CN106484324A - Method, system and RAID that a kind of RAID rebuilds - Google Patents
Method, system and RAID that a kind of RAID rebuilds Download PDFInfo
- Publication number
- CN106484324A CN106484324A CN201610822559.1A CN201610822559A CN106484324A CN 106484324 A CN106484324 A CN 106484324A CN 201610822559 A CN201610822559 A CN 201610822559A CN 106484324 A CN106484324 A CN 106484324A
- Authority
- CN
- China
- Prior art keywords
- band
- raid
- thread
- queue
- data
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0602—Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
- G06F3/0614—Improving the reliability of storage systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
- G06F3/0629—Configuration or reconfiguration of storage systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0668—Interfaces specially adapted for storage systems adopting a particular infrastructure
- G06F3/0671—In-line storage system
- G06F3/0683—Plurality of storage devices
- G06F3/0689—Disk arrays, e.g. RAID, JBOD
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Debugging And Monitoring (AREA)
Abstract
The invention discloses the method that RAID rebuilds, including：When RAID triggering is rebuild, RAID main thread creates three threads；All include band queue in three threads；The execution thread operation successively of the order according to band in corresponding band queue of three thread parallels, when all band is not contained in each band queue, completes RAID reconstruction；First thread is mainly responsible for read operation；Second thread is mainly responsible for carrying out XOR process to band successively according to the order of band in band queue；3rd thread mainly be responsible for according to band order successively by the data distributing in band to hot standby disk；The work executed in parallel that RAID is rebuild using the mode of streamline, rather than reuse single reconstruct thread RAID is rebuild, the concurrency of CPU is given full play to, the speed for rebuilding RAID is improved, greatly shortens reconstruction time；The invention also discloses the system of RAID reconstruction and RAID, with the effect above.
Description
Technical field
The present invention relates to computer memory technical field, method, system and RAID that more particularly to a kind of RAID rebuilds.
Background technology
RAID (Redundant Arrays of Independent Disks, RAID), raid-array, be
By the relatively inexpensive disk of a lot of prices, a huge disk group of capacity is combined into, is provided produced by data using indivedual disks
The whole disk system efficiency of addition effect promoting.Using this technology, data are cut into many sections, be stored in each respectively
On hard disk.While RAID also provides different degrees of redundant ability (except RAID0), when one piece of disk in array breaks down
When, by redundant data, can be by the data recovery in faulty disk to HotSpare disk.So as to ensure that the complete of user data
Property.
RAID can be divided into different grades according to redundant ability and performance.At present conventional have RAID0, RAID1,
RAID10, RAID5, RAID6, RAID7 etc..
The technology of realizing of RAID mainly has two kinds, and one kind is mirror image technology, and as RAID1, one kind is banding technique, as
RAID5、RAID6.Mirror image technology is the mirror image that can pass through faulty disk when one piece of disk failure directly by data copy to heat
On standby disk, so as to recover faulty disk.Banding technique be when one piece of disk failure, it is possible to use banding technique, using other
Data in disk, recalculate the data on faulty disk on HotSpare disk, so as to recover faulty disk.
Recover data and be referred to as RAID rebuild (reconstruction) to the process of HotSpare disk.For mirror image RAID (such as RAID1,
RAID10 rebuild process) is relatively simple.It realize that process is main in two stages, the first stage is reading one
Data in individual normal mirror, second stage be by read data copy in HotSpare disk.For band RAID (for example
RAID5, RAID6) their rebuild process is relative complex.First stage reads the data in normal disk, second stage pair
The data of reading are calculated, and to recover to lose data, the phase III issues recovery data in HotSpare disk.
Tradition RAID is to process the above three stage using single thread at present, although have using asynchronous copy skill
Art, asynchronous computing calibration technology, but what a certain moment was carried out to the process of mirror image or band or serial.Processing a mirror image
Or during band, other mirror images or band can not be processed.Using traditional RAID, if a disk is broken down, carry out rebuilding and lead to
More than ten hour, or even more than 20 hour will often be spent.Therefore, the speed of RAID reconstruction how is improved, is art technology
Personnel's technical issues that need to address.
Content of the invention
It is an object of the invention to provide method, system and RAID that a kind of RAID rebuilds, will using the mode of streamline
The work executed in parallel that RAID rebuilds, rather than reuse single reconstruct thread RAID is rebuild, give full play to CPU's
Concurrency, improves the speed for rebuilding RAID, that is, greatly shortens reconstruction time.
For solving above-mentioned technical problem, the present invention provides a kind of method that RAID rebuilds, including：
When RAID triggering is rebuild, RAID main thread creates three threads；Wherein, all include the bar leader of a group of people in three threads
Row；The execution thread operation successively of the order according to band in corresponding band queue of three thread parallels, when each band queue
In when all not containing band, complete RAID reconstruction；
Wherein, first thread, for obtaining idle band according to disk sector, and issues reading according to idle strip width
Request of data, and the free time band is added in the band queue of the thread, when in all disks in the free time band
Digital independent finish after corresponding band is sent to the second thread；
Second thread, for carrying out XOR process successively to band according to the order of band in band queue, and will
Corresponding band is sent to the 3rd thread；
3rd thread, for the order according to band in band queue successively by the data distributing in band to hot standby magnetic
Disk.
Wherein, when RAID triggering is rebuild, RAID main thread creates three threads, including：
When RAID has failed disk, triggering RAID rebuilds, and RAID main thread creates three threads, and respectively by three
Thread carries out CPU binding.
Wherein, read band and data in magnetic disk is corresponded to, and corresponding band is sent to the second thread, including：
According to disk sector using fixed data length as step-length application free time band as current band；
After application success, according to initiation read operation request under current strip width and band is added to normal disk in RAID
Queue, in the data for initiating the next step-length of read operation request post processing；
After in the current band, read operation corresponding data successfully returns to the current band buffer area, work as described
Front band is sent to the second thread.
Wherein, XOR process is carried out to band, and corresponding band is sent to the 3rd thread, including：
XOR request is initiated to current band, and is initiating the next band of XOR request post processing；
After the current band successful execution XOR, the current band is sent to the 3rd thread.
Wherein, by the data distributing in band to hot standby disk, including：
The result data of XOR in current band is issued in hot standby disk.
The present invention also provides a kind of RAID system of reconstruction, including：
Main thread module, for when RAID triggering is rebuild, RAID main thread creates three threads；Wherein, three threads
In all include band queue；Grasp successively by execution thread for the order according to band in corresponding band queue of three thread parallels
Make, when all band not being contained in each band queue, complete RAID reconstruction；
First thread module, for obtaining idle band according to disk sector, and issues reading behaviour according to idle strip width
Ask, and the free time band is added in the band queue of the thread, when in all disks in the free time band
Corresponding band is sent to the second thread after finishing by digital independent；
Second threading models, for carrying out XOR process to band successively according to the order of band in band queue,
And corresponding band is sent to the 3rd thread；
3rd threading models, for the order according to band in band queue successively by the data distributing in band to hot standby
Disk.
Wherein, the main thread module is rebuild specifically for the triggering RAID when RAID has failed disk, RAID main line
Journey creates three threads, and respectively three threads is carried out CPU binding.
Wherein, the first thread module, including：
Band application unit, for according to disk sector using fixed data length as step-length application free time band as work as
Front band；
Read operation unit, for applying after success, reads behaviour to normal disk in RAID according to initiation under current strip width
Band queue is asked and adds, in the data for initiating the next step-length of read operation request post processing；
First transmitting element, successfully returns to the current band for working as read operation corresponding data in the current band
After buffer area, the current band is sent to the second thread.
Wherein, second threading models, including：
XOR unit, for initiating XOR request to current band, and locates after XOR request is initiated
The next band of reason；
Second transmitting element, for when the current band successful execution XOR after, the current band being sent
To the 3rd thread.
The present invention also provides a kind of RAID, including：The system that RAID according to any of the above-described rebuilds.
The method that RAID provided by the present invention rebuilds, including：When RAID triggering is rebuild, RAID main thread creates three
Thread；Wherein, all include band queue in three threads；Three thread parallels according to band in corresponding band queue
Order execution thread operation successively, when all band is not contained in each band queue, completes RAID reconstruction；Wherein, first thread,
For idle band being obtained according to disk sector, and issue read operation request according to idle strip width, and the free time bar
Band is added in the band queue of the thread, will corresponding bar after the digital independent in all disks in the free time band is finished
Band is sent to the second thread；Second thread, for carrying out XOR to band successively according to the order of band in band queue
Process, and corresponding band is sent to the 3rd thread；3rd thread, for the order according to band in band queue successively by bar
Data distributing in band is to hot standby disk；
It can be seen that, the process that rebuilds is divided 3 stages by the method, and processing procedure becomes apparent from, and using pipeline processes, fills
Divide and parallel work-flow is utilized, the processing speed of band is improve, the time of RAID reconstruction is shortened, traditional RAID is overcome in weight
When building, each band is serial process, and only one of which thread, it is impossible to give full play to the shortcoming of the advantage of multinuclear；This
The bright system for also providing RAID reconstruction and RAID, with above-mentioned beneficial effect, will not be described here.
Description of the drawings
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
Accompanying drawing to be used needed for technology description is had to be briefly described, it should be apparent that, drawings in the following description are only this
Inventive embodiment, for those of ordinary skill in the art, on the premise of not paying creative work, can also basis
The accompanying drawing of offer obtains other accompanying drawings.
The execution schematic diagram of the method that the RAID that Fig. 1 is provided by the embodiment of the present invention rebuilds；
The structured flowchart of the system that the RAID that Fig. 2 is provided by the embodiment of the present invention rebuilds.
Specific embodiment
The core of the present invention is to provide method, system and the RAID that a kind of RAID rebuilds, will using the mode of streamline
The work executed in parallel that RAID rebuilds, rather than reuse single reconstruct thread RAID is rebuild, give full play to CPU's
Concurrency, improves the speed for rebuilding RAID, that is, greatly shortens reconstruction time.
Purpose, technical scheme and advantage for making the embodiment of the present invention is clearer, below in conjunction with the embodiment of the present invention
In accompanying drawing, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is
The a part of embodiment of the present invention, rather than whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art
The every other embodiment obtained under the premise of creative work is not made, belongs to the scope of protection of the invention.
From the point of view of both the above situation, the process of reconstruction of RAID be fixed mode, each rank section boundary be clearly.
Each stage can be regarded as an independent job.Can be completed using an individual components (or thread) completely.Cause
This, the present embodiment proposes a kind of reconstruction model based on pipeline system, makes full use of the parallel computation of multinuclear, improves and rebuilds
The speed of RAID.So as to greatly shorten reconstruction time.Detailed process is as follows：
Fig. 1 is refer to, the method that the RAID provided by the present embodiment rebuilds can include：
When RAID triggering is rebuild, RAID main thread creates three threads；Wherein, all include the bar leader of a group of people in three threads
Row；The execution thread operation successively of the order according to band in corresponding band queue of three thread parallels, when each band queue
In when all not containing band, complete RAID reconstruction；
Specifically, when RAID has failed disk, triggering RAID rebuilds, and RAID main thread creates three threads, and difference
Three threads are carried out CPU binding.
Wherein, first thread, for obtaining idle band according to disk sector, and issues reading behaviour according to idle strip width
Ask, and the free time band is added in the band queue of the thread, when in all disks in the free time band
Corresponding band is sent to the second thread after finishing by digital independent；It is responsible for reading the data in normal disk.
Second thread, for carrying out XOR process successively to band according to the order of band in band queue, and will
Corresponding band is sent to the 3rd thread；It is responsible for data operation is read, loss of data is prevented, carries out data recovery.
3rd thread, for the order according to band in band queue successively by the data distributing in band to hot standby magnetic
Disk.It is responsible for by the data distributing after recovery in normal HotSpare disk.
Above-mentioned parallel thread process of reconstruction is described below so that RAID5 is rebuild as an example：
Each band can experienced three stages：First stage reads the data in normal disk, and second stage is to reading
Data are calculated, and to recover to lose data, the phase III issues recovery data in HotSpare disk.It is respectively each stage wound
A thread is built, if multi-core CPU, each thread binding is when a CPU core.So equivalent to each core, all independently
Do a kind of work.Also corresponding to each work has individual components to be processed.A bar is preserved in each thread
Band queue, when having band in the queue, the part just carries out the process work of sole duty to band.
From Fig. 1 it can also be seen that after operation is rebuild, more than three afterwards, each moment, three processes existed band number
Corresponding process operation is executed, i.e., when flowing full water conservancy project is made, each cycle can be with 3 bands of parallel processing.Performance is than original
Improve nearly 3 times or so.
Based on technique scheme, the method for the RAID reconstruction that the embodiment of the present invention is carried, will using the mode of streamline
The work executed in parallel that RAID rebuilds, rather than reuse single reconstruct thread RAID is rebuild, give full play to CPU's
Concurrency, improves the speed for rebuilding RAID, that is, greatly shortens reconstruction time.
Based on above-described embodiment, first thread reads band and corresponds to data in magnetic disk, and corresponding band is sent to the second line
Journey, including：
According to disk sector using fixed data length as step-length application free time band as current band；
After application success, according to initiation read operation request under current strip width and band is added to normal disk in RAID
Queue, in the data for initiating the next step-length of read operation request post processing；
After in the current band, read operation corresponding data successfully returns to the current band buffer area, work as described
Front band is sent to the second thread.
Specifically, first thread first has to apply for idle band according to fixed data length (such as 4k).Apply successfully it
Afterwards, just read operation is initiated to normal disk.Begin to process the data of next step-length (i.e. according to fixed number after issuing end
Apply for idle band according to length (such as 4k), after application success, just read operation is initiated to normal disk).When a band
After all of data are all successfully returned, just this band is handed to the next stage (i.e. the second thread) and go to process, while crying
The thread of awake second stage.
Based on above-described embodiment, the second thread carries out XOR process to band, and corresponding band is sent to the 3rd
Thread, including：
XOR request is initiated to current band, and is initiating the next band of XOR request post processing；
After the current band successful execution XOR, the current band is sent to the 3rd thread.
Specifically, the process thread of second stage starts to carry out XOR to band.As XOR is also usual
Calculated using asynchronous interface, so the groundwork of the second thread is to be that asynchronous processing preparation source data and operation result are slow
Area is rushed, and initiates XOR request.In addition also having the conversion of band state and abnormal operation to process (is for example used for working as XOR
When computing executes failure, corresponding concrete operations).Once initiate start to process next one band after XOR request.If different
Or computing runs succeeded, then this band is handed to the next stage (i.e. the 3rd thread), and wake the thread of phase III.
Based on above-described embodiment, the 3rd thread by the data distributing in band to hot standby disk, including：
The result data of XOR in current band is issued in hot standby disk.
Specifically, the process thread (i.e. the 3rd thread) of three phases starts to be issued to calculation result data in band
Disk.The groundwork in this stage is issued in HotSpare disk to the result data for calculating as BIO data.Such a heavy
The normal flowing water for building band just finishes.
Thread binding to CPU core is just released at the end of the reconstruction, while the thread is also carried out destroying, save space.
Illustrate by taking RAID5 data disks as an example below, RAID5 data disks have three pieces of disks to constitute, and are disk0 respectively,
Disk1, disk2, have 1 HotSpare disk S0.When disk2 disk breaks down, RAID will start to carry out disk0 and disk1 data
XOR, calculates the data of disk2, returns on HotSpare disk S0.Its detailed process is as follows：
1st, when for example disk2 disk breaks down after have a disk to break down in RAID, triggering reconstruct will be kicked out of by RAID.
2nd, RAID main thread creates 3 threads th1, th2, th3, and carries out CPU binding respectively.
3rd, thread th1 is responsible for traveling through all sectors of whole HotSpare disk, with 4k as step-length, apply band, read disk0 and
Data in disk1 are to band buffering area.Next band is just processed after initiating read requests；While waking th2.
4th, thread th2 is responsible for data tissue in disk0 and disk1 to transferring to XOR part to carry out XOR together.
Next band is just processed after initiating XOR request；While waking th3.
5th, thread th3 is responsible for reorganizing the result of XOR, and by data distributing in HotSpare disk S0.
6th, after the completion of the traversal of all sectors, all data have just been recovered.
Based on the method that technique scheme, RAID provided in an embodiment of the present invention are rebuild, the process that rebuilds is divided 3
Stage, processing procedure become apparent from, and using pipeline processes, make full use of parallel work-flow, improve the processing speed of band, contracting
The time that short RAID rebuilds, traditional RAID is overcome when rebuilding, each band is serial process, and only one of which
Thread, it is impossible to give full play to the shortcoming of the advantage of multinuclear.
The system rebuild by RAID provided in an embodiment of the present invention and RAID are introduced below, RAID weight described below
The method that the system that builds and RAID are rebuild with above-described RAID can be mutually to should refer to.
Refer to the structured flowchart of the system of the RAID reconstruction that Fig. 2, Fig. 2 are provided by the embodiment of the present invention；The system can
To include：
Main thread module 100, for when RAID triggering is rebuild, RAID main thread creates three threads；Wherein, three lines
All include band queue in journey；The order according to band in corresponding band queue of three thread parallels execution thread successively
Operation, when all band is not contained in each band queue, completes RAID reconstruction；
First thread module 200, for obtaining idle band according to disk sector, and issues reading according to idle strip width
Operation requests, and the free time band is added in the band queue of the thread, when in all disks in the free time band
Digital independent finish after corresponding band is sent to the second thread；
Second threading models 300, for carrying out at XOR to band according to the order of band in band queue successively
Reason, and corresponding band is sent to the 3rd thread；
3rd threading models 400, for arriving the data distributing in band according to the order of band in band queue successively
Hot standby disk.
Based on above-described embodiment, the main thread module 100 specifically for triggering RAID when RAID has failed disk
Rebuild, RAID main thread creates three threads, and respectively three threads is carried out CPU binding.
Based on above-described embodiment, the first thread module 200, including：
Band application unit, for according to disk sector using fixed data length as step-length application free time band as work as
Front band；
Read operation unit, for applying after success, reads behaviour to normal disk in RAID according to initiation under current strip width
Band queue is asked and adds, in the data for initiating the next step-length of read operation request post processing；
First transmitting element, successfully returns to the current band for working as read operation corresponding data in the current band
After buffer area, the current band is sent to the second thread.
Based on above-described embodiment, second threading models 300, including：
XOR unit, for initiating XOR request to current band, and locates after XOR request is initiated
The next band of reason；
Second transmitting element, for when the current band successful execution XOR after, the current band being sent
To the 3rd thread.
Based on above-described embodiment, the 3rd threading models 400 are specially the number of results of XOR in current band
According to being issued in hot standby disk.
Based on above-mentioned any embodiment, the present invention also provides a kind of RAID, including：According to above-mentioned any embodiment
The system that RAID rebuilds.
Wherein, the RAID can also be able to be distributed raid for tradition RAID.
In specification, each embodiment is described by the way of going forward one by one, and what each embodiment was stressed is and other realities
Apply the difference of example, between each embodiment identical similar portion mutually referring to.For device disclosed in embodiment
Speech, as which corresponds to the method disclosed in Example, so description is fairly simple, related part is referring to method part illustration
?.
Professional further appreciates that, in conjunction with the unit of each example of the embodiments described herein description
And algorithm steps, can with electronic hardware, computer software or the two be implemented in combination in, in order to clearly demonstrate hardware and
The interchangeability of software, generally describes composition and the step of each example in the above description according to function.These
Function is executed with hardware or software mode actually, the application-specific depending on technical scheme and design constraint.Specialty
Technical staff can use different methods to realize described function to each specific application, but this realization should
Think beyond the scope of this invention.
The step of method described in conjunction with the embodiments described herein or algorithm, directly can be held with hardware, processor
Capable software module, or the combination of the two is implementing.Software module can be placed in random access memory (RAM), internal memory, read-only deposit
Reservoir (ROM), electrically programmable ROM, electrically erasable ROM, register, hard disk, moveable magnetic disc, CD-ROM or technology
In any other form of storage medium well known in field.
Above method, system and the RAID that RAID provided by the present invention rebuilds is described in detail.Herein should
The principle of the present invention and embodiment are set forth with specific case, the explanation of above example is only intended to help and manages
The solution method of the present invention and its core concept.It should be pointed out that for those skilled in the art, without departing from
On the premise of the principle of the invention, some improvement and modification can also be carried out to the present invention, these improve and modification also falls into this
In bright scope of the claims.
Claims (10)
1. a kind of method that RAID rebuilds, it is characterised in that include：
When RAID triggering is rebuild, RAID main thread creates three threads；Wherein, all include band queue in three threads；
The execution thread operation successively of the order according to band in corresponding band queue of three thread parallels, when equal in each band queue
When band is not contained, RAID reconstruction is completed；
Wherein, first thread, for obtaining idle band according to disk sector, and issue read operation according to idle strip width please
Ask, and the free time band is added in the band queue of the thread, the data in all disks in the free time band
Corresponding band is sent to the second thread after finishing by reading；
Second thread, for carrying out XOR process successively to band according to the order of band in band queue, and will be corresponding
Band is sent to the 3rd thread；
3rd thread, for the order according to band in band queue successively by the data distributing in band to hot standby disk.
2. the method that RAID according to claim 1 rebuilds, it is characterised in that when RAID triggering is rebuild, RAID main line
Journey creates three threads, including：
When RAID has failed disk, triggering RAID rebuilds, and RAID main thread creates three threads, and respectively by three threads
Carry out CPU binding.
3. the method that RAID according to claim 2 rebuilds, it is characterised in that idle band is obtained according to disk sector,
And reading request of data is issued according to idle strip width, and the free time band is added in the band queue of the thread,
Corresponding band is sent to the second thread after the digital independent in all disks in the free time band is finished, including：
According to disk sector using fixed data length as step-length application free time band as current band；
After application success, according to initiation read operation request under current strip width and bar is added to lead group of people normal disk in RAID
Row, in the data for initiating the next step-length of read operation request post processing；
After read operation corresponding data successfully returns to the current band buffer area in the current band, work as preceding article by described
Band is sent to the second thread.
4. the method that RAID according to claim 3 rebuilds, it is characterised in that carry out XOR process to band, and
Corresponding band is sent to the 3rd thread, including：
XOR request is initiated to current band, and is initiating the next band of XOR request post processing；
After the current band successful execution XOR, the current band is sent to the 3rd thread.
5. the method that RAID according to claim 4 rebuilds, it is characterised in that by the data distributing in band to hot standby magnetic
Disk, including：
The result data of XOR in current band is issued in hot standby disk.
6. the system that a kind of RAID rebuilds, it is characterised in that include：
Main thread module, for when RAID triggering is rebuild, RAID main thread creates three threads；Wherein, in three threads all
Include band queue；The execution thread operation successively of the order according to band in corresponding band queue of three thread parallels,
When all band not being contained in each band queue, complete RAID reconstruction；
First thread module, for obtaining idle band according to disk sector, and issue read operation according to idle strip width please
Ask, and the free time band is added in the band queue of the thread, the data in all disks in the free time band
Corresponding band is sent to the second thread after finishing by reading；
Second threading models, for carrying out XOR process successively to band according to the order of band in band queue, and will
Corresponding band is sent to the 3rd thread；
3rd threading models, for the order according to band in band queue successively by the data distributing in band to hot standby magnetic
Disk.
7. the system that RAID according to claim 6 rebuilds, it is characterised in that the main thread module is specifically for working as
When RAID has failed disk, triggering RAID rebuilds, and RAID main thread creates three threads, and respectively three threads is carried out CPU
Binding.
8. the system that RAID according to claim 7 rebuilds, it is characterised in that the first thread module, including：
Band application unit, for according to disk sector using fixed data length as step-length application free time band as working as preceding article
Band；
Read operation unit, for applying after success, please according to initiation read operation under current strip width to normal disk in RAID
Band queue is asked and adds, in the data for initiating the next step-length of read operation request post processing；
First transmitting element, successfully returns to the current band caching for working as read operation corresponding data in the current band
Qu Hou, the current band is sent to the second thread.
9. the system that RAID according to claim 8 rebuilds, it is characterised in that second threading models, including：
XOR unit, for initiating XOR request, and under XOR request post processing is initiated to current band
One band；
Second transmitting element, for when the current band successful execution XOR after, the current band being sent to
Three threads.
10. a kind of RAID, it is characterised in that include：The system that RAID according to claim 6 or 9 any one rebuilds.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610822559.1A CN106484324A (en) | 2016-09-13 | 2016-09-13 | Method, system and RAID that a kind of RAID rebuilds |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610822559.1A CN106484324A (en) | 2016-09-13 | 2016-09-13 | Method, system and RAID that a kind of RAID rebuilds |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106484324A true CN106484324A (en) | 2017-03-08 |
Family
ID=58273727
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610822559.1A Pending CN106484324A (en) | 2016-09-13 | 2016-09-13 | Method, system and RAID that a kind of RAID rebuilds |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106484324A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108334280A (en) * | 2017-12-28 | 2018-07-27 | 创新科存储技术（深圳）有限公司 | A kind of RAID5 disks group fast reconstructing method and device |
| WO2020093205A1 (en) * | 2018-11-05 | 2020-05-14 | 深圳市欢太科技有限公司 | Deep learning computation method and related device |
| CN111427516A (en) * | 2020-03-27 | 2020-07-17 | 杭州宏杉科技股份有限公司 | RAID reconstruction method and device |
| CN116149576A (en) * | 2023-04-20 | 2023-05-23 | 北京大学 | Method and system for reconstructing disk redundant array oriented to server non-perception calculation |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101615146A (en) * | 2009-07-08 | 2009-12-30 | 中国科学院计算技术研究所 | Disc array online reconstruction system and method |
| CN103186438A (en) * | 2013-04-02 | 2013-07-03 | 浪潮电子信息产业股份有限公司 | Method of improving disk array data reconstruction efficiency |
| CN105183583A (en) * | 2015-07-31 | 2015-12-23 | 浪潮电子信息产业股份有限公司 | Method for data reconstruction of disk array, and disk array system |
| US20160072525A1 (en) * | 2011-12-30 | 2016-03-10 | Streamscale, Inc. | Accelerated erasure coding system and method |
| CN105487825A (en) * | 2015-12-08 | 2016-04-13 | 浙江宇视科技有限公司 | RAID array reconstruction method and device |
-
2016
- 2016-09-13 CN CN201610822559.1A patent/CN106484324A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101615146A (en) * | 2009-07-08 | 2009-12-30 | 中国科学院计算技术研究所 | Disc array online reconstruction system and method |
| US20160072525A1 (en) * | 2011-12-30 | 2016-03-10 | Streamscale, Inc. | Accelerated erasure coding system and method |
| CN103186438A (en) * | 2013-04-02 | 2013-07-03 | 浪潮电子信息产业股份有限公司 | Method of improving disk array data reconstruction efficiency |
| CN105183583A (en) * | 2015-07-31 | 2015-12-23 | 浪潮电子信息产业股份有限公司 | Method for data reconstruction of disk array, and disk array system |
| CN105487825A (en) * | 2015-12-08 | 2016-04-13 | 浙江宇视科技有限公司 | RAID array reconstruction method and device |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108334280A (en) * | 2017-12-28 | 2018-07-27 | 创新科存储技术（深圳）有限公司 | A kind of RAID5 disks group fast reconstructing method and device |
| CN108334280B (en) * | 2017-12-28 | 2021-01-08 | 深圳创新科技术有限公司 | RAID5 disk group fast reconstruction method and device |
| WO2020093205A1 (en) * | 2018-11-05 | 2020-05-14 | 深圳市欢太科技有限公司 | Deep learning computation method and related device |
| CN112714917A (en) * | 2018-11-05 | 2021-04-27 | 深圳市欢太科技有限公司 | Deep learning calculation method and related equipment |
| CN112714917B (en) * | 2018-11-05 | 2024-05-10 | 深圳市欢太科技有限公司 | Deep learning calculation method and related equipment |
| CN111427516A (en) * | 2020-03-27 | 2020-07-17 | 杭州宏杉科技股份有限公司 | RAID reconstruction method and device |
| CN116149576A (en) * | 2023-04-20 | 2023-05-23 | 北京大学 | Method and system for reconstructing disk redundant array oriented to server non-perception calculation |
| CN116149576B (en) * | 2023-04-20 | 2023-07-25 | 北京大学 | Method and system for reconstructing disk redundant array oriented to server non-perception calculation |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106484324A (en) | Method, system and RAID that a kind of RAID rebuilds | |
| CN110187830A (en) | A kind of method and system accelerating disk array reconstruction | |
| CN101727299B (en) | RAID5-orientated optimal design method for writing operation in continuous data storage | |
| CN102023902B (en) | Disc array reconstruction method | |
| CN103984642B (en) | The reconstruction of redundancy auxiliary memory caching | |
| CN101609420A (en) | Realize method and the redundant arrays of inexpensive disks and the controller thereof of rebuilding of disc redundant array | |
| CN107273048A (en) | A kind of method for writing data and device | |
| CN104102556B (en) | A kind of magnetic disk of virtual machine data backup and restoration methods | |
| JP2008004069A (en) | Method of improving input and output performance of raid system using matrix stripe cache | |
| WO2021151298A1 (en) | Data redundancy processing method and apparatus, device, and storage medium | |
| CN101251812A (en) | Method for cluster system data fault tolerance | |
| CN101729421B (en) | Storage method and device based on time division multiplex | |
| CN104137082B (en) | Method and device for writing data of first block size | |
| Zeng et al. | HRAID6ML: A hybrid RAID6 storage architecture with mirrored logging | |
| CN116719484A (en) | Method, device, equipment and medium for processing write data of disk array | |
| US20110185119A1 (en) | Parity generator for redundant array of independent discs type memory | |
| CN101587744A (en) | Multi-level data redundancy method of large scale FLASH memory array | |
| CN108920094A (en) | Wipe unit R AID method, apparatus, computer equipment and storage medium | |
| CN105930099B (en) | The fault-tolerant disk array of double plate of small write operation in a kind of elimination local parallel | |
| CN102103533A (en) | Method for reconstructing single disk in double-disk fault-tolerance disk array | |
| CN103488547A (en) | Rapid reconstruction method of RAID group fault hard disk | |
| CN206696842U (en) | A kind of RAID verifies generating means | |
| CN110532128A (en) | A kind of correcting and eleting codes coding and data reconstruction method for reducing data and updating cost | |
| Wu et al. | JOR: A journal-guided reconstruction optimization for RAID-structured storage systems | |
| CN105138468B (en) | A kind of method and apparatus improving virtual disk group reconstruction property |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170308 |