CN104205097B - A kind of De-weight method device and system - Google Patents
A kind of De-weight method device and system Download PDFInfo
- Publication number
- CN104205097B CN104205097B CN201380002564.2A CN201380002564A CN104205097B CN 104205097 B CN104205097 B CN 104205097B CN 201380002564 A CN201380002564 A CN 201380002564A CN 104205097 B CN104205097 B CN 104205097B
- Authority
- CN
- China
- Prior art keywords
- slitting
- data block
- data
- group
- block
- 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
Abstract
The invention provides a kind of data processing technique, for data deduplication, fingerprint detection is carried out to multiple data blocks, when the fingerprint of multiple data blocks is identical, the stability of slitting group where further detecting data block, retain the high data block of stability, or when slitting group reliability where data block is not high, data recovery is carried out to slitting group where data block.The reliability of slitting group where the present invention can improve data block, lifts data safety.
Description
Technical field
The present invention relates to field of storage, more particularly to a kind of duplicate removal technology.
Background technology
In field of storage, in order to save memory space, duplicate removal (De-duplicate) is a kind of is frequently used
Technology, its way is remaining and this number if necessary to when being stored to many parts of identical data, only store portion therein
No longer stored according to the data repeated.That is the data repeated are deleted, therefore this technology is also referred to as data de-duplication.
In the selection of granularity, the base unit of duplicate removal can be used as using data block file declustering into data block.With
Data block as duplicate removal base unit when, fingerprint, the content strong correlation of fingerprint and data block can be assigned for each data block.
When the fingerprint of two data blocks is identical, we can draw the content identical conclusion of the two data blocks, be gone by performing
Operate again, only store one of data block into storage system, another data block is not stored.
However, duplicate removal technology also brings the problem of Information Security is reduced, if this only a data is because deposit
Storage system failure and he damages, the security for being likely to result in data is substantially reduced or data are permanently lost.
The content of the invention
The present invention can improve Information Security.
In a first aspect, the present invention provides a kind of data processing method, in controller, this method to include:It is multiple when having
When data block has identical fingerprint, the address of data block according to the fingerprint queries is searched according to the data block address
Slitting group shared by the data block;Verify the independent Redundant Array of Inexpensive Disc RAID of the slitting group of the data block
Slitting state, the data block in the consistent slitting group of slitting is preserved according to check results;It is described to preserve slitting according to check results
Data block in consistent slitting group, including following at least one:If there is the consistent slitting group of slitting, and exist
The consistent slitting group of non-slitting, then retain the data block in the consistent slitting group of slitting, delete remaining described slitting
Data block in group;If there is no the consistent slitting group of slitting, and there is the slitting group degraded, then calculated by RAID
It is the number that is stored with the consistent slitting group of slitting, the consistent slitting group of the slitting that method, which is repaired the slitting group reparation of degradation,
According to block, the data block in remaining described slitting group is deleted.
Second aspect, the present invention provides a kind of data block processing unit, and the device includes:Fingerprint comparison module, for pair
The fingerprint of data block is compared in storage device;Address search module, when thering are multiple data blocks to have phase in the storage device
With fingerprint when, according to the address of the multiple data block of the fingerprint queries, institute is searched according to the multiple data block address
State the slitting group shared by multiple data blocks;Consistency desired result module, the slitting group for verifying the multiple data block
Independent Redundant Array of Inexpensive Disc RAID slitting states, preserve the data in the consistent slitting group of slitting according to check results
Block;The data block preserved according to check results in the consistent slitting group of slitting, including following at least one:If there is point
The consistent slitting group of bar, and there is the consistent slitting group of non-slitting, then retain the consistent slitting group of slitting
In data block, delete the data block in remaining described slitting group;If there is no the consistent slitting group of slitting, and deposit
Then it is the consistent slitting group of slitting, described point by the slitting group reparation of degradation by RAID algorithm reparation in the slitting group of degradation
Be stored with the data block in the consistent slitting group of bar, deletes the data block in remaining described slitting group.
The third aspect, the present invention provides a kind of data block processing method, in controller, this method to include:Inquiry is deposited
The fingerprint base of data block in equipment is stored up, when existing in storage device and the data block to be stored possesses the storages of identical fingerprints
During data block, the independent Redundant Array of Inexpensive Disc RAID slitting states of the detection place of the data storage block slitting group;Root
Data block storage, including following one or more of which strategies are carried out according to detection:Tactful A, if the slitting shape of the slitting group
State is that slitting is consistent, then does not store the data block to be stored;Tactful B, if the slitting state of the slitting group is slitting drop
Level, then:Store the data block to be stored, delete described in data storage block;Or, storage is according to RAID algorithm by the drop
Level slitting group is repaired to the block of data storage;Tactful C, if the slitting state of the slitting group is that slitting differs
Cause, then:Store the data block to be stored, delete described in data storage block;Or, if number does not occur for date classification unit
According to mistake, then the block of data storage is repaired according to RAID algorithm.
Fourth aspect, the present invention provides a kind of data block processing unit, and the device includes:Enquiry module 61, for inquiring about
The fingerprint base of data block in storage device;Consistency desired result module 62, for when in storage device exist and the number to be stored
When possessing the block of data storage of identical fingerprints according to block, the independent inexpensive disk of the detection place of the data storage block slitting group
Redundant array RAID slitting states;Data block storage, including following one or more of which strategies are carried out according to detection:Tactful A,
If the slitting state of the slitting group is that slitting is consistent, the data block to be stored is not stored;Tactful B, if described point
The slitting state of bar group is that slitting degrades, then:Store the data block to be stored, delete described in data storage block;Or, deposit
Storage is repaired according to RAID algorithm by the degradation slitting group to the block of data storage;Tactful C, if the slitting group
Slitting state be that slitting is inconsistent, then:Store the data block to be stored, delete described in data storage block;Or, if
Error in data does not occur for date classification unit, then the block of data storage is repaired according to RAID algorithm.
The present invention program, can in data block duplicate removal, while carry out Detection of Stability to slitting group where data block, when
When stability is not enough by way of storing or repairing again, slitting group stability where data block is improved, so as to improve
Information Security.
Brief description of the drawings
Technical scheme, will make to the accompanying drawing used required for embodiment below in order to illustrate the embodiments of the present invention more clearly
Simply introduce, drawings in the following description are only some embodiments of the present invention, it can also be obtained according to these accompanying drawings
His accompanying drawing.
Fig. 1 is application scenarios topological diagram example of the present invention;
Fig. 2 is a kind of data block processing method flow chart of the embodiment of the present invention;
Fig. 3 is a kind of data block processing method flow chart of the embodiment of the present invention;
Fig. 4 is a kind of data block processing unit embodiment schematic diagram;
Fig. 5 is a kind of data block processing unit embodiment schematic diagram.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, technical scheme is clearly and completely described,
Obviously, described embodiment is only a part of embodiment of the invention, rather than whole embodiments.Based in the present invention
The every other embodiment that embodiment is obtained, belongs to the scope of protection of the invention.
Storage system is generally collectively constituted by controller and storage device.Equivalent to one computer of controller, including processing
Device and internal memory, can be managed to storage device, and provide interface to main frame and storage device.Storage device is provided physically
Memory space, storage device for example can be by solid state hard disc (Solid State Disc, SSD), and serial connecting small is calculated
Machine system interface SCSI (Serial Attached SCSI, SAS) disk is constituted.When writing data, the write data requests of main frame are first sent out
Toward controller, it is then the memory space that this write request is distributed in storage device by controller, and will be carried in write request
Data to be written are sent to storage device and stored.When reading data, data read in controller from storage device, are then sent out by controller
To main frame.Controller and storage device can be physically separate equipment, can also storage device is integrated in the controller.
When storage device is integrated in the controller, the data interaction between controller and storage device turns into the data inside controller
Controller, can also now be called storage server by interaction.
In another topological structure, controller provides management and does not provide the transmission of data, storage device and main frame it
Between data exchange can also be applicable numerical procedure provided in an embodiment of the present invention without controller.
Duplicate removal technology can be divided into (On-line) online and offline (Off-line) two ways.Online mode is to storage
The space availability ratio of equipment is higher;Offline mode writing speed is faster.
Online mode, controller receives a write request, and write request carries new data block.It is written into and deposits in new data block
Store up before equipment, controller checks in storage device whether there is identical data block, if it does not exist, then storage new data
Block is to storage device;If it is present no longer storing this new data block, and set up the LUN and that for possessing this new data block
Index relative between the individual data block existed.This index relative for example can be pointer, need to read this when follow-up
During individual new data block, that data block existed can be read by pointer.
Offline mode, after controller receives a write request, the identical number no matter whether storage device has been stored with
According to block, first data block storage into storage device.Then deduplication operation is periodically performed, or during the equipment to be stored free time
Perform deduplication operation.During deduplication operation, the data block repeated only retains portion, and the storage that release repeated data block takes is empty
Between.That data block that the LUN for pointing to these data blocks is all pointed to retain.
Because repeated data only retains portion, the security of this only a data is just particularly important.Storage system
System can be come by independent Redundant Array of Inexpensive Disc (Redundant Arrays of Independent Disks, RAID)
Improve its Information Security.But the guarantee that RAID can be provided is also limited, when some RAID slitting reliability reduction,
The Information Security wherein stored can be caused to reduce.
In RAID technique, slitting (Stripe) is made up of several stripe units (Stripe Unit, SU), and composition is same
The SU of one slitting may belong to different physical storages, and slitting is also known as band.Belong to the SU of same slitting, Ke Yiyong
There is the memory space of formed objects.In order to manage conveniently, belong to the SU of same slitting, there can also be identical offset,
That is they are located at the same position of different memory.For example for RAID5 or RAID6, SU can be divided into data SU and
SU is verified, data SU is used for storage service data, and verification SU is used for the verification data of storage service data, and verification SU can also claim
For redundancy SU.Integer slitting can constitute logic unit (Logic Unit, LU), and LU can be deposited as the logic towards main frame
Unit is stored up, in convention, logic unit is also referred to as LUN (Logic Unit Number, LUN), and the present invention continues to use this
Plant convention.
As shown in figure 1, being application scenarios topological diagram example of the present invention.Controller 1 and the connection composition storage of storage device 2 system
System.Controller 1 is made up of processing 11 and caching 12, and be stored with computer instruction in caching 12, and the operation computer of processor 11 refers to
Order, corresponding operation is performed to storage device can complete the present invention.Storage device 2 is made up of multiple memories 21, Mei Gecun
Reservoir 21 provides 1 stripe unit SU and constitutes a slitting 211.During data storage, main frame issues data by controller 1,
Controller is stored the data in the slitting of storage device.One data block takes integer slitting, and controller 1 can be with data
Block is that granularity carries out duplicate removal to data.
Data in SU, which make a mistake or lost, is referred to as SU failures.In a slitting, when event occurs for one of SU
During barrier, by RAID algorithm, the data in failure SU can be recovered with the data stored in the SU not broken down, this
The reparation of the individual process, referred to as slitting for recovering data.Some RAID algorithms can carry out data recovery to single SU failures, have
RAID algorithm can carry out data recovery to greater number of SU failures.The number for the failure SU that can be recovered by RAID algorithm
Amount, being referred to as this slitting allows the SU quantity of failure, and such as RAID5 allows a SU failure, and RAID6 allows 2 SU failures.
The state of slitting includes:Slitting is consistent;Slitting degrades;Slitting is inconsistent;Slitting is failed.In these four states
The reliability of slitting is reduced successively.
Slitting is unanimously normal condition, and the data in all SU in slitting are all normal, that is to say, that each SU data
It can read out, and the data that obtained verification data is calculated by data SU data and is verified in SU are identical.Slitting
It is inconsistent, refer to that the data in each SU of slitting can be read out, but the data in data SU calculate what is obtained
Verification data is different with the data stored in verification SU.The inconsistent Producing reason of slitting is probably data SU corrupt datas,
Be probably verify SU corrupt data, or existing data SU corrupt datas have again verify SU corrupt datas.Due in verification SU
Storage is redundant data, therefore ought only verify SU corrupt data, when data SU data do not malfunction, and is not considered as that data are sent out
Loss is sent, verification SU data can be recalculated by data SU data.
Slitting degrades, and refers to faulty SU in slitting, but can recover the data in failure SU by remaining SU in slitting.
When slitting allows the SU more than one broken down, slitting, which degrades, can be further subdivided into multiple grades, and the SU of failure is got over
Many, reliability is lower.For example with RAID6 slitting, there are 2 SU failures of 1 SU failure or appearance and be referred to as slitting drop
Level, but when there are 2 slitting failures, the security of data is lower during 1 slitting failure than occurring in slitting.Slitting is failed, and is
Referring to has SU to occur in that failure, and the data that can not be out of order by remaining SU recoveries of slitting in SU in slitting, that is to say, that
There is partial data to occur in that permanent loss in slitting.SU breaks down, and refers to that logic or physics occurs in SU memory space
On mistake, cause data in SU can not read or can not completely read.
Data block (Block) is the elementary cell of duplicate removal.One data block can be stored in one or more slitting, deposited
The data block for storing up same slitting is referred to as a slitting group.One slitting group includes a slitting or multiple slittings.LUN number
Constituted according to by the LUN data pointed to.
The embodiment of the present invention is during duplicate removal is carried out, while the reliability step of the slitting where considering data block.
In offline duplicate removal, when there are multiple identical blocks, the slitting reliability step where each data block is checked, retains reliable
Data block in property highest slitting, deletes the data block in remaining slitting.When the slitting state of all slittings is not slitting
Unanimously, that is, each slitting occurs in that a certain degree of reliability reduction, can data block write-in slitting is consistent
In slitting, the data in remaining slitting are deleted.
During online duplicate removal, if be stored with storage device with data block identical data block to be stored, detection
Slitting reliability step where the data block stored, if slitting state is not that slitting is consistent, data block write-in point
In the consistent slitting of bar, the data block stored is deleted.
Data block take multiple slittings when, using the reliability step of the minimum slitting of reliability as whole slitting group can
By property grade.Using technology provided in an embodiment of the present invention, slitting can where further considering data block during duplicate removal
By property grade, data block is stored in the higher slitting of reliability.Improve the security of data block.
As shown in Fig. 2 being a kind of data block processing method flow chart, it can carry out deleting again offline.Method includes following step
Suddenly.
Step 31, the fingerprint of data block in storage system is compared controller.
Each data block possesses a fingerprint, for example, can use MD5, data content of the scheduling algorithms of SHA 128 to data block
Calculated, using the result of calculating as data block fingerprint.The data block for possessing identical fingerprints is identical block.Storage is set
The fingerprint of standby middle data block is stored in the fingerprint base of controller.
Step 32, when there are multiple data blocks to have identical fingerprint, the institute of this fingerprint is possessed according to the fingerprint queries
There is the address of data block, the slitting group according to shared by the data block address searches the data block.
When find it is multiple possess identical fingerprints data block when, can retain one of data block, delete remainder data
Block, to reduce the occupancy to storage device memory space.Controller is stored with mapping table, and record has fingerprint and fingerprint in mapping table
The storage address of the data block of representative.Controller can find the slitting group for storing this data block according to the address of data block.
Data block is stored in slitting group, and data block and slitting group are corresponded.Data block address can show as LUN skew, can
Positional information with slitting group where being converted into physical address, data block can be physical address.
Step 33, the independent Redundant Array of Inexpensive Disc RAID slitting states of the slitting group of the data block are verified,
According to check results, the data block in the consistent slitting group of slitting is preserved.
The data block preserved according to check results in the consistent slitting group of slitting, including following at least one schemes:
If there is the consistent slitting group of slitting, and there is the consistent slitting group of non-slitting, then retain slitting one
Data block in the slitting group caused is constant, deletes the data block in remaining described slitting group;It is consistent if there is no slitting
The slitting group, and exist degrade slitting group, then the slitting group degraded by RAID algorithm reparation, generation slitting it is consistent
Slitting group, delete the data block in remaining described slitting group.The verification of slitting state is carried out to each slitting group, slitting group
State minimum slitting of reliability in slitting group is determined.
Reparation described in the embodiment of the present invention, refers to when there is part SU error in data or SU failures occur in slitting,
RAID checking algorithms are used using the data in normal SU, the data being out of order in SU or corrupt data SU are recalculated,
In the data in normal SU and the slitting group of the new write storage device of the data recalculated.Slitting group after reparation
Slitting state is that slitting is consistent, and the security of data block is inconsistent higher than slitting degradation, slitting in slitting group.Number after reparation
According to the specific writing position of block, slitting group can be redistributed;In the case where disclosure satisfy that normal read-write, former point can also be write
Data in bar group, the former slitting group of covering.
A kind of strategy provided in an embodiment of the present invention is:When the slitting state for having any slitting group is that slitting is consistent, say
Data block in this bright slitting is reliable, and the verification that other slitting groups can no longer perform slitting state is directly deleted.
In existing duplicate removal technology, due to not performing slitting verification, therefore, if both there is the consistent slitting of slitting
, there is the consistent slitting group of non-slitting again in group.The data block in one of slitting group can so be retained at random.If retained
It is the data block in the slitting group that slitting is failed, and empties other slitting groups, then the content of data block will be unable to normally be read
Go out, that is to say, that deduplication operation causes loss of data.The consistent slitting group of non-slitting includes other in addition to slitting is consistent
State, such as slitting are inconsistent, slitting degradation or slitting fail.From the point of view of statistical result, in multiple deduplication operation, existing skill
Art can cause loss of data or Information Security to reduce.And the embodiment of the present invention can improve data compared to prior art can
By property.
The embodiment of the present invention also provides another strategy:If there is no the consistent slitting group of slitting, and deposit
In the slitting group of degradation, then it is the consistent slitting group of slitting the slitting group reparation of degradation by RAID algorithm, deletes remaining institute
State the data block in slitting group.
In existing duplicate removal technology, due to slitting group perform slitting verification, therefore can not know slitting group can
By property, the data block after duplicate removal may be stored in the slitting group of degradation, or in the not high slitting group of other stability.This
Inventive embodiments, for such a situation, can repair to degradation slitting group, make the data block finally retained consistent with slitting
Form be stored in slitting group.Therefore Information Security is improved.
It should be noted that both strategies are independent, one kind arbitrarily is performed, storage statistically can be just improved and set
Standby Information Security.Therefore, to performing the device or controller of this method, two kinds of strategies can all be supported, also may be used
With only one of which strategy.
The embodiment of the present invention also provides a kind of optional strategy:If all slitting groups are all failure slitting groups, to institute
State the data in failure slitting group in the consistent slitting of slitting and carry out duplicate removal, and delete the data in the slitting of failure.With it is preceding
As two kinds of strategies, this strategy is also independent, device or controller to performing this method, can be with this any one
Strategy, or any two kinds or three kinds of support are all supported.
When all slitting groups are all failure slitting groups, it is meant that each slitting group there occurs that permanent data is lost, according to
It is not enough to recover whole data block by single slitting group.Slitting group is made up of slitting, and the embodiment of the present invention can save part point
Data in bar.The slitting of redemption may can gather into a complete slitting group, even if the slitting group saved is not enough to gather into
One complete slitting group, retain the data in these slittings remain on it is meaningful, for example future might have it is new
When data write, the data newly write and the slitting stored can gather into a complete slitting group.Therefore this strategy is kept away
Exempt from or reduce the loss of data of data block.
Keeping the measure of slitting includes:If there is the consistent slitting of slitting, then the slitting consistent to these slittings is carried out
Duplicate removal;If there is the slitting of degradation, and the slitting consistent with the slitting of the slitting storage identical data degraded is not present, then makes
The slitting degraded with RAID algorithm reparation;Mistake does not occur if there is the data SU of inconsistent slitting, and inconsistent slitting
By mistake, and in the absence of the slitting consistent with the slitting that inconsistent slitting stores identical data or the slitting of degradation, then use
The slitting that RAID algorithm reparation degrades.
Step 34, the LUN of slitting group points to the consistent slitting group of the slitting where pointing to the data block.
LUN is by controller management and provides main frame and uses.Controller record has in the slitting group pointed by LUN, slitting group
Data chunk into LUN data, when main frame read data when, storage can be found by LUN and slitting group points relationship
Data block in LUN.During to data deduplication, partial data block is deleted, and the data block of reservation is public for these LUN
With, it is therefore desirable to the LUN of slitting group where pointing to those deleted data blocks, it is modified to point to the consistent slitting of slitting
Group.
According to the different situations described by step 33, when the implementation procedure Central Plains of step 33 originally there is the consistent slitting of slitting
During group, these LUN point to the slitting group that this there is originally;If the consistent slitting group of slitting is not present originally, by repairing
During the consistent slitting group of repetitive generation slitting, these LUN point to the consistent slitting group of slitting for repairing generation.
Optionally, the mapping of the storage address for the data block that the fingerprint recorded in controller and fingerprint are represented further is updated
Table.The address that data block is stored is updated to point to the consistent slitting group of slitting.When carrying out duplicate removal next time, this can be used
Corresponding relation finds the slitting group where data block, reaffirms the slitting state of slitting group.
Step 35, optionally, the reference number of times of the data block can also be updated, the increased numerical value of reference number of times, which is equal to, to be deleted
The quantity for the data block removed.Controller record has the number of times that data block is cited, and quotes number of times and points to this data to describe
The LUN of block quantity.When it is 0 to quote number of times, represent that no LUN needs to use this data block, this data block can be by
Delete.Step 35 and step 34 do not limit the precedence of execution, can be first carried out or the two is performed simultaneously with any.
As shown in figure 3, being a kind of data block processing method flow chart, it can carry out deleting again online.Method includes following step
Suddenly.This method can have controller execution, the specifically computer instruction in the computing device caching of controller.
Step 41, inquiry storage device in data block fingerprint base, when in storage device exist and the data to be stored
Block possesses the data storage block of identical fingerprints, the independent Cheap magnetic disc redundant of the detection place of the data storage block slitting group
Array RAID slitting states.
Controller first calculates the fingerprint of this data block to be stored before data block to be stored is stored in storage device,
Then check in fingerprint base whether there is data block fingerprint to be stored.If there is no, it is meant that this data to be stored
Block was not stored, and data block to be stored needs to store into storage device.If there is, it is meant that this data block has been deposited
Stored up, and further determined whether to need to store data block to be stored again.
Step 42, by detecting the RAID slitting states of data storage block place slitting group, if slitting is inconsistent,
Need to generate the consistent slitting group of slitting.Concrete measure, can replace data storage block with data block to be stored, store again
Data block to be stored where slitting group when the consistent slitting group of slitting;If slitting group where data storage block can be with
Repair, the slitting group where data storage block can also be repaired.If slitting state is that slitting is consistent, avoids the need for storage and treat
Data storage block, data storage block is also without change.
Strategy is specifically described below, the embodiment of the present invention can include it is following strategy any one or more.Should
With the controller of the embodiment of the present invention, there can be a variety of tactful functions of support simultaneously, can also only support one of which plan
Slightly.
Tactful A, if the slitting state of slitting group where data storage block is that slitting is consistent, does not store and described waits to deposit
Store up data block.Certainly, it is also possible if replacing data storage block with data block to be stored.
Tactful B, degrades if the slitting state of the slitting group is slitting, stores the data block to be stored, delete
The data storage block, or storage are repaiied according to RAID algorithm by the degradation slitting group to the block of data storage
It is multiple.Data in slitting group in degradation, its failure SU can be by RAID algorithm reparation.
Tactful C, if the slitting state of the slitting group is that slitting is inconsistent, stores the data block to be stored, deletes
Except the data storage block.Or it is determined that being broken down in the absence of data SU, that is to say, that when failure SU is verification SU entirely,
Repair slitting group.Specific repairing method, is that the data in verification SU are recalculated according to RAID algorithm, then by former data SU
Data and recalculate the slitting groups of the data, together write storage device in the verification SU of acquisition.Can be former data SU
Place slitting group or the slitting group applied again.
It is determined that being either with or without the data SU modes broken down:The fingerprint of data in data SU is calculated, if fingerprint and treated
The fingerprint of data storage block is identical, illustrates that corrupted data does not occur for data SU.Fingerprint can be calculated without using verification data.
Step 43, the LUN where data block to be stored is pointed into the consistent slitting group of the slitting.
Be stored with the consistent slitting group of a slitting after the processing of step 42, in storage device, and this slitting is consistent
Slitting group in the data block that stores it is identical with data block to be stored.According to the Different Strategies of step 42, if data storage
The slitting state of block is that slitting is consistent, then LUN points to slitting group where data storage block;If using data block to be stored
Replace data storage block, then LUN points to data block storage to be stored to the slitting group at place after storage device;If to
The slitting group of data storage block is repaired, then LUN points to the slitting group after repairing.
Optionally, the mapping of the storage address for the data block that record has fingerprint and fingerprint to represent in controller is further updated
Table.The address that data block is stored is updated to point to the consistent slitting group of slitting.When carrying out duplicate removal next time, this can be used
Corresponding relation finds the slitting group where data block, reaffirms slitting state.
Step 44, optionally, the reference number of times of block is updated the data.The reference number of times of data storage block is added 1.
As shown in figure 4, being a kind of data block processing unit schematic diagram.Data block processing unit 5 includes fingerprint comparison module
51, address search module 52, consistency desired result module 53 and index module 54.Counting module 55 can also be included.
Fingerprint comparison module 51, is compared for the fingerprint to data block in storage system.
Each data block possesses a fingerprint, for example, can use MD5, data content of the scheduling algorithms of SHA 128 to data block
Calculated, using the result of calculating as data block fingerprint.The data block for possessing identical fingerprints is identical block.Storage is set
The fingerprint of standby middle data block can be stored in the fingerprint base of controller.
Address search module 52, for when there is multiple data blocks to have identical fingerprint, according to the fingerprint queries
The address of data block, the slitting group according to shared by the data block address searches the data block.
When find it is multiple possess identical fingerprints data block when, can retain one of data block, delete remainder data
Block, to reduce the occupancy to storage device memory space.Controller is stored with mapping table, and record has fingerprint and fingerprint in mapping table
The storage address of the data block of representative.Controller can find the slitting group for storing this data block according to the address of data block.
Data block is stored in slitting group, and data block and slitting group are corresponded.
Consistency desired result module 53, the independent Redundant Array of Inexpensive Disc of the slitting group for verifying the data block
RAID slitting states, according to check results, preserve the data block in the consistent slitting group of slitting.
Consistency desired result module 53 can be also used for repairing slitting group.It is described to preserve point of slitting unanimously according to check results
Data block in bar group, including following at least one schemes:If there is the consistent slitting group of slitting, and exist
The consistent slitting group of non-slitting, the then data block retained in the consistent slitting group of slitting is constant, deletes described in remaining
Data block in slitting group;If there is no the consistent slitting group of slitting, and there is the slitting group degraded, then pass through
The slitting group that RAID algorithm reparation degrades, the consistent slitting group of generation slitting, deletes the data block in remaining described slitting group.It is right
Each slitting group carries out the verification of slitting state, state minimum slitting of reliability in slitting group of slitting group.
Reparation described in the embodiment of the present invention, refers to when there is part SU error in data or SU failures occur in slitting,
RAID checking algorithms are used using the data in normal SU, the data being out of order in SU or corrupt data SU are recalculated,
In the data in normal SU and the slitting group of the new write storage device of the data recalculated.Slitting group after reparation
Slitting state is that slitting is consistent, and the security of data block is inconsistent higher than slitting degradation, slitting in slitting group.Number after reparation
According to the specific writing position of block, slitting group can be redistributed;In the case where disclosure satisfy that normal read-write, former point can also be write
Data in bar group, the former slitting group of covering.
A kind of strategy provided in an embodiment of the present invention is:When the slitting state for having any slitting group is that slitting is consistent, say
Data block in this bright slitting is reliable, and the verification that other slitting groups can no longer perform slitting state is directly deleted.
In existing duplicate removal technology, due to not performing slitting verification, therefore, if both there is the consistent slitting of slitting
, there is the consistent slitting group of non-slitting again in group.The data block in one of slitting group can so be retained at random.If retained
It is the data block in the slitting group that slitting is failed, and empties other slitting groups, then the content of data block will be unable to normally be read
Go out, that is to say, that deduplication operation causes loss of data.The consistent slitting group of non-slitting includes other in addition to slitting is consistent
State, such as slitting are inconsistent, slitting degradation or slitting fail.From the point of view of statistical result, in multiple deduplication operation, existing skill
Art can cause loss of data or Information Security to reduce.And the embodiment of the present invention can improve data compared to prior art can
By property.
The embodiment of the present invention also provides another strategy:If there is no the consistent slitting group of slitting, and deposit
In the slitting group of degradation, then it is the consistent slitting group of slitting the slitting group reparation of degradation by RAID algorithm, deletes remaining institute
State the data block in slitting group.
In existing duplicate removal technology, due to slitting group perform slitting verification, therefore can not know slitting group can
By property, the data block after duplicate removal may be stored in the slitting group of degradation, or in the not high slitting group of other stability.This
Inventive embodiments, for such a situation, can repair to degradation slitting group, make the data block finally retained consistent with slitting
Form be stored in slitting group.Therefore Information Security is improved.
It should be noted that both strategies are independent, one kind arbitrarily is performed, storage statistically can be just improved and set
Standby Information Security.Therefore, to performing the device or controller of this method, two kinds of strategies can all be supported, also may be used
With only one of which strategy.
The embodiment of the present invention also provides a kind of optional strategy:If all slitting groups are all failure slitting groups, to institute
State the data in failure slitting group in the consistent slitting of slitting and carry out duplicate removal, and delete the data in the slitting of failure.With it is preceding
As two kinds of strategies, this strategy is also independent, device or controller to performing this method, can be with this any one
Strategy, or any two kinds or three kinds of support are all supported.
When all slitting groups are all failure slitting groups, it is meant that each slitting group there occurs that permanent data is lost, according to
It is not enough to recover whole data block by single slitting group.Slitting group is made up of slitting, and the embodiment of the present invention can save part point
Data in bar.The slitting of redemption may can gather into a complete slitting group, even if the slitting group saved is not enough to gather into
One complete slitting group, retain the data in these slittings remain on it is meaningful, for example future might have it is new
When data write, the data newly write and the slitting stored can gather into a complete slitting group.Therefore this strategy is kept away
Exempt from or reduce the loss of data of data block.
Keeping the measure of slitting includes:If there is the consistent slitting of slitting, then the slitting consistent to these slittings is carried out
Duplicate removal;If there is the slitting of degradation, and the slitting consistent with the slitting of the slitting storage identical data degraded is not present, then makes
The slitting degraded with RAID algorithm reparation;Mistake does not occur if there is the data SU of inconsistent slitting, and inconsistent slitting
By mistake, and in the absence of the slitting consistent with the slitting that inconsistent slitting stores identical data or the slitting of degradation, then use
The slitting that RAID algorithm reparation degrades.
Index module 54, and consistency desired result module 53 are connected, the LUN of slitting group where for will point to the data block
Point to the consistent slitting group of the slitting.
LUN is by controller management and provides main frame and uses.Controller record has in the slitting group pointed by LUN, slitting group
Data chunk into LUN data, when main frame read data when, storage can be found by LUN and slitting group points relationship
Data block in LUN.During to data deduplication, partial data block is deleted, and the data block of reservation is public for these LUN
With, it is therefore desirable to the LUN of slitting group where pointing to those deleted data blocks, it is modified to point to the consistent slitting of slitting
Group.
If storage device Central Plains originally there is slitting consistent slitting group, these LUN are pointed to this by index module 54
Originally the slitting group that there is;If the consistent slitting group of slitting is not present originally, the slitting consistent by repairing generation slitting
During group, index module 54, which points to these LUN, repairs the consistent slitting group of the slitting generated.
Optionally, index module 54 further updates the storage for the data block that the fingerprint recorded in controller and fingerprint are represented
The mapping table of address.The address that data block is stored is updated to point to the consistent slitting group of slitting., can when carrying out duplicate removal next time
To find the slitting group where data block using this corresponding relation, the slitting state of slitting group is reaffirmed.
Counting module 55, and consistency desired result module 53 are connected, and can update the reference number of times of the data block, quote secondary
The increased numerical value of number is equal to the quantity for the data block deleted.Controller record has the number of times that data block is cited, and quotes number of times and uses
With the quantity for the LUN for describing to point to this data block.When it is 0 to quote number of times, represent that no LUN needs to use this data
Block, this data block can be deleted.
As shown in figure 5, being a kind of data block processing unit schematic diagram, it can carry out deleting again online.Data block processing unit 6
Including:Enquiry module 61, consistency desired result module 62 and index module 63.Counting module 64 can also be included.
Enquiry module 61, the fingerprint base for inquiring about data block in storage device, determine in storage device whether there is and
The data block to be stored possesses the data storage block of identical fingerprints.
Controller first calculates the fingerprint of this data block to be stored before data block to be stored is stored in storage device,
Then check in fingerprint base whether there is data block fingerprint to be stored.If there is no, it is meant that this data to be stored
Block was not stored, and data block to be stored needs to store into storage device.If there is, it is meant that this data block has been deposited
Stored up, and further determined whether to need to store data block to be stored again.
Consistency desired result module 62, the RAID slitting states of slitting group where for detecting data storage block, if point
Bar is inconsistent, then needs to generate the consistent slitting group of slitting.Concrete measure, can replace data storage with data block to be stored
Slitting consistent slitting group when block, slitting group where the data block to be stored stored again;If where data storage block
Slitting group can repair, the slitting group where data storage block can also be repaired.If slitting state is that slitting is consistent, just
Data block to be stored need not be stored, data storage block is also without change.
Consistency desired result module 62 also has the function of repairing slitting group or slitting.Below to consistency desired result module 62
Processing strategy be specifically described, the embodiment of the present invention can include it is following strategy any one or more.Using the present invention
The controller of embodiment, can have a variety of tactful functions of support simultaneously, can also only support one of which strategy.
Tactful A, if the slitting state of slitting group where data storage block is that slitting is consistent, does not store and described waits to deposit
Store up data block.Certainly, it is also possible if replacing data storage block with data block to be stored.
Tactful B, degrades if the slitting state of the slitting group is slitting, stores the data block to be stored, delete
The data storage block, or storage are repaiied according to RAID algorithm by the degradation slitting group to the block of data storage
It is multiple.Data in slitting group in degradation, its failure SU can be by RAID algorithm reparation.
Tactful C, if the slitting state of the slitting group is that slitting is inconsistent, stores the data block to be stored, deletes
Except the data storage block.Or it is determined that being broken down in the absence of data SU, that is to say, that when failure SU is verification SU entirely,
Repair slitting group.Specific repairing method, is that the data in verification SU are recalculated according to RAID algorithm, then by former data SU
Data and recalculate the slitting groups of the data, together write storage device in the verification SU of acquisition.Can be former data SU
Place slitting group or the slitting group applied again.
It is determined that being either with or without the data SU modes broken down:The fingerprint of data in data SU is calculated, if fingerprint and treated
The fingerprint of data storage block is identical, illustrates that corrupted data does not occur for data SU.Fingerprint can be calculated without using verification data.
Index module 63, for the LUN where data block to be stored to be pointed into the consistent slitting group of the slitting.
After the processing of consistency desired result module 62, be stored with the consistent slitting group of a slitting, this slitting in storage device
The data block stored in consistent slitting group is identical with data block to be stored.According to the Different Strategies of step 42, if stored
The slitting state of data block is that slitting is consistent, then LUN points to slitting group where data storage block;If using number to be stored
Data storage block is replaced according to block, then LUN points to data block storage to be stored to the slitting group at place after storage device;If
Slitting group to data storage block is repaired, then LUN points to the slitting group after repairing.
Optionally, index module 63 is further used for the data block that record in renewal controller has fingerprint and fingerprint to represent
The mapping table of storage address.The address that data block is stored is updated to point to the consistent slitting group of slitting.Duplicate removal is carried out in next time
When, this corresponding relation can be used to find the slitting group where data block, slitting state is reaffirmed.
Counting module 64, the reference number of times for updating the data block.The reference number of times of data storage block is added 1.Count
Module 64 is optional.
The embodiment of the present invention can also provide a kind of data block processing system, including data block processing unit 5 and storage
Equipment, storage device is used for data storage block.Data block processing unit can be controller or integrated in the controller soft
Part or hardware.
The embodiment of the present invention can also provide a kind of data block processing system, including data block processing unit 6 and storage
Equipment, storage device is used for data storage block.Data block processing unit can be controller or integrated in the controller soft
Part or hardware.
It will be recognized by those of ordinary skill in the art that the possibility implementation of various aspects of the invention or various aspects
System, method or computer program product can be embodied as.Therefore, each aspect of the present invention or various aspects
Possible implementation can using complete hardware embodiment, complete software embodiment (including firmware, resident software etc.), or
The form of the embodiment of integration software and hardware aspect, collectively referred to herein as " circuit ", " module " or " system ".In addition,
The possibility implementation of each aspect of the present invention or various aspects can use the form of computer program product, computer journey
Sequence product refers to be stored in the computer readable program code in computer-readable medium.
Computer-readable medium can be computer-readable signal media or computer-readable recording medium.Computer can
Read storage medium including but not limited to electronics, magnetic, optics, electromagnetism, infrared or semiconductor system, equipment either device or
Foregoing is any appropriately combined, such as random access memory (RAM), read-only storage (ROM), the read-only storage of erasable programmable
Device (EPROM or flash memory), optical fiber, portable read-only storage (CD-ROM).
Processor in computer reads the computer readable program code being stored in computer-readable medium so that place
Reason device is able to carry out function action specified in the combination of each step or each step in flow charts;Generation is implemented in block diagram
Each piece or each piece of combination specified in function action device.
Claims (16)
1. a kind of data processing method, in controller, it is characterised in that this method includes:
When there is multiple data blocks to have identical fingerprint, the address of data block according to the fingerprint queries, according to the number
The slitting group shared by the data block is searched according to block address;
The independent Redundant Array of Inexpensive Disc RAID slitting states of the slitting group of the data block are verified, according to check results
Preserve the data block in the consistent slitting group of slitting;
The data block preserved according to check results in the consistent slitting group of slitting, including following at least one:
If there is the consistent slitting group of slitting, and there is the consistent slitting group of non-slitting, then retain slitting one
Data block in the slitting group caused, deletes the data block in remaining described slitting group;
If there is no the consistent slitting group of slitting, and there is the slitting group degraded, then will by RAID algorithm reparation
The slitting group reparation of degradation is the data block that is stored with the consistent slitting group of slitting, the consistent slitting group of the slitting, is deleted
Except the data block in slitting group remaining described.
2. data processing method according to claim 1, it is characterised in that described consistent according to check results preservation slitting
Slitting group in data block, further comprise:
If all slitting groups are all failure slitting groups, the data in the slitting consistent to slitting in the failure slitting group are entered
Row duplicate removal, deletes the data in the slitting failed in the failure slitting group.
3. data processing method according to claim 1, it is characterised in that further comprise after this method:
The LUN of slitting group points to the consistent slitting group of the slitting where the data block being pointed to.
4. data processing method according to claim 1, it is characterised in that further comprise after this method:
The reference number of times of the data block is updated, the number of times of quoting after renewal is equal to all recovered normal value plus the data block deleted
Quantity.
5. a kind of data block processing unit, it is characterised in that the device includes:
Fingerprint comparison module 51, is compared for the fingerprint to data block in storage device;
Address search module 52, when there is multiple data blocks to have identical fingerprint in the storage device, is looked into according to the fingerprint
The address of the multiple data block is ask, the slitting according to shared by the multiple data block address searches the multiple data block
Group;
Consistency desired result module 53, the independent Redundant Array of Inexpensive Disc of the slitting group for verifying the multiple data block
RAID slitting states, the data block in the consistent slitting group of slitting is preserved according to check results;
The data block preserved according to check results in the consistent slitting group of slitting, including following at least one:
If there is the consistent slitting group of slitting, and there is the consistent slitting group of non-slitting, then retain slitting one
Data block in the slitting group caused, deletes the data block in remaining described slitting group;
If there is no the consistent slitting group of slitting, and there is the slitting group degraded, then will by RAID algorithm reparation
The slitting group reparation of degradation is the data block that is stored with the consistent slitting group of slitting, the consistent slitting group of the slitting, is deleted
Except the data block in slitting group remaining described.
6. data block processing unit according to claim 5, it is characterised in that described to preserve slitting one according to check results
Data block in the slitting group of cause, further comprises:
If all slitting groups are all failure slitting groups, the data in the slitting consistent to slitting in the failure slitting group are entered
Row duplicate removal, deletes the data in the slitting failed in the failure slitting group.
7. data block processing unit according to claim 5, it is characterised in that described device further comprises:
Index module 54, the LUN of slitting group points to the consistent slitting group of the slitting where for will point to the data block.
8. data block processing unit according to claim 5, it is characterised in that described device further comprises:
Counting module 55, the count value for updating the data block, the count value after renewal is equal to all recovered normal value plus deletion
The data block quantity.
9. a kind of storage system, include any one of the claim 6-8 data block processing unit, and the storage is set
It is standby.
10. a kind of data block processing method, in controller, it is characterised in that this method includes:
The fingerprint base of data block in storage device is inquired about, possesses identical fingerprints with data block to be stored when existing in storage device
During data storage block, the independent Redundant Array of Inexpensive Disc RAID slitting shapes of the detection place of the data storage block slitting group
State;
Data block storage, including following one or more of which strategies are carried out according to detection:
Tactful A, if the slitting state of the slitting group is that slitting is consistent, the data block to be stored is not stored;
Tactful B, degrades if the slitting state of the slitting group is slitting,:The data block to be stored is stored, deletes described
Data storage block;Or, the block of data storage is repaired by the degradation slitting group according to RAID algorithm;
Tactful C, if the slitting state of the slitting group is that slitting is inconsistent,:The data block to be stored is stored, institute is deleted
State data storage block;Or, if error in data does not occur for date classification unit, stored according to RAID algorithm to described
Data block is repaired.
11. data block processing method according to claim 10, it is characterised in that also include after methods described:
By the LUN belonging to the data block to be stored, the data block stored according to detection is pointed to.
12. data block processing method according to claim 10, it is characterised in that also include after methods described:
The reference number of times of the block of data storage is added 1.
13. a kind of data block processing unit, it is characterised in that the device includes:
Enquiry module 61, the fingerprint base for inquiring about data block in storage device;
Consistency desired result module 62, for when existing in storage device and data block to be stored possesses the storage numbers of identical fingerprints
During according to block, the independent Redundant Array of Inexpensive Disc RAID slitting states of the detection place of the data storage block slitting group;
Data block storage, including following one or more of which strategies are carried out according to detection:
Tactful A, if the slitting state of the slitting group is that slitting is consistent, the data block to be stored is not stored;
Tactful B, degrades if the slitting state of the slitting group is slitting,:The data block to be stored is stored, deletes described
Data storage block;Or, the block of data storage is repaired by the degradation slitting group according to RAID algorithm;
Tactful C, if the slitting state of the slitting group is that slitting is inconsistent,:The data block to be stored is stored, institute is deleted
State data storage block;Or, if error in data does not occur for date classification unit, stored according to RAID algorithm to described
Data block is repaired.
14. data block processing unit according to claim 13, it is characterised in that described device further comprises:
Index module 63, for by the LUN belonging to the data block to be stored, pointing to the data stored according to detection
Block.
15. data block processing unit according to claim 13, it is characterised in that described device further comprises:
Counting module 64, adds 1 for the reference number of times to the block of data storage.
16. a kind of storage system, include any one of the claim 13-15 data block processing unit, and the storage
Equipment.
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2013/091170 WO2015100639A1 (en) | 2013-12-31 | 2013-12-31 | De-duplication method, apparatus and system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104205097A CN104205097A (en) | 2014-12-10 |
CN104205097B true CN104205097B (en) | 2017-08-25 |
Family
ID=52088178
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201380002564.2A Active CN104205097B (en) | 2013-12-31 | 2013-12-31 | A kind of De-weight method device and system |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN104205097B (en) |
WO (1) | WO2015100639A1 (en) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105760108B (en) * | 2014-12-16 | 2018-12-07 | 华为数字技术(苏州)有限公司 | A kind of method and apparatus of data storage |
CN104822137B (en) * | 2015-04-14 | 2019-06-11 | 宇龙计算机通信科技(深圳)有限公司 | A kind of received method of information and terminal |
CN110557657A (en) * | 2018-05-30 | 2019-12-10 | 视联动力信息技术股份有限公司 | data processing method and system based on video network |
CN112650628A (en) * | 2020-12-30 | 2021-04-13 | 浪潮云信息技术股份公司 | High-availability and expandable data deduplication method |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102221982A (en) * | 2011-06-13 | 2011-10-19 | 北京卓微天成科技咨询有限公司 | Method and system for implementing deletion of repeated data on block-level virtual storage equipment |
CN102460398A (en) * | 2009-06-08 | 2012-05-16 | 赛门铁克公司 | Source classification for performing deduplication in a backup operation |
CN103150260A (en) * | 2011-11-25 | 2013-06-12 | 华为数字技术(成都)有限公司 | Method and device for deleting repeating data |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8412682B2 (en) * | 2006-06-29 | 2013-04-02 | Netapp, Inc. | System and method for retrieving and using block fingerprints for data deduplication |
US8515909B2 (en) * | 2008-04-29 | 2013-08-20 | International Business Machines Corporation | Enhanced method and system for assuring integrity of deduplicated data |
US8082228B2 (en) * | 2008-10-31 | 2011-12-20 | Netapp, Inc. | Remote office duplication |
US8510643B2 (en) * | 2009-12-23 | 2013-08-13 | Nvidia Corporation | Optimizing raid migration performance |
CN101789977B (en) * | 2010-02-08 | 2012-07-25 | 北京同有飞骥科技股份有限公司 | Teledata copying and de-emphasis method based on Hash coding |
-
2013
- 2013-12-31 CN CN201380002564.2A patent/CN104205097B/en active Active
- 2013-12-31 WO PCT/CN2013/091170 patent/WO2015100639A1/en active Application Filing
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102460398A (en) * | 2009-06-08 | 2012-05-16 | 赛门铁克公司 | Source classification for performing deduplication in a backup operation |
CN102221982A (en) * | 2011-06-13 | 2011-10-19 | 北京卓微天成科技咨询有限公司 | Method and system for implementing deletion of repeated data on block-level virtual storage equipment |
CN103150260A (en) * | 2011-11-25 | 2013-06-12 | 华为数字技术(成都)有限公司 | Method and device for deleting repeating data |
Non-Patent Citations (3)
Title |
---|
Understanding data deduplication ratios;Dutch M;《SNIA Data Management Forum》;20080731;第1-13页 * |
带重复数据删除的大规模存储系统可靠性保证;顾瑜等;《清华大学学报(自然科学版)》;20100531;第50卷(第5期);第739-744页 * |
重复数据删除技术的发展及应用;王树鹏;《中兴通讯技术》;20101031;第16卷(第5期);第9-14页 * |
Also Published As
Publication number | Publication date |
---|---|
CN104205097A (en) | 2014-12-10 |
WO2015100639A1 (en) | 2015-07-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109725822B (en) | Method, apparatus and computer program product for managing a storage system | |
US10365983B1 (en) | Repairing raid systems at per-stripe granularity | |
CN101916173B (en) | RAID (Redundant Array of Independent Disks) based data reading and writing method and system thereof | |
US7487289B2 (en) | Apparatus and method for detecting disk write omissions | |
US8943358B2 (en) | Storage system, apparatus, and method for failure recovery during unsuccessful rebuild process | |
US8738963B2 (en) | Methods and apparatus for managing error codes for storage systems coupled with external storage systems | |
US8356292B2 (en) | Method for updating control program of physical storage device in storage virtualization system and storage virtualization controller and system thereof | |
US7908512B2 (en) | Method and system for cache-based dropped write protection in data storage systems | |
CN110413205B (en) | Method, apparatus and computer readable storage medium for writing to disk array | |
US10120769B2 (en) | Raid rebuild algorithm with low I/O impact | |
KR101870521B1 (en) | Methods and systems for improving storage journaling | |
CN107924291B (en) | Storage system | |
CN105468479B (en) | A kind of disk array RAID bad block processing methods and device | |
CN102799533B (en) | Method and apparatus for shielding damaged sector of disk | |
CN104205097B (en) | A kind of De-weight method device and system | |
US10204003B2 (en) | Memory device and storage apparatus | |
US11874741B2 (en) | Data recovery method in storage medium, data recovery system, and related device | |
CN101866307A (en) | Data storage method and device based on mirror image technology | |
CN101236482A (en) | Method for processing data under degrading state and independent redundancy magnetic disc array system | |
US20050091452A1 (en) | System and method for reducing data loss in disk arrays by establishing data redundancy on demand | |
CN102981778A (en) | Redundant array of independent disks (RAID) array reconstruction method and device thereof | |
JP2006139478A (en) | Disk array system | |
CN108141229A (en) | Damage the efficient detection of data | |
US10642508B2 (en) | Method to limit impact of partial media failure of disk drive and detect/report the loss of data for objects due to partial failure of media | |
US20150347224A1 (en) | Storage control apparatus and method therefor |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |