CN106293543A - A kind of based on incidentally reclaim watt record a RAID write ordered method and device - Google Patents
A kind of based on incidentally reclaim watt record a RAID write ordered method and device Download PDFInfo
- Publication number
- CN106293543A CN106293543A CN201610726634.4A CN201610726634A CN106293543A CN 106293543 A CN106293543 A CN 106293543A CN 201610726634 A CN201610726634 A CN 201610726634A CN 106293543 A CN106293543 A CN 106293543A
- Authority
- CN
- China
- Prior art keywords
- write
- band
- incidentally
- data block
- reclaimed
- 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.)
- Granted
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
- G06F3/061—Improving I/O performance
-
- 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
- G06F3/0638—Organizing or formatting or addressing of data
- G06F3/064—Management of blocks
-
- 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)
- Signal Processing For Digital Recording And Reproducing (AREA)
Abstract
The present invention proposes a kind of to write ordered method and device, Computer Storage field based on watt record RAID incidentally reclaimed, and this device includes configuration management module, for being managed by the structure of circular logging by band;The most persistently cache module, for receiving all non-sequential write requests, when the most persistently, caching triggers lower brush when being written to water level, wherein band is divided into data band and verification band, lower brush is carried out in units of group, write between group does not produces writes covering, and in group, the process of writing of non-whole piece band triggers small letter process, does incidentally reclaimer operation when little write-read simultaneously.The present invention starts small letter after all little write-reads all complete again write, in band, during little write-read, incidentally do recovery read, the valid data read will be reclaimed and to play the valid data of brush to merge into the sequential write of bulk during small letter is write, avoid the RMW process produced with interior small letter, and promote the overall performance of system.
Description
Technical field
The present invention relates to Computer Storage field, write ordered particularly to a kind of based on watt record RAID incidentally reclaimed
Method and device.
Background technology
Surface density as the traditional magnetic disk of primary storage medium quickly will be close to 1TB/In2The limit, also implies that
The memory space of single-deck has been difficult to have greatly improved again, in order to solve this problem, in recent years, and a kind of new disk skill
Art is proposed in the industry, i.e. watt record (Shingled Magnetic Recording, SMR) technology, and it uses partial stack adjacent
The mode of magnetic track promotes the memory density of disk further, but, the mode of this superposition magnetic track is increasing disk storage appearance
But disk is made directly to update in original place while amount, because updating the data write on a tracks can cover its adjacent track
On data, for avoiding the damage of adjacent track data, need make reading amendment write (Read-Modify-Write, RMW), as
As name WD, first reading section data from disk, then revise a part of data therein, the most again will amendment
After write back disk together with unmodified data, the part data wherein read to ensure will not cover on disk when writing back
The valid data of other positions, therefore, the worst in the case of, if disk is filled with valid data, RMW may need operation
Whole disk, causes serious writing amplification, reduces the non-sequential write performance of disk greatly, write amplification to reduce, controlled by RMW
System within the specific limits, reduces write performance expense, disk can be divided into continuous print region, i.e. band (Band), as shown in Figure 1
(assume to write that a tracks (Track) is farthest only can cover the data on next tracks, i.e. coverage rate K=2, each Band by
Article two, Track composition), the addressing between Band is continuous print, but needs reserved certain space (Gap) between Band, it is ensured that write
Optional position on Band0, all without any data covered on Band1, controls at one in this manner it is possible to amplification will be write
In the range of Band, therefore write the size (Band Size) that magnification level depends on Band, owing to needing with Band partitioning discs
Headspace between adjacent Band, wastes disk storage capacity, and for reducing disk size waste, Band Size can not be the least,
Because Band Size is the least, on disk, the quantity of Band is the most, and disk space waste is the most serious, even if Band is Size
Being set to about tens MB, each write request of user may cause RMW, needs whole Band in the worst case
On data all read, revised and be written back, performance also can decline to a great extent.
For RAID5 6 or the RAID system of more higher-dimension tape verifying, under non-sequential write field scape, due to depositing of small letter
, each band of disk can introduce the most non-sequential read-write operation, due to the non-sequential write performance of watt recording disc
Very poor, the non-sequential write performance of RAID system entirety can be affected by serious.
Summary of the invention
For the deficiencies in the prior art, the present invention propose a kind of based on incidentally reclaim watt record a RAID write ordered method
And device.
The present invention propose a kind of based on incidentally reclaim watt record a RAID write ordered device, including:
Configuration management module, for being managed band by the structure of circular logging;
The most persistently cache module, is used for receiving all non-sequential write requests, when the most persistently caching is written to water level
Triggering lower brush, band is wherein divided into data band and verification band, lower brush is carried out in units of group, and the write between group does not produces to be write
Covering, in group, the process of writing of non-whole piece band triggers small letter process, does incidentally reclaimer operation when little write-read simultaneously.
There is on each disk band one group of composition of identical reel number.
Described circular logging includes head position, tail position, after small letter runs through, by the write request write head of small letter
Position, portion, and head position is moved forward, incidentally reclaim and read one section from described tail position to described head position sequential scan
The data of continuous space, scanning comprises invalid data block, extracts valid data block from the data block read, and will effectively
Data block writes described head position, thus accomplishes to be converted into the random write falling in band and random write random write and order
Write.
Distance N is there is in described head position with described tail position.
When carrying out non-sequential write request, described data band is relative with the data block that will write with described verification band
The data block answered and check block read, and combine the check block that the described data block that will write calculating is new, and by described will
The data block of write is interior with described new check block write band.
The present invention also propose a kind of based on incidentally reclaim watt record a RAID write ordered method, including:
Step 1, is managed band by the structure of circular logging;
Step 2, receives all non-sequential write requests, and when the most persistently, caching triggers lower brush when being written to water level, wherein will
Band is divided into data band and verification band, and lower brush is carried out in units of group, and the write between group does not produces writes covering, non-whole piece in group
The process of writing of band triggers small letter process, does incidentally reclaimer operation when little write-read simultaneously.
There is on each disk band one group of composition of identical reel number.
Described circular logging includes head position, tail position, after small letter runs through, by the write request write head of small letter
Position, portion, and head position is moved forward, incidentally reclaim and read one section from described tail position to described head position sequential scan
The data of continuous space, scanning comprises invalid data block, extracts valid data block from the data block read, and will effectively
Data block writes described head position, thus accomplishes to be converted into the random write falling in band and random write random write and order
Write.
Distance N is there is in described head position with described tail position.
When carrying out non-sequential write request, described data band is relative with the data block that will write with described verification band
The data block answered and check block read, and combine the check block that the described data block that will write calculating is new, and by described will
The data block of write is interior with described new check block write band.
From above scheme, it is an advantage of the current invention that:
The present invention assembles non-sequential write request by the most persistently caching, will fall into the little write-read in each band when lower brush
Write two processes that are split into small letter, i.e. after all little write-reads all complete, start small letter again and write, in band during little write-read
Incidentally do recovery to read, the valid data read will be reclaimed and to play the valid data of brush to merge into bulk during small letter is write
Sequential write, it is to avoid the RMW process that in band, small letter produces, and promote the overall performance of system.
Accompanying drawing explanation
Fig. 1 is that disc space is by band division figure;
Fig. 2 is circular logging structure chart in band;
Fig. 3 is system architecture diagram;
Fig. 4 is the band random write and sequential write figure incidentally reclaimed.
Detailed description of the invention
The structure that the present invention have employed circular logging to the band of watt recording disc is managed, as in figure 2 it is shown, write request is write
Entering head pointer position, and moved forward by head pointer, black region is valid data block, and shadow region produces owing to updating operation
Invalid data block, owing to watt recording technique does not support that original place updates, invalid data block can only by reclaimer operation could again by
Utilizing, reclaimer operation is the data reading one section of continuous space at tail pointer to head pointer direction sequential scan, and scanning needs bag
Containing invalid data block, from the data block read, extract valid data block, and by valid data block write head pointer position,
Thus discharge the space that invalid data block takies, it is achieved the release in space in band.In band, circular logging structure guarantees to write is suitable
Sequence, it is to avoid the RMW operation of watt recording disc, present invention improves over the small letter flow process of tradition RAID so that writing of small letter
Journey becomes the sequential write in band, thus promotes watt record non-sequential write performance of RAID system.
The present invention adds a floor height speed persistently caching (such as: SSD) on tradition RAID, all non-sequential for receiving
Write request, when at a high speed persistently caching triggers lower brush when being written to water level, such as Fig. 3, and is divided into data band and verification band by band, under
Brushing with group (Group) is that unit carries out (having band one group of composition of identical reel number in Fig. 3 in dotted line frame in different disk), group
Between write will not produce and write covering, in group, non-whole piece band is write and can be triggered small letter process, does recovery when little write-read and reads, such as figure
4, introduce during small letter in band and once reclaim read operation, after all small letter read operations all complete in band, small letter to be write
Enter the data in band together with reclaiming the valid data read at the head pointer of the write band of order, thus accomplish to fall in band
Random write and random write be converted into random write and sequential write, only introduced the random expense read that reclaims in a sub-band, but will
In band, multiple random write merge for sequential write, it is to avoid the RMW process of watt recording disc, and the small letter mistake to tradition RAID5
Cheng Jinhang is ordered.
The present invention propose a kind of based on incidentally reclaim watt record a RAID write ordered device, including:
Configuration management module, for being managed band by the structure of circular logging;
The most persistently cache module, is used for receiving all non-sequential write requests, when the most persistently caching is written to water level
Triggering lower brush, band is wherein divided into data band and verification band, lower brush is carried out in units of group, and the write between group does not produces to be write
Covering, in group, the process of writing of non-whole piece band triggers small letter process, does incidentally reclaimer operation when little write-read simultaneously.
There is on each disk band one group of composition of identical reel number.
Described circular logging includes head position, tail position, after small letter runs through, by the write request write head of small letter
Position, portion, and head position is moved forward, incidentally reclaim and read one section from described tail position to described head position sequential scan
The data of continuous space, scanning comprises invalid data block, extracts valid data block from the data block read, and will effectively
Data block writes described head position, thus accomplishes to be converted into the random write falling in band and random write random write and order
Write.
Distance N is there is in described head position with described tail position.
When carrying out non-sequential write request, described data band is relative with the data block that will write with described verification band
The data block answered and check block read, and combine the check block that the described data block that will write calculating is new, and by described will
The data block of write is interior with described new check block write band.
The detailed description of the invention of the present invention is given below, in conjunction with accompanying drawing, the present invention is made that detailed description, wherein this
Bright include non-sequential write request assemble and incidentally reclaim write ordered two parts.
Non-sequential write request is assembled:
As it is shown on figure 3, all non-sequential write (D8 ', D3 ', D1 ') are write in the most lasting caching (such as SSD), when persistently
After caching is written into certain water level (water level), brush operation under triggering, it is assumed that data block D8 ', D3 ', D1 ' should be write
Enter in Disk0:Band1, owing to being that non-full band is write, RAID5 small letter flow process (assume be read rewrite) can be triggered, need to first by
Check block P8, P3, P1 old in data block D8 old in Disk0:Band1, D3, D1 and Disk1:Band1 reads (random in band
Read), with new D8 ', D3 ', D1 ' calculate verification P8 ', P3 ', the P1 ' made new advances, then new data block and check block write back in band.
Writing of incidentally reclaiming is ordered:
Such as Fig. 4, to read old D8, D3, D1, recovery read and write new D8 ', D3 ', D1 ', reclaim the valid data read and be
Example, during little write-read, random write D8, D3, D1 are original little write-read expenses, incidentally reclaim and introduce the most extra random write, i.e. return
Receive and read, when all little write-reads and after reclaiming and reading all to complete in band, by suitable for the valid data and new D8 ', D3 ', D1 ' reclaiming reading
At the head pointer of sequence write band, the maximum distance write farthest between head pointer and the tail pointer that distance is circular logging in Fig. 4,
Write covering owing to a watt recording disc adjacent track can produce, certain distance N therefore between head pointer and tail pointer, must be kept, its
The size of middle N had both depended on that coverage rate K write in above-mentioned watt of record, i.e. write the track number that a tracks can cover farthest
Amount, again with on adjacent track the skew of sector about (physical location of the sector on adjacent track is not necessarily to radius
Direction alignment), the skill level that the product that therefore N value is concrete with manufacturer uses is relevant, the size ≈ magnetic track of N in the present invention
Number of sectors * (K-1), N is in units of sector.
For the RMW process in units of band being not optimised, little write-read is original expense, and writing on of small letter carries interior right and wrong
Sequential write, the worst situation needs all to read the data in whole band, is written back after merging with new data, and the present invention is only
Introduce once the random expense reclaiming reading, and the random write of original small letter is converted to sequential write, improve system
Overall performance.
Claims (10)
1. write ordered device based on watt record RAID incidentally reclaimed for one kind, it is characterised in that including:
Configuration management module, for being managed band by the structure of circular logging;
The most persistently cache module, is used for receiving all non-sequential write requests, and when the most persistently, caching triggers when being written to water level
Lower brush, is wherein divided into data band and verification band by band, and lower brush is carried out in units of group, and the write between group does not produces to write to be covered
Lid, in group, the process of writing of non-whole piece band triggers small letter process, does incidentally reclaimer operation when little write-read simultaneously.
2. write ordered device based on watt record RAID incidentally reclaimed as claimed in claim 1, it is characterised in that each disk
On have identical reel number band composition one group.
3. as claimed in claim 1 write ordered device based on watt record RAID incidentally reclaimed, it is characterised in that described in follow
Ring daily record includes head position, tail position, after small letter runs through, by the write request write head position of small letter, and by head
Position, portion moves forward, and incidentally reclaims the number reading one section of continuous space from described tail position to described head position sequential scan
According to, scanning comprises invalid data block, extracts valid data block, and valid data block is write institute from the data block read
State head position, thus accomplish the random write falling in band and random write are converted into random write and sequential write.
4. write ordered device based on watt record RAID incidentally reclaimed as claimed in claim 3, it is characterised in that described head
Distance N is there is in position, portion with described tail position.
5. write ordered device based on watt record RAID incidentally reclaimed as claimed in claim 1, it is characterised in that when carrying out
During non-sequential write request, the data block corresponding with the data block that will write with described verification band by described data band and school
Test block to read, and combine the described data block that will write and calculate new check block, and by the described data block that will write and
In described new check block write band.
6. write ordered method based on watt record RAID incidentally reclaimed for one kind, it is characterised in that including:
Step 1, is managed band by the structure of circular logging;
Step 2, receives all non-sequential write requests, and when the most persistently, caching triggers lower brush when being written to water level, is wherein drawn by band
Being divided into data band and verification band, lower brush is carried out in units of group, and the write between group does not produces writes covering, non-whole piece band in group
Process of writing triggers small letter process, does incidentally reclaimer operation when little write-read simultaneously.
7. write ordered method based on watt record RAID incidentally reclaimed as claimed in claim 6, it is characterised in that each disk
On have identical reel number band composition one group.
8. as claimed in claim 6 write ordered method based on watt record RAID incidentally reclaimed, it is characterised in that described in follow
Ring daily record includes head position, tail position, after small letter runs through, by the write request write head position of small letter, and by head
Position, portion moves forward, and incidentally reclaims the number reading one section of continuous space from described tail position to described head position sequential scan
According to, scanning comprises invalid data block, extracts valid data block, and valid data block is write institute from the data block read
State head position, thus accomplish the random write falling in band and random write are converted into random write and sequential write.
9. write ordered method based on watt record RAID incidentally reclaimed as claimed in claim 8, it is characterised in that described head
Distance N is there is in position, portion with described tail position.
10. write ordered method based on watt record RAID incidentally reclaimed as claimed in claim 6, it is characterised in that when entering
During the non-sequential write request of row, by described data band with in described verification band and the corresponding data block of the data block that will write and
Check block reads, and the data block that will write described in combination calculates new check block, and by the described data block that will write
Interior with described new check block write band.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610726634.4A CN106293543B (en) | 2016-08-25 | 2016-08-25 | It is a kind of that ordered method and device is write based on watt record RAID incidentally recycled |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610726634.4A CN106293543B (en) | 2016-08-25 | 2016-08-25 | It is a kind of that ordered method and device is write based on watt record RAID incidentally recycled |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106293543A true CN106293543A (en) | 2017-01-04 |
CN106293543B CN106293543B (en) | 2019-03-12 |
Family
ID=57676109
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610726634.4A Active CN106293543B (en) | 2016-08-25 | 2016-08-25 | It is a kind of that ordered method and device is write based on watt record RAID incidentally recycled |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106293543B (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106970764A (en) * | 2017-02-24 | 2017-07-21 | 中国科学院计算技术研究所 | An a kind of watt record RAID based on band version writes ordered method and system |
CN108376053A (en) * | 2018-01-23 | 2018-08-07 | 中国科学院计算技术研究所 | A kind of newer watt of record in support original place writes ordered method and device |
CN109509487A (en) * | 2017-09-15 | 2019-03-22 | 株式会社东芝 | Hard disk device and its control method |
CN109582222A (en) * | 2018-10-31 | 2019-04-05 | 华中科技大学 | A kind of method for cleaning persistently cached in host perception watt recording disc |
CN109710184A (en) * | 2018-12-19 | 2019-05-03 | 中国人民解放军国防科技大学 | Hierarchical hybrid storage method and system for tile record disk perception |
CN110471615A (en) * | 2018-05-10 | 2019-11-19 | 杭州宏杉科技股份有限公司 | A kind of method for writing data and device |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101075211A (en) * | 2007-06-08 | 2007-11-21 | 马彩艳 | Flash memory management based on sector access |
CN102087586A (en) * | 2009-12-02 | 2011-06-08 | 成都市华为赛门铁克科技有限公司 | Data processing method and device |
CN102185929A (en) * | 2011-06-02 | 2011-09-14 | 杭州华三通信技术有限公司 | Storage area network (SAN)-resource-based video monitoring data storage method and device |
CN102662607A (en) * | 2012-03-29 | 2012-09-12 | 华中科技大学 | RAID6 level mixed disk array, and method for accelerating performance and improving reliability |
CN103049222A (en) * | 2012-12-28 | 2013-04-17 | 中国船舶重工集团公司第七0九研究所 | RAID5 (redundant array of independent disk 5) write IO optimization processing method |
CN103257932A (en) * | 2012-01-17 | 2013-08-21 | 国际商业机器公司 | Method and system for managing data in computer readable cache system |
CN104035729A (en) * | 2014-05-22 | 2014-09-10 | 中国科学院计算技术研究所 | Block device thin-provisioning method for log mapping |
CN104040508A (en) * | 2012-01-17 | 2014-09-10 | 国际商业机器公司 | Populating a first stride of tracks from a first cache to write to a second stride in a second cache |
CN104461373A (en) * | 2013-09-16 | 2015-03-25 | 杭州宏杉科技有限公司 | Method and device for processing data of RAID (redundant array of independent disks) |
CN104503706A (en) * | 2014-12-23 | 2015-04-08 | 中国科学院计算技术研究所 | Data storing method and data reading method based on disk array |
-
2016
- 2016-08-25 CN CN201610726634.4A patent/CN106293543B/en active Active
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101075211A (en) * | 2007-06-08 | 2007-11-21 | 马彩艳 | Flash memory management based on sector access |
CN102087586A (en) * | 2009-12-02 | 2011-06-08 | 成都市华为赛门铁克科技有限公司 | Data processing method and device |
CN102185929A (en) * | 2011-06-02 | 2011-09-14 | 杭州华三通信技术有限公司 | Storage area network (SAN)-resource-based video monitoring data storage method and device |
CN103257932A (en) * | 2012-01-17 | 2013-08-21 | 国际商业机器公司 | Method and system for managing data in computer readable cache system |
CN104040508A (en) * | 2012-01-17 | 2014-09-10 | 国际商业机器公司 | Populating a first stride of tracks from a first cache to write to a second stride in a second cache |
CN102662607A (en) * | 2012-03-29 | 2012-09-12 | 华中科技大学 | RAID6 level mixed disk array, and method for accelerating performance and improving reliability |
CN103049222A (en) * | 2012-12-28 | 2013-04-17 | 中国船舶重工集团公司第七0九研究所 | RAID5 (redundant array of independent disk 5) write IO optimization processing method |
CN104461373A (en) * | 2013-09-16 | 2015-03-25 | 杭州宏杉科技有限公司 | Method and device for processing data of RAID (redundant array of independent disks) |
CN104035729A (en) * | 2014-05-22 | 2014-09-10 | 中国科学院计算技术研究所 | Block device thin-provisioning method for log mapping |
CN104503706A (en) * | 2014-12-23 | 2015-04-08 | 中国科学院计算技术研究所 | Data storing method and data reading method based on disk array |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106970764A (en) * | 2017-02-24 | 2017-07-21 | 中国科学院计算技术研究所 | An a kind of watt record RAID based on band version writes ordered method and system |
CN106970764B (en) * | 2017-02-24 | 2019-08-16 | 中国科学院计算技术研究所 | It is a kind of based on band version watt record a RAID write ordered method and system |
CN109509487A (en) * | 2017-09-15 | 2019-03-22 | 株式会社东芝 | Hard disk device and its control method |
CN109509487B (en) * | 2017-09-15 | 2021-01-26 | 株式会社东芝 | Hard disk device and control method thereof |
CN108376053A (en) * | 2018-01-23 | 2018-08-07 | 中国科学院计算技术研究所 | A kind of newer watt of record in support original place writes ordered method and device |
CN108376053B (en) * | 2018-01-23 | 2020-02-28 | 中国科学院计算技术研究所 | Tile record writing sequencing method and device supporting in-place updating |
CN110471615A (en) * | 2018-05-10 | 2019-11-19 | 杭州宏杉科技股份有限公司 | A kind of method for writing data and device |
CN110471615B (en) * | 2018-05-10 | 2022-08-02 | 杭州宏杉科技股份有限公司 | Data writing method and device |
CN109582222A (en) * | 2018-10-31 | 2019-04-05 | 华中科技大学 | A kind of method for cleaning persistently cached in host perception watt recording disc |
CN109582222B (en) * | 2018-10-31 | 2020-11-24 | 华中科技大学 | Method for cleaning persistent cache in host sensing tile recording disk |
CN109710184A (en) * | 2018-12-19 | 2019-05-03 | 中国人民解放军国防科技大学 | Hierarchical hybrid storage method and system for tile record disk perception |
CN109710184B (en) * | 2018-12-19 | 2022-02-18 | 中国人民解放军国防科技大学 | Hierarchical hybrid storage method and system for tile record disk perception |
Also Published As
Publication number | Publication date |
---|---|
CN106293543B (en) | 2019-03-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106293543A (en) | A kind of based on incidentally reclaim watt record a RAID write ordered method and device | |
CN106970764B (en) | It is a kind of based on band version watt record a RAID write ordered method and system | |
CN103135940B (en) | Implementing enhanced fragmented stream handling in a shingled disk drive | |
US8681443B2 (en) | Shingle-written magnetic recording (SMR) device with hybrid E-region | |
CN104049908B (en) | Intermediate storage based on dynamic particle | |
CN108958656B (en) | Dynamic stripe system design method based on RAID5 solid state disk array | |
US5764880A (en) | Method and system for rebuilding log-structured arrays | |
CN103049222B (en) | A kind of RAID5 writes IO optimized treatment method | |
US20110161557A1 (en) | Distributed media cache for data storage systems | |
CN107037976A (en) | Storage device and its method of work | |
CN105955664B (en) | A kind of reading/writing method of watt record conversion layer based on segment structure | |
US20100070733A1 (en) | System and method of allocating memory locations | |
US8862856B2 (en) | Implementing remapping command with indirection update for indirected storage | |
CN1914687A (en) | Quick access to disc management information | |
US20190004739A1 (en) | Managing Shingled Magnetic Recording (SMR) Zones in a Hybrid Storage Device | |
CN106951375A (en) | The method and device of snapped volume is deleted within the storage system | |
CN103970669A (en) | Method for accelerating physical-to-logic address mapping of recycling operation in solid-state equipment | |
CN107145301B (en) | Data access method and device based on hybrid disk | |
CN106354658A (en) | Method for reducing memory resource occupation of mapping tables in hybrid mapping algorithm | |
CN113190180A (en) | Storage device based on mixed media and distributed storage system | |
CN108376053A (en) | A kind of newer watt of record in support original place writes ordered method and device | |
CN105824579B (en) | A method of incidentally read-write reduction watt record writes amplification expense | |
CN107728943B (en) | Method for delaying generation of check optical disc and corresponding data recovery method | |
CN107402890A (en) | A kind of data processing method and system based on Solid-state disc array and caching | |
CN104461930A (en) | Cache writing method and device |
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 |