CN103858092B - A kind of data migration method and device - Google Patents
A kind of data migration method and device Download PDFInfo
- Publication number
- CN103858092B CN103858092B CN201380002566.1A CN201380002566A CN103858092B CN 103858092 B CN103858092 B CN 103858092B CN 201380002566 A CN201380002566 A CN 201380002566A CN 103858092 B CN103858092 B CN 103858092B
- Authority
- CN
- China
- Prior art keywords
- extent
- disk
- lun
- disk group
- moving out
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F12/00—Accessing, addressing or allocating within memory systems or architectures
- G06F12/02—Addressing or allocation; Relocation
- G06F12/08—Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
- G06F12/0802—Addressing of a memory level in which the access to the desired data or data block requires associative addressing means, e.g. caches
- G06F12/0866—Addressing of a memory level in which the access to the desired data or data block requires associative addressing means, e.g. caches for peripheral storage systems, e.g. disk cache
- G06F12/0868—Data transfer between cache memory and other subsystems, e.g. storage devices or host systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F12/00—Accessing, addressing or allocating within memory systems or architectures
- G06F12/02—Addressing or allocation; Relocation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/08—Error detection or correction by redundancy in data representation, e.g. by using checking codes
- G06F11/10—Adding special bits or symbols to the coded information, e.g. parity check, casting out 9's or 11's
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/16—Error detection or correction of the data by redundancy in hardware
- G06F11/20—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements
- G06F11/2053—Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements where persistent mass storage functionality or persistent mass storage control functionality is redundant
- G06F11/2094—Redundant storage or storage space
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2201/00—Indexing scheme relating to error detection, to error correction, and to monitoring
- G06F2201/81—Threshold
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2212/00—Indexing scheme relating to accessing, addressing or allocation within memory systems or architectures
- G06F2212/26—Using a specific storage system architecture
- G06F2212/261—Storage comprising a plurality of storage devices
- G06F2212/262—Storage comprising a plurality of storage devices configured as RAID
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a kind of Data Transference Technology, comprising: the load information obtaining the first disk group; When described load information reach move out threshold value time, at least one digital independent of moving out in Extent of destination logical unit LUN in the first disk group in the buffer memory of described controller; Obtain at least one Extent that moves in second disk group, the quantity of the described Extent of moving into is identical with the quantity of the described Extent that moves out, to move into described in the RAID grade of Extent of moving out described in each and one the RAID grade one_to_one corresponding of Extent, described in the Extent that moves into be idle Extent; To move out described in the mapping relations of the described LUN of moving out and the described Extent of moving out are updated to the mapping relations of LUN and the described Extent that moves into.
Description
Technical field
The present invention relates to memory technology, particularly a kind of Data Transference Technology.
Background technology
In field of storage, in order to improve the read or write speed of data or strengthen the security of data, one is called that the technology of inexpensive disk array (RedundantArraysofIndependentDisks, RAID) is used widely.In this art, according to the requirement of certain RIAD rank, multiple disk is linked together, form a RAID group.
According to small computer system interface (SmallComputerSystemInterface, SCSI), or fiber channel (FibreChannel, FC), internet small computer system interface net iSCSI (internetSmallComputerSystemInterface, etc. SCSI) agreement of SCSI is encapsulated, or other similar agreements, can according to the capacity of RAID group, logically transverse cuts is done to RAID group, this RAID group is divided into multiple logical block (LogicUnit).In field of storage, also call logical block with LUN (logical unit number, LogicUnitNumber) traditionally, unless stated otherwise, the present invention continues to use this custom.
In RAID group, each disk provides the storage space of formed objects for this LUN, and each unit cut down just is called a LUN.Each LUN can supply host access separately, and like this, an original RAID group just can employ as multiple RAID group in logic.
Disk group is made up of at least one RAID group, because in disk group, number of disks is difficult to direct expansion, and disk set up vertical at the beginning of, be difficult to determine the actual demand amount of user to resource, therefore often occur that resource that disk group can provide is difficult to the situation of meeting consumers' demand.
Summary of the invention
The invention provides a kind of Data Transference Technology, the Data Migration across disk group can be realized.
First aspect, the invention provides a kind of data migration method, be applied in controller, described controller respectively with the first disk group, second disk group connects, described first disk group and described second disk group include at least 2 disks, described first disk group DG and described second disk group include disk, each described disk comprises multiple piecemeal Chunk, wherein, at least 2 from same DG and the Chunk of different disk, according to Redundant Array of Inexpensive Disc RAID algorithm groups component block group CKG, each CKG is divided into multiple length Extent, the storage space of each Extent is provided jointly by all Chunk of CKG, to same Extent, the storage size that each Chunk provides is identical, the method comprises: the load information obtaining the first disk group, when described load information reach move out threshold value time, at least one digital independent of moving out in Extent of destination logical unit LUN in the first disk group in the buffer memory of described controller, obtain at least one Extent that moves in second disk group, the quantity of the described Extent of moving into is identical with the quantity of the described Extent that moves out, to move into described in the RAID grade of Extent of moving out described in each and one the RAID grade one_to_one corresponding of Extent, described in the Extent that moves into be idle Extent, to move out described in the mapping relations of the described LUN of moving out and the described Extent of moving out are updated to the mapping relations of LUN and the described Extent that moves into.
In the first possible implementation of first aspect, load information comprises: bandwidth B PS per second higher than first threshold, with idle storage space lower than at least one in Second Threshold.
Second aspect, the invention provides a kind of data migration device, for moving the data of the first disk group to second disk group, described first disk group DG and described second disk group include disk, each described disk comprises multiple piecemeal Chunk, described first disk group and described second disk group include at least 2 disks, wherein, at least 2 from same DG and the Chunk of different disk, according to Redundant Array of Inexpensive Disc RAID algorithm groups component block group CKG, each CKG is divided into multiple length Extent, the storage space of each Extent is provided jointly by all Chunk of CKG, to same Extent, the storage size that each Chunk provides is identical, this device comprises: parameter acquiring unit, for obtaining the load information of the first disk group, data-reading unit, is connected with described parameter acquiring unit 41, when described load information reach move out threshold value time, at least one digital independent of moving out in Extent of destination logical unit LUN in the first disk group in the buffer memory of described controller, data write unit, be connected with described data-reading unit, to move into Extent at least one acquisition in second disk group, the quantity of the described Extent of moving into is identical with the quantity of the described Extent that moves out, to move into described in the RAID grade of Extent of moving out described in each and one the RAID grade one_to_one corresponding of Extent, described in the Extent that moves into be idle Extent, LUN updating block, for the mapping relations of move out described in the mapping relations of the described LUN of moving out and the described Extent of moving out being updated to LUN and the described Extent that moves into.
In the first possible implementation of second aspect, load information comprises: bandwidth B PS per second higher than first threshold, with idle storage space lower than at least one in Second Threshold.
Accompanying drawing explanation
In order to be illustrated more clearly in embodiment of the present invention technical scheme, be briefly described by the accompanying drawing used below required for embodiment, the accompanying drawing in the following describes is only some embodiments of the present invention, can also obtain other accompanying drawing according to these accompanying drawings.
Fig. 1 is embodiment of the present invention data migration method process flow diagram;
Fig. 2 is the invention process csr controller schematic diagram;
Fig. 3 is embodiment of the present invention data migration device structural drawing.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, be clearly and completely described technical scheme of the present invention, obviously, described embodiment is only the present invention's part embodiment, instead of whole embodiments.Based on the every other embodiment that the embodiment in the present invention obtains, all belong to the scope of protection of the invention.
In traditional RAID technique, first according to selected RAID rank, multiple disk combination is formed RAID group together, then using each disk in RAID group as least unit, each disk takes out equal-sized a part of space and jointly forms LUN.LUN is supplied to main frame read-write as unit, because LUN is across all disks, so when data write by main frame, data can be write in whole RAID group; During read data, provide data to main frame by whole RAID group.
Major part RAID rank both provides the function of data reconstruction.Such as in RAID3, there is a disk storage checking data, all the other disk storage business datums.After having a disk to there occurs fault, the data in failed disk can be recovered by checking algorithm by all the other disks.But, because the storage space of single disk is increasing, and the number of disks formed in RAID group is often few, such as form a RAID group by 6 disks, if therefore there is a disk to there occurs fault, can only participate in reconstruct by all the other disks that quantity in RAID group is few, this time may be very very long, such as reconstruct the data block of 1 terabyte (TB), need to expend 10 hours.If have new disk failures during reconstruct, the loss of data that cannot recover may be caused to storage system.
With directly set up using disk as granularity in prior art RAID group unlike.In the present invention program, first each disk in multiple disk is cut into equal-sized piecemeal (CK, Chunk), such as 6 disks form a disk group (DiskGroup, DG), disk each in DG is cut into the Chunk that size is 64MB (megabyte), then RAID group is set up using Chunk as granularity, such as 3 Chunk form a RAID group, the RAID group be made up of these Chunk, we are called piecemeal group (ChunkGroup, CKG).Do not have in same disk and belong to multiple Chunk in same CKG, in other words, a disk is at most a CKG provides a Chunk.Each disk not in DG will provide Chunk for CKG, and namely, in CKG, the quantity of Chunk can be fewer than the quantity of disk in DG.
It should be noted that, technology involved in the present invention is not limited to storage medium itself, therefore except disk, is equally applicable to other media such as solid state hard disc, erasable optical disk, floppy disk, tape, for convenience of description, the embodiment of the present invention is illustrated with disk.The said data of the embodiment of the present invention, such as picture, video, database data, audio frequency, text etc.
The RAID relation that now between disk, not recomposition is traditional, but the disk group (such as 96 disks) that can form larger number of disks, the Chunk on disk can therewith in disk group the Chunk of different disk form CKG.Each CKG can have oneself RAID rank, and like this, the Chunk on a disk can belong to multiple piecemeal groups of multiple RAID rank.
Each Chunk in CKG is divided into the storage space of fritter, a block space of formed objects is provided by each Chunk, a length (Extent) can be formed, the size such as 256KB (kilobyte) of an Extent.Use this way, a CKG can be divided into multiple Extent, form ExtentPool (length pond).Extent is the elementary cell forming LUN of the present invention, chooses one or more Extent and can form LUN for host access from ExtentPool.When LUN is made up of multiple Extent, form the Extent of this LUN from same CKG, also can from different CKG, the RAID rank forming the Extent of same LUN can be the same or different, the Raid rank of Extent with its from the RAID rank of CKG identical.The storage size belonging to the different Extent of same LUN is identical, can facilitate during Data Migration and realize mutual exclusion.
In addition, due in traditional RAID technique, are LUNID+ LBA (Logical Block Addressing) (LBA) addresses to the addressing mode of data in LUN.If every 6 Chunk form a RAID group, the storage space of so each Extent is also from 6 disks.But because LUN is made up of different Extent, and the disk that each Extent crosses over can be different, so in this case, the disk that the number of disks that LUN crosses over can be crossed over than traditional RAID group is more.Therefore, when there being disk failures, the number of disks participating in reconstruct is also more, and reconstitution time shortens greatly.The foundation of embodiment of the present invention Chunk, CKG, Extent and LUN can have been come by controller.Jointly be made up of controller and storage medium in storage system, storage medium provides storage space, and controller manages storage space, and the reading be responsible for data in storage space and write.The disk that Extent crosses over, refers to the disk that the storage space of Extent is originated.
The embodiment of the present invention can realize the Data Migration across DG.Multiple DG can be had in a storage system, each DG is made up of the disk of multiple identical type, the quantity required according to RAID algorithm in DG selects CKG, CKG is made up of the block on hard disk different in DG, the space of LUN is made up of the space of several CKG, so the data of LUN can be distributed in the multiple hard disks in DG.
In the embodiment of the present invention, same disk group belongs to same failure domain, and different disk groups belongs to different failure domains.If that is there is the disk of any amount to send fault in a disk group, when recovering data, haveing nothing to do with the disk of other disk groups, relying on other disks of disk group inside to recover.
By the division of CKG, there is direct or indirect relation in disk and other disks of same disk group.Such as one total disk A, B, C, D and E amounts to 5 disks and forms a DG, and the Chunk so forming a CKG can from wherein any more than 2 or 2 disks.The Chunk of such as CKG1 can from these 3 disks of A, B and C, the Chunk of CKG2 from the Chunk of these 4 disks of A, B, C and D, CKG3 from these three disks of C, D, E ... because disk A and disk C provides Chunk to form CKG1 jointly, therefore by CKG1 and CKG2, disk A, C establish direct contact; By CKG3, disk C and E establishes and contacts directly; By contacting directly and contacting directly between disk C, E between disk A, C, form the indirect association of disk A, E.
As Fig. 1, concrete introduction is done to the Data Migration across DG below.
Step 11, controller obtains bandwidth (BandPerSecond, BPS) per second and the remanence disk space of a DG.Have the LUN that can move away in one DG, be called the LUN that moves out, the LUN that moves out maps and is made up of multiple Extent, and there are mapping relations with its Extent of composition.Wherein, controller manages a DG, the 2nd DG.The data of moving out in any Extent of any LUN can reduce the load of a DG.One DG and the 2nd DG can be made up of the storage medium of same size, such as identical read or write speed storage medium.
When each execution input operation or output function, controller can be recorded, and adds up the sum of the input/output operations in the time period, averages and just can obtain BPS per second.And total storage space of controller inquiry DG, and the disk space that each write operation takies, DG remaining space can be obtained.
Step 12, when BPS is higher than first threshold, and remanence disk space lower than Second Threshold time, read the buffer memory of data to controller of at least one Extent in LUN of moving out.
The introduction of this step be the trigger condition of Data Migration.BPS is the read/write operation on average bandwidth taken per second.Such as according to the setting of threshold value, as BPS >=100MB/S, and residual memory space≤10%, just trigger migration, in the LUN that then moves out, the data of at least one Extent are to the buffer memory of controller.The quantity of the Extent of sense data can be set by the user, and also can arrange threshold value respectively, then calculates according to algorithm and obtains.Such as can arrange when meet BPS≤50MB/S and residual memory space >=50% just no longer moves.So can stop the threshold value of migration that the quantity calculating or roughly estimate the Extent needing migration is set according to these two.
Step 13, obtains the Extent information of the 2nd DG, will read the data of controller cache, in the Extent of write the 2nd DG.The Extent of sense data from a DG, and the Extent writing data in the 2nd DG has identical RAID rank and memory capacity.The Extent information and the step 12 that obtain the 2nd DG also can any one first perform, or executed in parallel.The Extent be written into does not store valid data, can be called idle Extent.The RAID rank of Extent is exactly the RAID rank of the CKG belonging to Extent.
If the DG of controller management is except a DG, multiple in addition, one can be selected arbitrarily as the 2nd DG from this multiple DG, also can choose according to certain condition, such as arrange threshold value of moving into, the judgement parameter of threshold value of moving into is identical with the judgement parameter of threshold value of moving out.When threshold value of moving out considers that whether BPS is higher than first threshold, then threshold value of moving into considers that whether BPS is lower than the 3rd threshold value.When threshold value of moving out considers that whether BPS is higher than first threshold, and whether free space is less than Second Threshold; Threshold value of then moving into considers that whether BPS is lower than the 3rd threshold value, and whether free space is greater than the 4th threshold value.Such as, can set DG that BPS and residual memory space all meet certain condition such as, as the 2nd DG, BPS≤20MB/S and residual memory space >=80% just can become by the DG of data of moving into.Selection any one DG except a DG is as the 2nd DG, and statistically also can reach the present invention is beneficial effect.
Step 14, upgrading the Extent that LUN maps that moves out, specifically, is the mapping the Extent with sense data in a DG, change into the 2nd DG in write the mapping of the Extent of data.The storage space of the Extent that moves out of sense data in a DG can be discharged, become idle Extent, for other LUN.
In the embodiment of the present invention, the first disk group and second disk group include disk, and each disk comprises multiple piecemeal Chunk.CKG is made up of at least 2 Chunk and forms according to RAID algorithm, forms the different disk of Chunk from same DG of same CKG.Each CKG is divided into multiple length Extent, and the storage space of each Extent is provided jointly by all Chunk of CKG, and to same Extent, the storage size that each Chunk provides is identical.
LUN is made up of Extent, by mapping relations, and the Extent that record LUN comprises.By upgrading mapping relations, making a part of Extent in LUN from the 2nd DG, completing the migration to data in LUN.
In the present embodiment, relate to BPS, these two parameters of remanence disk space, these two parameters can be referred to as load information, and meet threshold value with these two parameters simultaneously, namely as BPS >=100MB/S, and residual memory space≤10%, as trigger data migration operation.In other embodiments, also only can obtain a wherein parameter, move with regard to trigger data when this parameter meets threshold value.Or consider two parameters simultaneously, two parameters weight being separately set, obtaining a total numerical value after two parameter weightings, then comparing by total numerical value and total threshold value, to judge whether to punish data migration operation.
The embodiment of the present invention adopts the mapping relations of LUN-Extent to carry out addressing to the content of LUN, data Replica to be migrated to object DG, then changes the Data Migration that these mapping relations just can realize across DG.And in traditional RAID, carry out addressing with LUNID+LAB address, the Data Migration across DG cannot be completed.
As Fig. 2, describe from hardware, above-mentioned Data Migration step can be performed by the processor 21 of controller 2, and have program stored therein in the internal memory 22 of controller 2 instruction, and processor 21 execution of program instructions realizes this method.Controller 2 is connected with disk group 31 disk group 32, can be specifically that internal memory 22 is connected with disk group 31 and disk group 32.
As Fig. 3, the present invention also provides a kind of data migration device 4, can apply above-mentioned moving method.Data migration device 4 comprises parameter acquiring unit 41, data-reading unit 42, data write unit 43 and LUN updating block 44.Data migration device 4 is in the controller integrated, can move between disk group to data.
Parameter acquiring unit 41, for obtaining read-write operation per second time (Input/OutputOperationsPerSecond, BPS) and the remanence disk space of a DG.
Have the LUN that can move away in one DG, be called the LUN that moves out, also can be called the LUN that moves out, the LUN that moves out is mapped with multiple Extent.Wherein, controller manages a DG, the 2nd DG.The data of moving out in any Extent of any LUN can reduce the load of a DG.One DG and the 2nd DG can be made up of the storage medium of same size, such as identical read or write speed storage medium.
When each execution input operation or output function, can be recorded in controller buffer memory all, the sum of the input/output operations in the time period is added up, average and just can obtain BPS per second.And parameter acquiring unit 41 inquires about total storage space of DG, and the disk space that each write operation takies, DG remaining space can be obtained.
Data-reading unit 42, is connected with parameter acquiring unit 41, for when BPS is higher than first threshold, and remanence disk space lower than first threshold time, read the buffer memory of data to controller of at least one Extent in LUN of moving out.
The trigger condition of threshold value as Data Migration can be set.Such as according to the setting of threshold value, as BPS >=100MB/S, and residual memory space≤10%, just trigger migration, in the LUN that then moves out, the data of at least one Extent are to the buffer memory of controller.The quantity of the Extent of sense data can be set by the user, and also can arrange threshold value respectively, then calculates according to algorithm and obtains.Such as can arrange when meet BPS≤50MB/S and residual memory space >=50% just no longer moves.So can stop the threshold value of migration that the quantity calculating or roughly estimate the Extent needing migration is set according to these two.
Data write unit 43, is connected with data-reading unit 42, for obtaining the Extent information of the 2nd DG, will read the data of controller cache, in the Extent of write the 2nd DG.
The Extent of sense data from a DG, and the Extent writing data in the 2nd DG has identical RAID rank and memory capacity.The Extent be written into does not store valid data, can be called idle Extent.The RAID rank of Extent is exactly the RAID rank of the CKG belonging to Extent.
If DG has more than three or three, data write unit 43 arbitrarily can select one as the 2nd DG, also can choose according to certain condition, such as, arrange threshold value of moving into, the project of threshold value of moving into and threshold value coupling of moving out.When threshold value of moving out considers that whether BPS is higher than first threshold, then threshold value of moving into considers that whether BPS is lower than the 3rd threshold value.When threshold value of moving out considers that whether BPS is higher than first threshold, and whether free space is less than Second Threshold; Threshold value of then moving into considers that whether BPS is lower than the 3rd threshold value, and whether free space is greater than the 4th threshold value.
LUN updating block 44, is connected with data write unit 43, for upgrading the Extent that the LUN that moves out maps, specifically, is the mapping the Extent with sense data in a DG, change into the 2nd DG in write the mapping of the Extent of data.LUN updating block 44 can also further, the storage space of the Extent that moves out of sense data in release the one DG, become idle Extent.
LUN is made up of Extent, by mapping relations, and the Extent that record LUN comprises.By upgrading mapping relations, making a part of Extent in LUN from the 2nd DG, completing the migration to data in LUN.
In the embodiment of the present invention, the first disk group DG and second disk group include disk, and each disk comprises multiple piecemeal Chunk.CKG is made up of at least 2 Chunk and forms according to RAID algorithm, forms the different disk of Chunk from same DG of same CKG.Each CKG is divided into multiple length Extent, and the storage space of each Extent is provided jointly by all Chunk of CKG, and to same Extent, the storage size that each Chunk provides is identical
In the present embodiment, relate to BPS, these two parameters of remanence disk space, these two parameters can be referred to as load information, and meet threshold value with these two parameters simultaneously, namely as BPS >=100MB/S, and residual memory space≤10%, as trigger data migration operation.In other embodiments, also only can obtain a wherein parameter, move with regard to trigger data when this parameter meets threshold value.Or consider two parameters simultaneously, weight is set to two, obtains a total numerical value after weighting, then compare by total numerical value and total threshold value, to judge whether to punish data migration operation.
The embodiment of the present invention adopts the mapping relations of LUN-Extent to carry out addressing to the content of LUN, data Replica to be migrated to object DG, then changes the Data Migration that these mapping relations just can realize across DG.And in traditional RAID, carry out addressing with LUNID+LAB address, the Data Migration across DG cannot be completed.
In above-described embodiment, except BPS, read-write number of times (Input/OutputOperationsPerSecond per second can also be used, IOPS) as threshold value, BPS=IOPS × I/O size, I/O size is the data volume that single read/write requests is asked, such as such as 512B, 4KB, 1MB etc.
One of ordinary skill in the art will appreciate that the possible implementation of various aspects of the present invention or various aspects can be embodied as system, method or computer program.Therefore, the possible implementation of each aspect of the present invention or various aspects can adopt complete hardware embodiment, completely software implementation (comprising firmware, resident software etc.), or the form of the embodiment of integration software and hardware aspect, is all referred to as " circuit ", " module " or " system " here.In addition, the possible implementation of each aspect of the present invention or various aspects can adopt the form of computer program, and computer program refers to the computer readable program code be stored in computer-readable medium.
Computer-readable medium can be computer-readable signal media or computer-readable recording medium.Computer-readable recording medium is including but not limited to electronics, magnetic, optics, electromagnetism, infrared or semiconductor system, equipment or device, or aforesaid appropriately combined arbitrarily, as random access memory (RAM), ROM (read-only memory) (ROM), Erasable Programmable Read Only Memory EPROM (EPROM or flash memory), optical fiber, portable ROM (read-only memory) (CD-ROM).
Processor in computing machine reads the computer readable program code be stored in computer-readable medium, makes processor can perform the function action specified in the combination of each step or each step in flow charts; Generate the device implementing the function action specified in the combination of each block of block diagram or each piece.
Claims (12)
1. a data migration method, be applied in controller, described controller respectively with the first disk group, second disk group connects, it is characterized in that: described first disk group and described second disk group include at least 2 disks, each described disk comprises multiple piecemeal Chunk, wherein, at least 2 from same disk group and the Chunk of different disk, according to Redundant Array of Inexpensive Disc RAID algorithm groups component block group CKG, each CKG is divided into multiple length Extent, the storage space of each Extent is provided jointly by all Chunk of CKG, to same Extent, the storage size that each Chunk provides is identical, the method comprises:
Obtain the load information of the first disk group;
When described load information reach move out threshold value time, at least one digital independent of moving out in Extent of destination logical unit LUN in the first disk group in the buffer memory of described controller;
Obtain at least one Extent that moves in second disk group, the quantity of the described Extent of moving into is identical with the quantity of the described Extent that moves out, to move into described in the RAID grade of Extent of moving out described in each and one the RAID grade one_to_one corresponding of Extent, described in the Extent that moves into be idle Extent;
To move out described in the mapping relations of the described LUN of moving out and the described Extent of moving out are updated to the mapping relations of LUN and the described Extent that moves into.
2. method according to claim 1, is characterized in that, described load information reaches threshold value of moving out, and comprising:
Bandwidth B PS per second is higher than first threshold; With
Idle storage space is lower than Second Threshold.
3. method according to claim 1, is characterized in that, described load information reaches threshold value of moving out, and comprising:
Bandwidth B PS per second is higher than first threshold.
4. method according to claim 1, is characterized in that, described load information reaches threshold value of moving out, and comprising:
Idle storage space is lower than Second Threshold.
5. the method according to claim 1,2,3 or 4, is characterized in that, after the mapping relations of move out described in the mapping relations of the described LUN of moving out and the described Extent of moving out being updated to LUN and the described Extent that moves into, comprises further:
To move out described in release the storage space of Extent.
6. the method according to claim 1,2,3 or 4, is characterized in that:
The load information of described second disk group meets threshold value of moving into.
7. the method according to claim 1,2,3 or 4, is characterized in that:
The storage size belonging to the different Extent of same LUN is identical.
8. a data migration device, for moving the data of the first disk group to second disk group, it is characterized in that: described first disk group and described second disk group include disk, each described disk comprises multiple piecemeal Chunk, described first disk group and described second disk group include at least 2 disks, wherein, at least 2 from same disk group and the Chunk of different disk, according to Redundant Array of Inexpensive Disc RAID algorithm groups component block group CKG, each CKG is divided into multiple length Extent, the storage space of each Extent is provided jointly by all Chunk of CKG, to same Extent, the storage size that each Chunk provides is identical, this device comprises:
Parameter acquiring unit, for obtaining the load information of the first disk group;
Data-reading unit, is connected with described parameter acquiring unit 41, when described load information reach move out threshold value time, at least one digital independent of moving out in Extent of destination logical unit LUN in the first disk group in the buffer memory of described controller;
Data write unit, be connected with described data-reading unit, to move into Extent at least one acquisition in second disk group, the quantity of the described Extent of moving into is identical with the quantity of the described Extent that moves out, to move into described in the RAID grade of Extent of moving out described in each and one the RAID grade one_to_one corresponding of Extent, described in the Extent that moves into be idle Extent;
LUN updating block, for the mapping relations of move out described in the mapping relations of the described LUN of moving out and the described Extent of moving out being updated to LUN and the described Extent that moves into.
9. device according to claim 8, is characterized in that, described load information reaches threshold value of moving out, and comprising:
Bandwidth B PS per second is higher than first threshold; With
Idle storage space is lower than Second Threshold.
10. device according to claim 8, is characterized in that, described load information reaches threshold value of moving out, and comprising:
Bandwidth B PS per second is higher than first threshold.
11. devices according to claim 8, is characterized in that, described load information reaches threshold value of moving out, and comprising:
Idle storage space is lower than Second Threshold.
12. according to Claim 8,9, device described in 10 or 11, it is characterized in that:
The storage size belonging to the different Extent of same LUN is identical.
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/CN2013/089982 WO2015089798A1 (en) | 2013-12-19 | 2013-12-19 | Data migration method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103858092A CN103858092A (en) | 2014-06-11 |
| CN103858092B true CN103858092B (en) | 2016-03-09 |
Family
ID=50864332
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201380002566.1A Active CN103858092B (en) | 2013-12-19 | 2013-12-19 | A kind of data migration method and device |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN103858092B (en) |
| WO (1) | WO2015089798A1 (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109558068A (en) * | 2017-09-27 | 2019-04-02 | 北京忆恒创源科技有限公司 | Data migration method and migratory system |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105988727B (en) * | 2015-02-25 | 2021-04-16 | 中兴通讯股份有限公司 | Storage method and storage device based on RAID |
| US10002072B2 (en) * | 2015-05-18 | 2018-06-19 | Mediatek Inc. | Method and apparatus for controlling data migration in multi-channel memory device |
| CN106408684B (en) * | 2016-09-30 | 2019-10-18 | 北京新能源汽车股份有限公司 | A kind of method and device of fault message storage |
| WO2018094708A1 (en) | 2016-11-26 | 2018-05-31 | 华为技术有限公司 | Method for migrating data, host and solid static disk |
| CN106598500A (en) * | 2016-12-16 | 2017-04-26 | 郑州云海信息技术有限公司 | Hierarchical storage dynamic migration method |
| CN106775481B (en) * | 2016-12-23 | 2019-11-05 | 华为技术有限公司 | Method for reading data and equipment |
| WO2020215223A1 (en) * | 2019-04-23 | 2020-10-29 | 华为技术有限公司 | Distributed storage system and garbage collection method used in distributed storage system |
| CN111007988B (en) * | 2019-11-29 | 2022-10-18 | 苏州浪潮智能科技有限公司 | RAID internal wear balancing method, system, terminal and storage medium |
| CN114442942B (en) * | 2022-01-05 | 2023-05-09 | 杭州宏杉科技股份有限公司 | Data migration method, system, equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101084481A (en) * | 2004-12-20 | 2007-12-05 | 伊姆西公司 | A method to perform parallel data migration in a clustered storage environment |
| CN101587425A (en) * | 2009-06-16 | 2009-11-25 | 杭州华三通信技术有限公司 | A kind of method and device that increases magnetic disc redundant array redundancy |
| CN102184080A (en) * | 2011-04-29 | 2011-09-14 | 华中科技大学 | Curing band based data reading and writing method used after capacity expansion of RAID (Redundant Array of Independent Disk) system |
| US8452928B1 (en) * | 2006-06-29 | 2013-05-28 | Emc Corporation | Virtual array non-disruptive migration of extended storage functionality |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101546249A (en) * | 2008-03-26 | 2009-09-30 | 中兴通讯股份有限公司 | On-line capacity expansion method for disk arrays |
| US8639907B2 (en) * | 2010-07-22 | 2014-01-28 | Netgear, Inc. | Method and apparatus for dynamically adjusting memory capacity in accordance with data storage |
| CN102103629B (en) * | 2010-12-14 | 2012-09-26 | 西北工业大学 | Online data migration method |
-
2013
- 2013-12-19 CN CN201380002566.1A patent/CN103858092B/en active Active
- 2013-12-19 WO PCT/CN2013/089982 patent/WO2015089798A1/en active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101084481A (en) * | 2004-12-20 | 2007-12-05 | 伊姆西公司 | A method to perform parallel data migration in a clustered storage environment |
| US8452928B1 (en) * | 2006-06-29 | 2013-05-28 | Emc Corporation | Virtual array non-disruptive migration of extended storage functionality |
| CN101587425A (en) * | 2009-06-16 | 2009-11-25 | 杭州华三通信技术有限公司 | A kind of method and device that increases magnetic disc redundant array redundancy |
| CN102184080A (en) * | 2011-04-29 | 2011-09-14 | 华中科技大学 | Curing band based data reading and writing method used after capacity expansion of RAID (Redundant Array of Independent Disk) system |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109558068A (en) * | 2017-09-27 | 2019-04-02 | 北京忆恒创源科技有限公司 | Data migration method and migratory system |
| CN109558068B (en) * | 2017-09-27 | 2024-04-05 | 北京忆恒创源科技股份有限公司 | Data migration method and migration system |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2015089798A1 (en) | 2015-06-25 |
| CN103858092A (en) | 2014-06-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103858092B (en) | A kind of data migration method and device | |
| US11243706B2 (en) | Fragment management method and fragment management apparatus | |
| US10146435B2 (en) | Storage system and data write method | |
| US10521131B2 (en) | Storage apparatus and storage control apparatus | |
| US11157365B2 (en) | Method for processing stripe in storage device and storage device | |
| CN102937882B (en) | To effective access with the memory device using bitmap | |
| US20160179403A1 (en) | Storage controller, storage device, storage system, and semiconductor storage device | |
| TWI569139B (en) | Valid data merging method, memory controller and memory storage apparatus | |
| US20130290613A1 (en) | Storage system and storage apparatus | |
| WO2015015550A1 (en) | Computer system and control method | |
| KR101729636B1 (en) | File processing method and storage device | |
| CN106033320B (en) | Processing method and device for high-performance Redundant Array of Inexpensive Disks (RAID) mechanism | |
| KR20190143502A (en) | Method for storing data by storage device and storage device | |
| US10019315B2 (en) | Control device for a storage apparatus, system, and method of controlling a storage apparatus | |
| CN103019882B (en) | The RAID4 system of solid state hard disc | |
| TW201643690A (en) | Data storage system and specific instruction executing method thereof | |
| CN104679442A (en) | Method and device for improving performance of disk array | |
| US9158478B2 (en) | Storage system and storage control method | |
| KR20210076834A (en) | Enhanced ssd reliability | |
| US11379326B2 (en) | Data access method, apparatus and computer program product | |
| CN103631682A (en) | Data backup implement method and device | |
| CN103645995B (en) | Write the method and device of data | |
| CN106844229A (en) | The method for organizing of solid state hard disc firmware mapping table, system and device | |
| KR102032878B1 (en) | Method for correcting error of flash storage controller | |
| CN104461394A (en) | RAID (redundant array of independent disk) and method for reading data from same |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |