WO2015035768A1

WO2015035768A1 - System expansion method and apparatus for redundant array of independent disks (raid)

Info

Publication number: WO2015035768A1
Application number: PCT/CN2014/074775
Authority: WO
Inventors: 曾华荣; 魏明昌
Original assignee: 华为技术有限公司
Priority date: 2013-09-16
Filing date: 2014-04-04
Publication date: 2015-03-19
Also published as: CN103513938B; CN103513938A

Abstract

Embodiments of the present invention relate to the technical field of information networks. Provided are a system expansion method and apparatus for a redundant array of independent disks (RAID). The method and the apparatus can improve the utilization rate of system storage space and ensure a minimum data migration volume. The method comprises: when a storage node is added to an RAID system, calculating a new virtual storage hard disk distribution table according to a parameter of a virtual storage hard disk group in a current RAID system and to a distribution algorithm; when the new virtual storage hard disk distribution table is compared with an original virtual storage disk distribution table, a number X of data virtual storage hard disks in each virtual storage hard disk group of the current RAID system remains unchanged, and a number Y of verification virtual storage hard disks changes; and migrating or deleting the virtual storage hard disk in the virtual storage hard disk group according to the distribution location of the virtual storage hard disk in the new virtual storage hard disk distribution table.

Description

Redundant array of independent disks RAID system expansion method and device

Technical field

The present invention relates to the field of information network technologies, and in particular, to a method and an apparatus for expanding a redundant array of independent disks.

Background technique

In a communication network, each node device has a separate storage unit, which may be a local storage hard disk or a local LUN (Logical Unit Number), or may be a serial attached SCSI (SAS). Row connection SCSI

(Small Computer System Interface) The other dedicated storage unit belonging to this node that is connected to the bus. And in order to solve the data reliability and performance problems of a single disk, people have proposed RAID

(edundant Arrays of Inexpensive Disks, ¹ J) technology, which runs "software with inter-node RAID technology" on each node device, thus using each node device as a storage node and storing these The nodes form a virtual RAID system. Specifically, the technology is responsible for coordinating the storage disks of the storage nodes and organizing them into a storage space that can be uniformly accessed. The storage space here can be regarded as a virtual single large-capacity hard disk. At the same time, in order to meet the constraints of the RAID system formed by many storage nodes, the virtualized inter-node RAID scheme is usually used to organize the data. Specifically, the storage hard disk on each storage node is stored in a certain size. The hard disk space is divided into multiple virtual storage disks, and redundant relationships are established between the virtual storage disks on the different storage disks. Therefore, when data in a storage hard disk of a storage node is lost or damaged, the data of the storage node can be restored according to the redundancy relationship described above.

When a new storage node needs to be added to the RAID system, that is, when the RAID system is expanded, the existing technology generally maintains the redundancy relationship between the original virtual storage hard disks. The processing of data balancing between storage nodes makes the storage space utilization of the RAID system too low and the data migration amount is too large.

Therefore, in the case of RAID system expansion, how to ensure high RAID system space utilization while reducing data migration is a problem that the industry is expected to solve. Summary of the invention

Embodiments of the present invention provide a method and apparatus for expanding a redundant array of independent disks, thereby improving the utilization of the storage space of the system and reducing the amount of data migration.

In order to achieve the above object, embodiments of the present invention use the following technical solutions:

The first aspect provides a method for expanding a redundant array of independent disks, wherein the RAID system includes N storage nodes, and each of the N storage nodes is divided into virtual storage disks of the same size. The M virtual storage disks constitute a virtual storage disk group having a redundant relationship, and the virtual storage disk group is composed of an X block data virtual storage hard disk and a Y block parity virtual storage hard disk, M=X+Y, the data. The virtual storage hard disk stores the data block, and the verification virtual storage hard disk stores the verification block, and the method includes:

When the storage system is added to the RAID system, the new virtual storage hard disk distribution table is calculated according to the parameter and the distribution algorithm of the virtual storage disk group in the current RAID system; the new virtual storage hard disk distribution table is compared with the original virtual The storage disk distribution table, in which the number of data virtual storage disks in each virtual storage disk group in the current RAID system remains unchanged, and the number Y of the virtual storage disks is changed.

The virtual storage hard disks in the virtual storage disk group are migrated or deleted according to the distribution location of the virtual storage disks in the new virtual storage disk distribution table.

In a first possible implementation manner, according to the first aspect, when a storage node is added to the RAID system, a new virtual storage hard disk is calculated according to a parameter and a distribution algorithm of the virtual storage disk group in the current RAID system. The distribution table includes:

Obtaining parameters of the virtual storage disk group when the RAID system adds the storage node;

Calculating a new virtual storage hard disk distribution table of the virtual storage disk group by using the distribution algorithm and parameters of the virtual storage disk group.

In a second possible implementation manner, in combination with the first aspect or the first possible implementation manner, the parameters of the virtual storage disk group include at least one of the following: a storage node address ID list, a storage node weight list, a The identifier of the virtual storage disk group and the number of virtual storage disks in the virtual storage disk group; wherein the storage node weight list is for each storage The storage space of the storage node currently storing the hard disk.

In a third possible implementation manner, according to the first aspect, the virtual storage hard disk in the virtual storage disk group is migrated or deleted according to the distribution location of the virtual storage hard disk in the new virtual storage disk distribution table. The method includes: determining, according to the new virtual storage hard disk distribution table, whether the current virtual storage hard disk is a verified virtual storage hard disk that needs to be deleted;

If the current virtual storage hard disk is not a verification virtual storage hard disk that needs to be deleted, determining, according to the new virtual storage hard disk distribution table, whether the current virtual storage hard disk is located and the new virtual storage hard disk distribution table. The position of the match;

If the current virtual storage hard disk is a verification virtual storage hard disk that needs to be deleted, the verification data in the current virtual storage hard disk and the current virtual storage hard disk is deleted.

In a fourth possible implementation, according to the third possible implementation manner, determining, according to the new virtual storage hard disk distribution table, whether the current virtual storage hard disk is located and the new virtual storage hard disk After the locations in the distribution table are consistent, the method further includes: if the current virtual storage hard disk does not match the location in the virtual storage hard disk distribution table, copying the data in the current virtual storage hard disk to the newly added storage Deleting the data in the current virtual storage hard disk and the current virtual storage hard disk;

If the current virtual storage hard disk matches the location in the virtual storage hard disk distribution table, the current virtual storage hard disk location is kept unchanged.

In a second aspect, a RAID expansion system for a redundant array of independent disks is provided. The RAID system includes N storage nodes, and each of the N storage nodes is divided into virtual storage disks of the same size. The M virtual storage disks constitute a virtual storage disk group having a redundant relationship, and the virtual storage disk group is composed of an X block data virtual storage hard disk and a Y block parity virtual storage hard disk, M=X+Y, the data. The virtual storage hard disk stores a data block, and the verification virtual storage hard disk stores a check block, where the device includes:

a distribution table calculation unit, configured to calculate a new virtual node according to a parameter and a distribution algorithm of the virtual storage disk group in the current RAID system when the storage node is added to the RAID system The storage virtual disk distribution table; the new virtual storage hard disk distribution table is compared with the original virtual storage hard disk distribution table, and the number of data virtual storage hard disks in each virtual storage disk group in the current RAID system remains unchanged, and the verification is performed. The number of virtual storage hard disks Y changes;

And a migration unit, configured to migrate or delete the virtual storage hard disk in each virtual storage disk group according to a distribution location of each virtual storage disk in the new virtual storage disk distribution table calculated by the distribution table computing unit.

In a first possible implementation manner, according to the second aspect, the distribution table calculation unit includes:

a parameter obtaining module, configured to obtain a parameter of the virtual storage disk group when the storage node is added to the RAID system;

a distribution table calculation module, configured to calculate, by using the distribution algorithm and parameters of the virtual storage disk group obtained by the parameter obtaining unit, the new virtual storage disk distribution table of the virtual storage disk group.

In a second possible implementation manner, in combination with the second aspect or the first possible implementation manner, the parameters of the virtual storage disk group include at least one of the following: a storage node address ID list, a storage node weight list, a The identifier of the virtual storage disk group and the number of virtual storage disks in the virtual storage disk group; wherein the storage node weight list is the usage rate of the storage space of the current storage disk of each storage node.

In a third possible implementation, according to the second aspect, the migration unit includes: a first determining module, configured to determine, according to the new virtual storage hard disk distribution table, whether the current virtual storage hard disk is a school that needs to be deleted Verify the virtual storage hard disk;

a second determining module, configured to determine, according to the new virtual storage hard disk distribution table, whether the location of the current virtual storage hard disk and the new virtual storage hard disk are distributed according to the new virtual storage hard disk distribution table The positions in the table match;

And a migration module, configured to delete the verification data in the current virtual storage hard disk and the current virtual storage hard disk if the first determining module determines that the result is yes.

In a fourth possible implementation, according to the third possible implementation manner, the migration module is further configured to: if the second determining module determines that the result is negative, The data in the current virtual storage hard disk is copied to the storage hard disk of the newly added storage node, and the data in the current virtual storage hard disk and the current virtual storage hard disk at the original location is deleted; If the judgment result of the second judging module is yes, the position of the current virtual storage hard disk is kept unchanged.

The RAID system expansion method and device provided by the embodiment of the present invention, when the RAID system is expanded, the new calculation is performed according to the parameters and the distribution algorithm of the virtual storage disk group in the current RAID system (that is, the RAID system after the storage node is added) The virtual storage disk distribution table is such that the number of data virtual storage disks in each virtual storage disk group in the current RAID system is unchanged, the number of virtual storage disks Y is changed, and then the virtual disks in the new virtual storage disk are distributed. The storage location of the storage disk is used to migrate or delete the virtual storage disk in each virtual storage disk group, thereby improving the utilization of the system storage space, and compared with the prior art, in the case of the same space utilization, The invention balances the data of the expanded RAID system by keeping the number of X unchanged, and only modifying the number of Ys, thereby greatly reducing the amount of data and the amount of calculation.

DRAWINGS

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly described below. It is obvious that the drawings in the following description are only some embodiments of the present invention. Other drawings may also be obtained from those of ordinary skill in the art in view of the drawings.

1 is a schematic flowchart of a method for expanding a RAID system according to an embodiment of the present invention; FIG. 2 is a schematic flowchart of another method for expanding a RAID system according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of a virtual storage hard disk distribution of a virtual storage disk group according to an embodiment of the present disclosure;

4 is a schematic diagram of a new distribution of a virtual storage hard disk of another virtual storage disk group according to an embodiment of the present invention;

FIG. 5 is a schematic structural diagram of a RAID system expansion device according to an embodiment of the present invention; FIG. 6 is a schematic structural diagram of another RAID system expansion device according to an embodiment of the present invention; Figure

FIG. 7 is a schematic structural diagram of a RAID system expansion device according to another embodiment of the present invention.

detailed description

The technical solutions in the embodiments of the present invention will be clearly described in conjunction with the drawings in the embodiments of the present invention. It is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present invention without creative efforts are within the scope of the present invention.

Nowadays, to solve the problems of reliability and performance in a single block of data in the disk, there has been RAID (Redundant Arrays of Inexpensive Disks, Redundant Array of Independent Disks bad ¹ J) technique, i.e. the use of the hard disk technology to store a plurality of RAID storage node Form a virtual RAID system that can be accessed uniformly. Specifically, the present invention is applicable to redundancy between node devices, and is also applicable to redundancy of a single storage hard disk in a single node device. Therefore, when the storage node is a node device, the RAID system is composed of several The storage hard disk corresponding to each node device in the node device is composed; when the storage node is any single storage hard disk in one node device, the RAID system is composed of each single storage hard disk in one node device. The foregoing node device is a variety of terminal devices (eg, a PC, a server, etc.), and the storage hard disk corresponding to each node device is composed of at least one single hard disk.

Specifically, the foregoing RAID system includes N storage nodes. And each storage node is composed of at least one storage hard disk. The storage hard disk may be a local hard disk or a local LUN (Logical Unit Number), or may be connected through a network or a Serial Attached SCSI (SAS) bus. The dedicated storage unit for this node. Secondly, each of the N storage nodes is divided into virtual storage disks of the same size. The virtual storage disks may be referred to as CK (chunk), and the M CKs constitute virtual storage with a certain redundancy relationship. CKG (chunk Group), wherein the number of CKs in each CKG in the RAID system is not necessarily the same, and the number of CKs selected from each storage hard disk is not necessarily the same for each CKG. Specifically, the above CKG is virtual stored by X block data The storage hard disk and the Y-block verification virtual storage hard disk are composed of Μ=Χ+Υ, wherein the data virtual storage hard disk stores the data block, and the verification virtual storage hard disk stores the verification block. Based on the above description of the RAID system, the present invention provides a new RAID system expansion method and apparatus.

Specifically, as shown in FIG. 1 , the RAID system expansion method specifically includes the following steps:

101. When a storage node is added to the RAID system, the RAID system expansion device calculates a new virtual storage hard disk distribution table according to the parameters and the distribution algorithm of the virtual storage disk group in the current RAID system.

The parameter of the virtual storage disk group includes at least one of the following: a storage node address ID list, a storage node weight list, an identifier of the virtual storage disk group, and a number of virtual storage disks in each virtual storage disk group; The storage node weight list is the usage rate of the storage space of the storage disk of the current storage node. The identifier of the virtual storage disk group is the identifier of each virtual storage disk group of the RAID system, and each of the RAID systems The number of virtual storage disks in a virtual storage disk group can be different. The above-mentioned new virtual storage hard disk distribution table includes the distribution position of each virtual storage hard disk in the current RAID system (that is, the RAID system after the storage node is added), specifically, the above-mentioned new virtual storage hard disk distribution table is compared. The original virtual storage disk distribution table. The number of data virtual storage disks in each virtual storage disk group in the current RAID system remains unchanged. The number of virtual storage disks Y is changed. In addition, when expanding a RAID system, you can add not only one storage node but also multiple storage nodes at the same time.

After the storage node is added to the RAID system, the parameters of each virtual storage disk are obtained, and the new virtual storage disk distribution table is calculated according to the parameters and the distribution algorithm of the obtained virtual storage disk group. At this time, the obtained new virtual storage hard disk distribution table is compared with the original virtual storage hard disk distribution table, and according to the calculation formula of the space utilization ratio of the RAID system (X/(X+Y)), the present invention can be seen. By keeping the number of data virtual storage disks X in each virtual storage disk group unchanged, and dynamically adjusting the number of verified virtual storage disks Y, thereby increasing the space utilization of the RAID system, and because the number of X is maintained, Change, only modify the number of Y, reduce the amount of data that needs to be migrated, and in addition, since the redundancy algorithm supports the reduction of Y, it is not necessary to recalculate the check, thereby ensuring high space utilization, so that The amount of computation of the RAID system is greatly reduced. 102. The AID system expansion device migrates or deletes the virtual storage hard disk in each virtual storage disk group according to the distribution location of each virtual storage disk in the new virtual storage disk distribution table.

Specifically, after the storage node is added to the RAID system, part of the data of the original storage node needs to be migrated to the newly added storage node to maintain the balance of the data of each storage node. Before the data migration, the data needs to be newly acquired. The parameters of the virtual storage hard disk are calculated by a distribution algorithm to calculate a new virtual storage hard disk distribution table and compared with the current virtual storage hard disk distribution to know which data needs to be migrated or deleted.

The RAID system expansion method provided by the embodiment of the present invention, when the RAID system is expanded, the new virtual storage calculated according to the parameters and the distribution algorithm of the virtual storage disk group in the current RAID system (that is, the RAID system after the storage node is added) The disk distribution table is such that the number of data virtual storage disks in each virtual storage disk group in the current RAID system is unchanged. The number of virtual storage disks Y is changed, and then the virtual storage disks in the new virtual storage disk distribution table are followed. The distribution location of the virtual storage hard disk in each virtual storage disk group is migrated or deleted, thereby improving the utilization of the system storage space, and compared with the prior art, in the case of the same space utilization, due to the present invention By keeping the number of X unchanged, only the number of Ys is modified to perform data equalization on the expanded RAID system, thereby greatly reducing the amount of data and the amount of computation.

An embodiment of the present invention provides a method for expanding a redundant array of independent disks. As shown in FIG. 2, the method for expanding a volume includes the following steps:

201. When a storage node is added to the RAID system, the RAID system expansion device obtains parameters of the virtual storage disk group.

Specifically, after the storage node is added to the RAID system, the RAID system expansion device first acquires a storage object directory stored in the current RAID system (that is, a RAID system after the storage node is added), where the storage object directory includes the storage in the RAID system. The storage address of each storage object sequentially reads the address of each storage object in the storage object directory, thereby sequentially traversing each virtual storage disk group corresponding to each storage object address in the storage object directory, and further Obtain the storage disk group parameters corresponding to each storage disk group in the RAID system. Specifically, taking any storage object as an example, first, reading the address of any storage object from the storage object directory, thereby searching according to the storage corresponding address. All the virtual storage disk groups in which the storage object data is stored are stored, and then each virtual storage disk group corresponding to any storage object is traversed in order to obtain the storage disk group of each storage disk group of the storage object. parameter. Of course, when the number of the virtual storage disks in the virtual storage disk is the same, the same storage object only needs to obtain the virtual storage disk group parameter of the virtual storage disk group corresponding to the storage object. The parameters of all virtual storage disk groups corresponding to the storage object. The foregoing storage object includes any one or more of the following: a file, a logical unit LUN, and an object. It should be noted that the user can access all the storage objects in the RAID system through each storage node.

Specifically, the new virtual storage hard disk distribution table is compared with the original virtual storage hard disk distribution table. The number of data virtual storage hard disks in each virtual storage disk group in the current RAID system remains unchanged, and the number of virtual storage hard disks is verified. Y changes. In addition, when expanding a RAID system, you can add not only one storage node but also multiple storage nodes at the same time.

202. The AID system expansion device calculates a new virtual storage hard disk distribution table of the virtual storage disk group by using a distribution algorithm and parameters of the virtual storage disk group.

Wherein, the new virtual storage hard disk distribution table is compared with the original virtual storage hard disk distribution table, and according to the calculation formula of the space utilization ratio of the RAID system (X/(X+Y)), the present invention is maintained by The number of data virtual storage disks X in each virtual storage disk group is unchanged, and the number of verified virtual storage disks Y is dynamically adjusted, thereby improving the space utilization of the RAID system, and since the number of Xs is kept unchanged, only Modifying the number of Y reduces the amount of data that needs to be migrated. In addition, since the redundancy algorithm supports the reduction of Y, it is not necessary to recalculate the check, thereby ensuring high space utilization, making the RAID system The amount of calculation is greatly reduced.

Optionally, the parameters of the virtual storage disk group include the following at least one of the following: a storage node address ID list, a storage node weight list, an identifier of the virtual storage disk group, and a number of virtual storage disks in each virtual storage disk group. The storage node weight list is the usage rate of the storage space of the current storage disk of each storage node, and the identifier of the virtual storage disk group is the identifier of each virtual storage disk group of the RAID system, and the RAID system The number of virtual storage disks in each virtual storage disk group can be different.

203. The AID system expansion device determines the current state according to the new virtual storage hard disk distribution table. Whether the virtual storage disk in the virtual storage disk group is the verification virtual storage disk that needs to be deleted.

If the current virtual storage hard disk is not a verification virtual storage hard disk that needs to be deleted, go to the step

204. If the current virtual storage hard disk is a verification virtual storage hard disk that needs to be deleted, then go to step 207.

Specifically, when the virtual storage hard disk in the virtual storage disk group is determined to be a virtual storage hard disk to be deleted, the virtual storage hard disk is traversed by each virtual storage hard disk in the current RAID system. In the traversal, the storage hard disk of a storage node may be sequentially traversed according to the order of the array or randomly, or the storage hard disk of each storage node may be traversed at the same time; then, when traversing the storage hard disk of a storage node, Each virtual storage disk group is traversed in turn, and each virtual storage disk group can also be traversed in parallel at the same time. When traversing a virtual storage disk group, you can traverse each virtual storage disk in the virtual storage disk group in turn, or traverse each virtual storage disk in the virtual storage disk group in parallel.

204. The AID system expansion device determines, according to the new virtual storage hard disk distribution table, whether the location of the current virtual storage hard disk matches the position in the new virtual storage hard disk distribution table.

If the current virtual storage hard disk does not match the location in the virtual storage hard disk distribution table, then go to step 205. If the current virtual storage hard disk matches the position in the virtual storage hard disk distribution table, then go to step 206.

205. The RAID system expansion device copies the data in the current virtual storage hard disk to the storage hard disk of the newly added storage node, and deletes the data in the current virtual storage hard disk and the current virtual storage hard disk at the original location.

Specifically, if the virtual storage hard disk CK0 in the existing distribution table of the virtual storage hard disk is located at the storage node a, the virtual storage hard disk CK1 is located at the storage node b. In the new virtual storage hard disk distribution table, the CK0 is located at the storage node c, and the CK1 is at the storage node d. Then CK0 needs to migrate from storage node a to storage node c, and CK1 needs to migrate from storage node b to storage node d.

206. The RAID system expansion device keeps the current virtual storage hard disk position unchanged.

207. The RAID system expansion device deletes the verification data in the current virtual storage hard disk and the current virtual storage hard disk. Specifically, after the storage node is added to the RAID system, part of the data of the original storage node is migrated to the newly added storage node to maintain the balance of the data of each storage node. The parameters of the storage disk group are calculated, and a new virtual storage disk distribution table is calculated by using a distribution algorithm, and the new virtual storage disk distribution table is compared with the specific distribution location of the current virtual storage disk, so as to know which data needs to be migrated. Those that need to be deleted to implement the expansion method of the present invention.

It should be noted that the expansion method provided by the present invention is applicable not only to a RAID system composed of multiple storage nodes, but also to a RAID system composed of multiple storage hard disks in a single storage node. There is no limitation here, and the expansion method of the present invention can be implemented as long as the scenario of the RAID system is satisfied.

In addition, when the user needs to write data in a file newly created in the RAI D system, firstly, the location of the file in the RAID system, that is, the file address of the file, is required, and secondly, each location corresponding to the location is calculated. The virtual storage disk group corresponds to which storage node, and the identifier of each virtual storage disk group, and then each virtual disk corresponding to the file is viewed according to the parameters of each virtual storage disk group corresponding to the obtained file. Whether the storage disk group is established, if it is established, write data to the corresponding location in the virtual storage disk group, and update the verification; if the RAID system does not find each virtual storage disk corresponding to the file The identifier of the group is calculated by using a distribution algorithm, and the virtual storage hard disk in the virtual storage disk group is specifically distributed on which storage node, and then the virtual storage hard disk is created in each storage node, and each file corresponding to the file is marked. A virtual storage disk group has been created, and finally the data is written to each virtual storage corresponding to the file. The corresponding position in the disk stack, and update the checksum.

The RAID system expansion method provided by the embodiment of the present invention, when the RAID system is expanded, the new virtual storage calculated according to the parameters and the distribution algorithm of the virtual storage disk group in the current RAID system (that is, the RAID system after the storage node is added) The disk distribution table is such that the number of data virtual storage disks in each virtual storage disk group in the current RAID system is unchanged. The number of virtual storage disks Y is changed, and then the virtual storage disks in the new virtual storage disk distribution table are followed. The distribution location of the virtual storage hard disk in each virtual storage disk group is migrated or deleted, thereby improving the utilization of the system storage space, and compared with the prior art, in the case of the same space utilization, due to the present invention By keeping the number of X not Change, only modify the number of Y to data equalization of the expanded RAID system, thereby greatly reducing the amount of data transferred and the amount of calculation.

In this embodiment, the virtual storage hard disk CK distribution table of the virtual storage disk group CKG shown in FIG. 3 is taken as an example. As can be seen from FIG. 3, the CKG is composed of 12 CKs of three storage nodes, wherein the redundancy relationship of each CK in the CKG is 8+4. After adding a storage node, a new CK distribution table of the CKG as shown in FIG. 4 is obtained by calculation. From FIG. 4, it can be seen that the redundancy relationship of each CK in the CKG becomes 8+3, so that the space utilization rate is 8/12 changed to 8/11. When the storage node is added again, the space utilization can be increased again. Of course, when the storage node is increased to a certain extent, the space utilization cannot be improved.

Specifically, the capacity expansion and data migration process of this embodiment is as follows:

First, according to the above CKG, the original redundancy relationship of each CK is 8+4. After calculation by the CK distribution algorithm, a new CK distribution table of each CK in the CKG can be obtained, and the existing CK distribution position is as shown in FIG. The comparison of the new CK distribution table shown shows that CK0~CK2 on the storage node 0 on the CKG, CK4~CK6 on the storage node 1, and CK8 CK10 on the storage node 2 do not need to be migrated because the position is unchanged.

Secondly, since the redundancy relationship of the CKG before the expansion is 8+4, when the expansion becomes 4 storage nodes, it is calculated according to the capacity equalization method (8+4) /4 that each node needs to store 3 CKs. . At this time, when one storage node fails, only three CKs are lost. Then, the number of redundant CKs of each CK in the CKG does not need to be maintained at 4, and is directly reduced to 3, and the reliability of the RAID system is not affected. Requirements (ie, one storage node is allowed to fail or two hard disks in a storage node's storage hard disk are allowed to fail at the same time). Therefore, comparing FIG. 3 with FIG. 4, it can be seen that CK3 needs to be migrated from storage node 0 to newly added storage node 3, and CK7 needs to be migrated from storage node 1 to newly added storage node 3, and CK11 is not The redundant redundancy check CK required can be deleted directly.

It should be noted that, since the number of data CK is unchanged, the number of check CKs is reduced, and the check does not need to be recalculated (since the EC algorithm supports that the number of data CK does not change and the number of check CKs decreases, the check is not recalculated) , which reduces the amount of data and the amount of calculations that are migrated. In addition, since there is no need to calculate the check, the expansion device does not need to read the CKs whose positions are unchanged, such as the data on CK0~CK2, CK4~CK6, CK8 CK10, which can save a lot of Hard disk IO (Input/Output, input/output) and network bandwidth, as well as a large number of CPU (Central Processing Unit) resources, to reduce the impact on the business when expanding. In addition, since the number of check CKs is reduced, the inter-node migration that may occur due to the deleted check CK is avoided, and the hard disk 10 and the network bandwidth are saved; the deleted check CK space can be recycled, and the system is also improved. Space utilization.

Table 1

In addition, from the space utilization change table shown in Table 1 above, it can be seen that when the storage node is increased, the data CK is kept unchanged, and the verification CK is lowered, the RAID system space utilization is remarkably increased. Therefore, when the storage node increases, the reliability of the system can be ensured by reducing the value of the check CK. At the same time, when the storage node is reduced, the reliability of the system can also be ensured by increasing the value of the check CK.

An embodiment of the present invention provides a RAID system expansion device, wherein the RAID system includes N storage nodes, and each of the N storage nodes is mapped. Divided into virtual storage disks of the same size, the M virtual storage disks form a virtual storage disk group with redundant relationship. The virtual storage disk group is composed of an X-block data virtual storage hard disk and a Y-block verification virtual storage hard disk, M=X. +Y, the data virtual storage hard disk storage data block, the verification virtual storage hard disk storage check block, as shown in FIG. 5 and 6, specifically, the RAID system expansion device is used to implement the above RAID system expansion method, As shown in FIG. 5, the expansion device 3 includes: a distribution table calculation unit 31 and a migration unit 32, wherein:

The distribution table calculation unit 3 1 is configured to calculate a new virtual storage hard disk distribution table according to parameters and a distribution algorithm of the virtual storage disk group in the current RAID system when a storage node is added in the RAID system.

The new virtual storage hard disk distribution table is compared with the original virtual storage hard disk distribution table. The number of data virtual storage hard disks in each virtual storage disk group in the current RAID system remains unchanged, and the number of virtual storage hard disks Y is verified. Variety.

The migration unit 32 is configured to migrate or delete the virtual storage hard disk in each virtual storage disk group according to the distribution location of each virtual storage disk in the new virtual storage disk distribution table calculated by the distribution table computing unit 31.

The RAID system expansion device provided by the embodiment of the present invention calculates a new virtual storage according to a parameter and a distribution algorithm of a virtual storage disk group in a current RAID system (ie, a RAID system after the storage node is added) when the RAID system is expanded. The disk distribution table is such that the number of data virtual storage disks in each virtual storage disk group in the current RAID system is unchanged. The number of virtual storage disks Y is changed, and then the virtual storage disks in the new virtual storage disk distribution table are followed. The distribution location of the virtual storage hard disk in each virtual storage disk group is migrated or deleted, thereby improving the utilization of the system storage space, and compared with the prior art, in the case of the same space utilization, due to the present invention By keeping the number of X unchanged, only the number of Ys is modified to perform data equalization on the expanded RAID system, thereby greatly reducing the amount of data and the amount of computation.

Optionally, as shown in FIG. 6, the distribution table calculation unit 31 includes: a parameter acquisition module 3 1 1 and a distribution table calculation module 312, where:

The parameter obtaining module 31 is configured to obtain parameters of the virtual storage disk group when the storage node is added to the RAID system.

a distribution table calculation module 3 12, configured to obtain by the distribution algorithm and the parameter acquisition unit 3 1 1 The parameters of the virtual storage disk group calculate the new virtual storage disk distribution table of the virtual storage disk group.

Optionally, the virtual storage disk group parameter includes the following at least one of the following: a storage node address ID list, a storage node weight list, an identifier of the virtual storage disk group, and a number of virtual storage disks in the virtual storage disk group; The node weight list is the usage rate of the storage space of the storage disk currently stored by each storage node.

Optionally, as shown in FIG. 6, the migration unit 32 includes: a first determining module 321, a second determining module 322, and a migration module 323, where:

The first determining module 321, is configured to determine, according to the new virtual storage hard disk distribution table, whether the current virtual storage hard disk is a verified virtual storage hard disk that needs to be deleted.

The second determining module 322 is configured to determine, according to the new virtual storage hard disk distribution table, whether the location of the current virtual storage hard disk matches the position in the new virtual storage hard disk distribution table, if the first determining module 321 determines that the result is no.

The migration module 323 is configured to delete the verification data in the current virtual storage hard disk and the current virtual storage hard disk if the first determination module 321 determines that the result is yes.

Optionally, the migration module 323 is further configured to: if the second determination module 322 determines that the result is no, copy the data in the current virtual storage hard disk to the storage hard disk of the newly added storage node, and Data deletion in the virtual storage hard disk and the current virtual storage hard disk; if the second determination module 322 determines that the result is yes, the current virtual storage hard disk location is kept unchanged.

The RAID system expansion device provided by the embodiment of the present invention calculates a new virtual storage according to a parameter and a distribution algorithm of a virtual storage disk group in a current RAID system (ie, a RAID system after the storage node is added) when the RAID system is expanded. The disk distribution table is such that the number of data virtual storage disks in each virtual storage disk group in the current RAID system is unchanged. The number of virtual storage disks Y is changed, and then the virtual storage disks in the new virtual storage disk distribution table are followed. The distribution location of the virtual storage hard disk in each virtual storage disk group is migrated or deleted, thereby improving the utilization of the system storage space, and compared with the prior art, in the case of the same space utilization, due to the present invention By keeping the number of X unchanged, only the number of Ys is modified to perform data equalization on the expanded RAID system, thereby greatly reducing the amount of data and the amount of computation. An embodiment of the present invention provides a RAID system expansion device, as shown in FIG. 7, wherein the RAID system includes N storage nodes, and each of the N storage nodes is divided into the same size. The virtual storage hard disk and the M virtual storage hard disks form a virtual storage disk group with redundant relationship. The virtual storage disk group is composed of an X-block data virtual storage hard disk and a Y-block verification virtual storage hard disk, M=X+Y, The data virtual storage hard disk stores the data block, and the verification virtual storage hard disk stores the verification block. The expansion device is configured to implement the foregoing expansion method of the RAID system. Specifically, the expansion device includes: the processor 41 and the communication unit 42. The processor 41 and the communication interface 42 described above are coupled together by a bus system. The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only one thick line is shown in Figure 7, but it does not mean that there is only one bus or one type of bus. among them:

The processor 41 can be: a central processing unit (CPU), an application specific integrated circuit (ASIC), a digital signal processor (DSP), an off-the-shelf programmable gate array (FPGA), or the like. Programmable logic device. The communication unit 42 is for providing instructions and data information to the processor 41, and the processor 41 is connected via a bus.

The specific functions of the above devices are as follows.

The communication unit 42 is for providing the processor 41 with data and instructions based on the RAID system. The processor 41 is configured to calculate a new virtual storage hard disk distribution table according to a parameter and a distribution algorithm of the virtual storage disk group acquired by the communication unit 42 when the storage node is added in the RAID system.

The new virtual storage hard disk distribution table is compared with the original virtual storage hard disk distribution table. The number of data virtual storage hard disks in each virtual storage disk group in the current RAID system (that is, the RAID system after the storage node is added) remains unchanged. Change, verify that the number of virtual storage hard disks Y changes.

The processor 41 is further configured to migrate or delete the virtual storage hard disk in each virtual storage disk group according to the distribution location of each virtual storage disk in the new virtual storage disk distribution table. Except.

Optionally, the processor 41 is specifically configured to: when the storage node is added to the RAID system, receive the instruction sent by the communication unit 42, obtain the parameter of the virtual storage disk group through the communication unit 42; and use the distribution algorithm and the parameter of the virtual storage disk group Calculate the new virtual storage hard disk distribution table for the virtual storage disk group.

Optionally, the processor 41 is configured to determine, according to the new virtual storage hard disk distribution table, whether the current virtual storage hard disk is a verified virtual storage hard disk that needs to be deleted; if the current virtual storage hard disk is not a verified virtual storage hard disk that needs to be deleted According to the new virtual storage hard disk distribution table, it is determined whether the current virtual storage hard disk is in the same position as the new virtual storage hard disk distribution table; if the current virtual storage hard disk is the verified virtual storage hard disk to be deleted, the current deletion is deleted. Verification data in the virtual storage hard disk and the current virtual storage hard disk.

Optionally, the processor 41 is further configured to: if the current virtual storage hard disk does not match the location in the virtual storage hard disk distribution table, copy the data in the current virtual storage hard disk to the storage hard disk of the newly added storage node, The data in the current virtual storage hard disk and the current virtual storage hard disk in the original location is deleted; if the current virtual storage hard disk matches the position in the virtual storage hard disk distribution table, the current virtual storage hard disk location is maintained.

The RAID system expansion device provided by the embodiment of the present invention calculates a new virtual storage according to a parameter and a distribution algorithm of a virtual storage disk group in a current RAID system (ie, a RAID system after the storage node is added) when the RAID system is expanded. The disk distribution table is such that the number of data virtual storage disks in each virtual storage disk group in the current RAID system is unchanged. The number of virtual storage disks Y is changed, and then the virtual storage disks in the new virtual storage disk distribution table are followed. The distribution location of the virtual storage hard disk in each virtual storage disk group is migrated or deleted, thereby improving the utilization of the system storage space, and compared with the prior art, in the case of the same space utilization, due to the present invention By keeping the number of X unchanged, only the number of Ys is modified to perform data equalization on the expanded RAID system, thereby greatly reducing the amount of data and the amount of computation. It will be apparent to those skilled in the art that, for convenience and brevity of description, only the division of each functional module described above is exemplified. In practical applications, the above-mentioned function assignment can be completed by different functional modules as needed. The internal structure of the device is divided into different functional modules to perform all or part of the functions described above. For the specific working process of the system, the device and the unit described above, refer to the corresponding process in the foregoing method embodiment, and details are not described herein again.

In the several embodiments provided herein, it should be understood that the disclosed systems, apparatus, and methods may be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules or units is only a logical function division. In actual implementation, there may be another division manner, for example, multiple units or components may be used. Combined or can be integrated into another system, or some features can be ignored, or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be electrical, mechanical or other form.

The units described as separate components may or may not be physically separate. The components displayed as units may or may not be physical units, i.e., may be located in one place, or may be distributed over multiple network units. Some or all of the units may be selected according to actual needs to achieve the objectives of the solution of the embodiment. In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a standalone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application, in essence or the contribution to the prior art, or all or part of the technical solution may be embodied in the form of a software product stored in a storage medium. The instructions include a plurality of instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor to perform all or part of the steps of the methods described in various embodiments of the present application. The foregoing storage medium includes: a USB flash drive, a removable hard disk, a read-only memory (ROM), A medium that can store program codes, such as a random access memory (RAM), a magnetic disk, or an optical disk.

The above embodiments are only used to illustrate the technical solutions of the present application, and are not limited thereto. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that The technical solutions described in the embodiments are modified, or some of the technical features are equivalently replaced; and the modifications or substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

WO 2015/035768 -fcrt -mf _| PCT/CN2014/074775 Claims

1. A redundant array of independent disks RAID system expansion method, characterized in that the RAID system includes N storage nodes, and each storage hard disk in the N storage nodes is divided into virtual storage of the same size Hard disk, M virtual storage hard disks form a virtual storage hard disk group with redundant relationship, the virtual storage hard disk group is composed of X data virtual storage hard disks and Y block verification virtual storage hard disks, M=X+Y, the The data virtual storage hard disk stores data blocks, and the verification virtual storage hard disk stores verification blocks. The method includes: when the RAID system adds storage nodes, according to the parameters and distribution of the virtual storage hard disk group in the current RAID system. The algorithm calculates a new virtual storage hard disk distribution table; compared with the original virtual storage hard disk distribution table, the new virtual storage hard disk distribution table maintains the number X of data virtual storage hard disks in each virtual storage hard disk group in the current RAID system. remains unchanged, verify that the number Y of virtual storage hard disks changes;

Migrate or delete the virtual storage hard disks in the virtual storage hard disk group according to the distribution positions of the virtual storage hard disks in the new virtual storage hard disk distribution table.

2. The method according to claim 1, characterized in that when a storage node is added to the RAID system, a new virtual storage hard disk is calculated according to the parameters and distribution algorithm of the virtual storage hard disk group in the current RAID system. The distribution table includes:

When the RAID system adds the storage node, obtain the parameters of the virtual storage hard disk group;

A new virtual storage hard disk distribution table of the virtual storage hard disk group is calculated through the distribution algorithm and the parameters of the virtual storage hard disk group.

3. The method according to claim 1 or 2, characterized in that the parameters of the virtual storage hard disk group include at least one of the following: a storage node address ID list, a storage node weight list, a parameter of the virtual storage hard disk group, The identification and the number of virtual storage hard disks in the virtual storage hard disk group; wherein the storage node weight list is the storage space usage of the current storage hard disk of each storage node.

4. The method according to claim 1, characterized in that, the virtual storage hard disks in the virtual storage hard disk group are migrated or deleted according to the distribution positions of the virtual storage hard disks in the new virtual storage hard disk distribution table. include: Determine whether the current virtual storage hard disk is a verification virtual storage hard disk that needs to be deleted according to the new virtual storage hard disk distribution table; if the current virtual storage hard disk is not a verification virtual storage hard disk that needs to be deleted, then according to the new The virtual storage hard disk distribution table determines whether the location of the current virtual storage hard disk matches the position in the new virtual storage hard disk distribution table; if the current virtual storage hard disk is a verification virtual storage hard disk that needs to be deleted, delete it. The current virtual storage hard disk and the verification data in the current virtual storage hard disk.

5. The method according to claim 4, wherein the step of determining, based on the new virtual storage hard disk distribution table, whether the current virtual storage hard disk location is consistent with the location of the new virtual storage hard disk distribution table is After the locations match, it also includes: If the current virtual storage hard disk does not match the location in the virtual storage hard disk distribution table, copy the data in the current virtual storage hard disk to the storage hard disk of the newly added storage node. , and delete the current virtual storage hard disk and the data in the current virtual storage hard disk; if the current virtual storage hard disk matches the position in the virtual storage hard disk distribution table, keep the current The location of the virtual storage hard disk remains unchanged.

6. An independent disk redundant array RAID system expansion device, characterized in that the RAID system includes N storage nodes, and each storage hard disk in the N storage nodes is divided into virtual storage of the same size. Hard disk, M virtual storage hard disks form a virtual storage hard disk group with a redundant relationship, the virtual storage hard disk group is composed of X data virtual storage hard disks and Y block verification virtual storage hard disks, M=X+Y, the The data virtual storage hard disk stores the data blocks, the verification virtual storage hard disk stores the verification blocks, and the device includes: a distribution table calculation unit, used to perform the calculation according to the current RAID system when a storage node is added to the RAID system. The parameters and distribution algorithm of the virtual storage hard disk group calculate a new virtual storage hard disk distribution table; The new virtual storage hard disk distribution table is compared with the original virtual storage hard disk distribution table. Each virtual storage hard disk group in the current RAID system The number X of virtual storage hard disks for medium data remains unchanged, and the number Y of verification virtual storage hard disks changes;

Migration unit, used for the new virtual storage calculated according to the distribution table calculation unit The virtual storage hard disks in each virtual storage hard disk group are migrated or deleted based on the distribution position of each virtual storage hard disk in the hard disk distribution table.

7. The device according to claim 6, characterized in that the distribution table calculation unit includes:

A parameter acquisition module, used to obtain the parameters of the virtual storage hard disk group when the storage node is added to the RAID system;

A distribution table calculation module, configured to calculate the new virtual storage hard disk distribution table of the virtual storage hard disk group through the parameters of the virtual storage hard disk group obtained by the distribution algorithm and the parameter acquisition unit.

8. The device according to claim 6 or 7, characterized in that the parameters of the virtual storage hard disk group include at least one of the following: a storage node address ID list, a storage node weight list, a parameter of the virtual storage hard disk group, The identification and the number of virtual storage hard disks in the virtual storage hard disk group; wherein the storage node weight list is the storage space usage of the current storage hard disk of each storage node.

9. The device according to claim 6, wherein the migration unit includes: a first judgment module, configured to judge whether the current virtual storage hard disk needs to be deleted based on the new virtual storage hard disk distribution table. Check the virtual storage hard drive;

The second judgment module is configured to, if the judgment result of the first judgment module is no, judge whether the location of the current virtual storage hard disk is consistent with the distribution of the new virtual storage hard disk according to the new virtual storage hard disk distribution table. The positions in the table match; a migration module, configured to delete the current virtual storage hard disk and the verification data in the current virtual storage hard disk if the judgment result of the first judgment module is yes.

10. The device according to claim 9, characterized in that: the migration module is also configured to copy the data in the current virtual storage hard disk to the newly added in the storage hard disk of the storage node, and delete the current virtual storage hard disk at the original location and the data in the current virtual storage hard disk; used to keep if the judgment result of the second judgment module is yes. The current virtual storage hard disk location remains unchanged.