in FIG. 4 , or to perform steps S81 to S82 and steps S85 to S85 to the embodiment shown in FIG. 8 . Step S89, and/or other processes for supporting the techniques described herein.

Wherein, all relevant contents of the steps involved in the above method embodiments can be cited in the functional descriptions of the corresponding functional modules, which will not be repeated here.

The division of modules in the embodiments of the present application is schematic, and is only a logical function division. In actual implementation, there may be other division methods. In addition, the functional modules in the various embodiments of the present application may be integrated into one processing unit. In the device, it can also exist physically alone, or two or more modules can be integrated into one module. The above-mentioned integrated modules can be implemented in the form of hardware, and can also be implemented in the form of software function modules.

FIG. 13 shows a data storage device 1300 of a storage system provided by an embodiment of the present application, wherein the data storage device 1300 may be the controller of the storage system in the embodiment shown in FIG. 4 or FIG. The function of the controller of the storage system in the method provided in the embodiment of FIG. 4 of the present application; the data storage device 1300 may also be a controller capable of supporting the storage system to implement the storage in the method provided by the embodiment shown in FIG. 4 or FIG. 8 of the present application. A device that functions as a controller of a system. Wherein, the data storage device 1300 may be a chip system. In this embodiment of the present application, the chip system may be composed of chips, or may include chips and other discrete devices.

The data storage device 1300 includes at least one processor 1320 for implementing or supporting the data storage device 1300 to implement the function of the controller of the storage system in the method provided by the embodiment shown in FIG. 4 or FIG. 8 of the present application. Exemplarily, the processor 1320 may generate N second storage partition groups corresponding to the second topology. For details, please refer to the detailed description in the method example, which will not be repeated here.

Data storage device 1300 may also include at least one memory 1330 for storing program instructions and/or data. Memory 1330 and processor 1320 are coupled. The coupling in the embodiments of the present application is an indirect coupling or communication connection between devices, units or modules, which may be in electrical, mechanical or other forms, and is used for information exchange between devices, units or modules. Processor 1320 may cooperate with memory 1330. Processor 1320 may execute program instructions stored in memory 1330 . At least one of the at least one memory may be included in the processor. When the processor 1320 executes the program instructions in the memory 1330, the method shown in FIG. 4 or FIG. 8 may be implemented.

The data storage device 1300 may also include a communication interface 1310 for communicating with other devices through a transmission medium, so that the devices used in the data storage device 1300 may communicate with other devices. Illustratively, the other device may be a client. The processor 1320 may use the communication interface 1310 to send and receive data.

The specific connection medium between the communication interface 1310 , the processor 1320 , and the memory 1330 is not limited in the embodiments of the present application. In the embodiment of the present application, the memory 1330, the processor 1320, and the communication interface 1310 are connected through a bus 1340 in FIG. 13. The bus is represented by a thick line in FIG. 13, and the connection between other components is only for schematic illustration. , is not limited. The bus can be divided into an address bus, a data bus, a control bus, and the like. For ease of presentation, only one thick line is used in FIG. 13, but it does not mean that there is only one bus or one type of bus.

In this embodiment of the present application, the processor 1320 may be a general-purpose processor, a digital signal processor, an application-specific integrated circuit, a field programmable gate array or other programmable logic device, a discrete gate or transistor logic device, or a discrete hardware component, which may implement Alternatively, each method, step, and logic block diagram disclosed in the embodiments of the present application are executed. A general purpose processor may be a microprocessor or any conventional processor or the like. The steps of the methods disclosed in conjunction with the embodiments of the present application may be directly embodied as executed by a hardware processor, or executed by a combination of hardware and software modules in the processor.

In this embodiment of the present application, the memory 1330 may be a non-volatile memory, such as a hard disk drive (HDD) or a solid-state drive (SSD), etc., or may be a volatile memory (volatile memory), such as random access Access memory (random-access memory, RAM). Memory is, but is not limited to, any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer. The memory in this embodiment of the present application may also be a circuit or any other device capable of implementing a storage function, for storing program instructions and/or data.

Embodiments of the present application also provide a computer-readable storage medium, including instructions, which, when executed on a computer, cause the computer to execute the method executed by the controller of the storage system in the embodiment shown in FIG. 4 or FIG. 8 .

Embodiments of the present application also provide a computer program product, including instructions, which when run on a computer, cause the computer to execute the method executed by the controller of the storage system in the embodiment shown in FIG. 4 or FIG. 8 .

An embodiment of the present application provides a chip system, where the chip system includes a processor, and may further include a memory, for implementing the function of the controller of the storage system in the foregoing method. The chip system can be composed of chips, and can also include chips and other discrete devices.

The methods provided in the embodiments of the present application may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present application are generated. The computer may be a general purpose computer, a special purpose computer, a computer network, network equipment, user equipment, or other programmable apparatus. The computer instructions may be stored in or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions may be downloaded from a website site, computer, server or data center Transmission to another website site, computer, server or data center via wired (eg coaxial cable, optical fiber, digital subscriber line, DSL for short) or wireless (eg infrared, wireless, microwave, etc.) A computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. that contains an integration of one or more available media. The available media can be magnetic media (eg, floppy disks, hard disks, magnetic tape), optical media (eg, digital video disc (DVD) for short), or semiconductor media (eg, SSD), and the like.

Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims (17)

1. a data storage method of a storage system, characterized in that, comprising: When the topology of the storage system is updated from the first topology to the second topology, N second storage partition groups corresponding to the second topology are generated, and each of the N second storage partition groups The second storage partition group corresponds to P second storage devices, and N and P are positive integers; wherein, the first topology is different from the second topology; Using the N second storage partition groups to update the M first storage partition groups stored in the storage system, the M first storage partition groups correspond to the first topology, and the M first storage partition groups Each first storage partition group in the storage partition group corresponds to K first storage devices, and M and K are positive integers; the values of P and K are different; After receiving the new data to be stored, the new data to be stored is stored in a second storage device corresponding to at least one of the N second storage partition groups. 2 . The method according to claim 1 , wherein the value of P is the same as the data redundancy mode configured by the storage system under the second topology structure, and the value of K is the same as that of the storage system in the second topology. 3 . The data redundancy modes configured in the first topology structure are the same. 3. The method according to claim 1 or 2, wherein when P is greater than K, each first storage device in the L first storage devices corresponding to each first storage partition group stores A part of the data is the same as the data stored in one of the P second storage devices corresponding to the M second storage partition groups, and L is a positive integer smaller than P. 4. The method according to claim 3, wherein the method further comprises: Part of the data in each original data is migrated according to one second storage partition group in the N second storage partition groups, and each original data is stored in the M first storage partition groups. For K storage devices corresponding to one storage partition group, part of data in each of the original data is different from data stored in each of the L second storage devices. 5. The method according to claim 1 or 2, wherein when P is less than K, each first storage device in the S first storage devices corresponding to each first storage partition group stores The data is the same as a part of data stored in one of the P second storage devices corresponding to the M second storage partition groups, and S is a positive integer smaller than P. 6. The method according to claim 5, wherein the method further comprises: Part of the data in each original data is migrated according to one second storage partition group in the N second storage partition groups, and each original data is stored in the M first storage partition groups. For K storage devices corresponding to one storage partition group, part of data in each original data is different from any part of data stored in each of the P second storage devices. 7. The method according to claim 1 or 2, wherein the P second storage devices corresponding to any one of the N second storage partition groups are different from the M second storage devices. The K first storage devices included in any one of the first storage partition groups in a storage partition group are different. 8. The method according to claim 7, wherein the method further comprises: Data migration is performed for each original data according to one second storage partition group in the N second storage partition groups, and the each original data is that the storage system in the first topology structure, according to the At least one first storage partition group in the M first storage partition groups stores data in the storage system. 9. A data storage device for a storage system, comprising a communication interface and a processor, wherein: The processor is configured to generate N second storage partition groups corresponding to the second topology structure when the topology structure of the storage system is updated from the first topology structure to the second topology structure, and the N second topological structure Each second storage partition group in the storage partition group corresponds to P second storage devices, and N and P are positive integers; wherein, the first topology structure is different from the second topology structure; The processor is further configured to use the N second storage partition groups to update the M first storage partition groups stored in the storage system, the M first storage partition groups and the first topology structure Correspondingly, each first storage partition group in the M first storage partition groups corresponds to K first storage devices, and M and K are positive integers; P and K have different values; The processor is further configured to, after receiving the new data to be stored through the communication interface, store the new data to be stored in at least one second storage partition with the N second storage partition groups in the second storage device corresponding to the group. 10 . The apparatus according to claim 9 , wherein the value of P is the same as the data redundancy mode configured by the storage system under the second topology structure, and the value of K is the same as that of the storage system in the second topology structure. 11 . The data redundancy modes configured in the first topology structure are the same. 11. The apparatus according to claim 9 or 10, wherein when P is greater than K, each first storage device in the L first storage devices corresponding to each first storage partition group stores A part of the data is the same as the data stored in one of the P second storage devices corresponding to the M second storage partition groups, and L is a positive integer smaller than P. 12. The apparatus of claim 11, wherein the processor is further configured to: Part of the data in each original data is migrated according to one second storage partition group in the N second storage partition groups, and each original data is stored in the M first storage partition groups. For K storage devices corresponding to one storage partition group, part of data in each of the original data is different from data stored in each of the L second storage devices. 13. The apparatus according to claim 9 or 10, wherein when P is less than K, each first storage device in the S first storage devices corresponding to each first storage partition group stores The data is the same as a part of data stored in one of the P second storage devices corresponding to the M second storage partition groups, and S is a positive integer smaller than P. 14. The apparatus of claim 13, wherein the processor is further configured to: Part of the data in each original data is migrated according to one second storage partition group in the N second storage partition groups, and each original data is stored in the M first storage partition groups. For K storage devices corresponding to one storage partition group, part of data in each original data is different from any part of data stored in each of the P second storage devices. 15. The apparatus according to claim 9 or 10, wherein the P second storage devices corresponding to any one of the N second storage partition groups are different from the M second storage devices. The K first storage devices included in any one of the first storage partition groups in a storage partition group are different. 16. The apparatus of claim 15, wherein the processor is further configured to: Data migration is performed for each original data according to one second storage partition group in the N second storage partition groups, and the each original data is that the storage system in the first topology structure, according to the At least one first storage partition group in the M first storage partition groups stores data in the storage system. 17. A computer storage medium, wherein the computer storage medium stores instructions that, when executed on a computer, cause the computer to execute the method according to any one of claims 1-8.

Images