CN103729149A - Data storage method - Google Patents
Data storage method Download PDFInfo
- Publication number
- CN103729149A CN103729149A CN201310750346.9A CN201310750346A CN103729149A CN 103729149 A CN103729149 A CN 103729149A CN 201310750346 A CN201310750346 A CN 201310750346A CN 103729149 A CN103729149 A CN 103729149A
- Authority
- CN
- China
- Prior art keywords
- data
- ssd
- writing
- random
- write
- 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
Images
Landscapes
- Signal Processing For Digital Recording And Reproducing (AREA)
Abstract
The invention relates to a data storage method which includes the steps of merging and storing data written at random in an SSD; writing the data written in the SSD into an array RAID set. After the embodiment is applied, the data written at random are recorded through the SSD, the data are then written into the array RAID set at a time, and the reading-writing performance of an RAID5 is further guaranteed.
Description
Technical field
The application relates to field of computer technology, more specifically, relates to a kind of method of storing data.
Background technology
Array (RAID) the 5th, the storage scheme that a kind of memory property, data security and carrying cost are taken into account.It is the half-way house of RAID0 and RAID1 that RAID5 can be understood as.RAID5 can provide data security guarantee for system, but degree of protection is than mirror image (Mirror) is low, disk space utilization factor is higher than Mirror.RAID5 has the akin data reading speed with RAID0, just many parity informations, and the speed of data writing is slightly slower than single disk is carried out to write operation.Due to the corresponding parity information of a plurality of data, the disk space utilization factor of RAID5 is higher than RAID1 simultaneously, and carrying cost is relatively low.
At least use 3 hard disks to set up RAID5 disk array.When having data to write hard disk, according to the mode of 1 hard disk, be exactly the magnetic track of this piece hard disk of writing direct.If RAID5, data write according to algorithm and are divided into 3 parts, then write respectively 3 hard disks, when writing, also can on these 3 hard disks, write check information, reading out data content from 3 hard disks respectively when reading data writing, then carry out verification by checking information.When wherein having 1 hard disk to occur damaging, the data that can store from other 2 hard disks calculate the data content of the 3rd hard disk.That is to say that this storage mode of raid5 has only allowed a hard disk to break down, and need to change while breaking down as early as possible.After changing faulty hard disk, the data that write between age at failure can be carried out verification again.If go bad again 1 hard disk in unresolved fault, that is exactly catastrophic fault.
Utilize RAID5 storage data when running into when writing at random, because will make to read to revise to write (RMW) or read to rebuild, write (RCW) and operate, aforesaid operations has had a strong impact on the readwrite performance of RAID5.
Summary of the invention
The embodiment of the present invention proposes a kind of method of storing data, the data of utilizing solid state hard disc record to write at random, the readwrite performance of the then hard disk in one-time write RAID5, and then assurance RAID5.
The technical scheme of the embodiment of the present invention is as follows:
Store a method for data, described method comprises:
Storage after the data of writing are at random merged in solid-state hard disk SSD;
The data that are stored in SSD are write in array RAID group.
Describedly before being merged in SSD, the data of writing at random further comprise:
SSD is divided into meta-data region and data field, and meta-data region records the information of the data of writing at random in SSD, the data that data area stores is write at random.
Described method further comprises: data field is cut apart at least one data block to the data block that sign has been used.
Described data field is cut apart and comprised at least one data block: data field is divided into at least one data block according to the size of band in RAID5.
The information of the data of writing at random in described SSD comprises: the size of the data write at random of storage, the data of writing are at random memory location and data memory location in RAID group of writing at random in SSD.
Described method further comprises: when RAID group is write data, while writing less than band, by described, less than strip data, be cached in SSD.
Describedly by being stored in data in SSD, writing RAID5 group and comprise:
When band meets full strip data, the data in described band are write in RAID group, remove the data in band described in SSD.
Describedly by being stored in data in SSD, writing RAID group and comprise:
In default time threshold, be stored in the data of writing at random in SSD and do not revise, the described data of writing are at random write in RAID group.
Describedly by being stored in data in SSD, writing RAID group and comprise:
To only need RCW operation, and the data that are stored in the band of data volume maximum in SSD write in RAID group.
Describedly by being stored in data in SSD, writing RAID group and comprise:
When the data volume of RMW operation is identical, will in disk, there be the data of a plurality of continuous strips position to write in RAID group.
From technique scheme, can find out, will need the data of RMW or RCW in embodiments of the present invention, storage after merging in SSD; The data that are stored in SSD are write in RAID group.The data of utilizing solid state hard disc record to write at random, then write the hard disk in RAID group, and then guarantee the readwrite performance of RAID5.
Accompanying drawing explanation
Fig. 1 is the method flow schematic diagram of storage data.
Embodiment
For making the object, technical solutions and advantages of the present invention express clearlyer, below in conjunction with drawings and the specific embodiments, the present invention is further described in more detail.
In embodiments of the present invention, by the data of writing at random, storage after merging in SSD; The data that are stored in SSD are write in RAID group.The data of utilizing solid state hard disc record to write at random, then, in one-time write RAID group, guarantee the readwrite performance of RAID5.
In addition, when data meet certain condition in being stored in SSD, be just stored in RAID group, if do not meet this condition, be not temporarily stored in SSD, thereby avoided repeatedly reading the data in RAID group.And then fully guaranteed the readwrite performance of RAID5.
Below in conjunction with accompanying drawing 1, describe technical scheme of the present invention in detail, specifically comprise the following steps:
101, by the data of writing at random, storage after merging in SSD.
SSD is the hard disk made from solid-state electronic storage chip array, control module and storage unit, consists of.The advantages such as SSD has fast, the shockproof crash resistance of read or write speed, low-power consumption, noiselessness, operating temperature range is large and light.Consider the mechanical hard disk adopting in RAID5, for fear of the readwrite performance that affects RAID5, can will need the data of RMW or RCW, in SSD, merge in the disk that stores again RADI group into.
SSD has the high performance feature of random read-write, and the data of writing at random that can S set SD some be affected to performance are first write SSD the inside and done and merge, and then these data are write in RAID group.To write at random data and merge in SSD, then when full band, write again in RAID group, do like this with respect to prior art, greatly reduce RMW and RCW operation, improve the readwrite performance of RAID5.
First, SSD is divided into meta-data region and data field, and meta-data region records the information of data in SSD, the position that in SSD, the size of the information of data storage data and this information are stored.For example, the size of storage data is 1M, is stored in the first data block.
The data that data area stores is write at random, the data when data of writing at random in the present invention also comprise data small letter.Data field is cut apart at least one data block to the data block of having used by bit map (bitmap) table sign, the corresponding whole SSD in all bit position data field in whole bitmap.Data block size can be arranged to RAID5 in band in the same size, the object of doing is like this to facilitate data conversion storage to RAID5, avoids splitting data block.
In addition, when RAID group is write data, while writing less than band, will be cached in SSD less than strip data.Data during wherein, less than band belong to writes data at random.
So, meta-data region further record in RAID group the data write at random in SSD in each member's dish in SSD memory location, the data write are at random in the size of the memory location of SSD and the data write at random.For example, each data to be written needs metadata need to record following information:
Disk_numb: the position of writing is at random at which disk;
Stripe_pos: the position of writing is at random which pillar location in RAID5;
Data_offset: the position of writing at random deviation post in band;
Data_size: the position size of the data of writing at random;
Block_numb: the data of writing at random exist in which data block of SSD.
By metadata, can check that each disk has the position of the data of writing at random.According to the strip_pos position of data, also can check the data that have been buffered in SSD, which data is in RAID5 in same band.
102, the data that are stored in SSD are write in RAID group.
In SSD, all data are all to take the data block of stripe size to exist in SSD as storage cell.The size of a data block is identical with the size of band in RAID5.Band described below is data block in SSD.
When band meets full strip data, the data in this band are write in RAID group, remove the data in band described in SSD.
In addition, all right setup times threshold value if some data of writing at random writes after SSD, was not revised in time threshold again, and the data of this being write at random write in RAID group.
When if client's business is not in a hurry, select the relatively full strip data that RCW just can complete that only needs that in SSD, in a band, data are filled out preferentially to write in RAID group.Due to a situation that band is fuller, RCW operation is relatively less.
In SSD data, there is the band of a plurality of diverse locations, when needing the data volume of RMW operation identical, those have the continuous strip data of a plurality of pillar locations priority processing in some or certain disk.Because a plurality of pillar locations are continuous, it is also continuous corresponding to each data in magnetic disk, better performances when correspondingly data are are sequentially read and write.
The above, be only preferred embodiment of the present invention, is not intended to limit protection scope of the present invention.Within the spirit and principles in the present invention all, any modification of doing, be equal to replacement, improvement etc., within all should being included in protection scope of the present invention.
Claims (10)
1. a method of storing data, is characterized in that, described method comprises:
Storage after the data of writing are at random merged in solid-state hard disk SSD;
The data that are stored in SSD are write in array RAID group.
2. according to the method for claim 1 storage data, it is characterized in that, describedly further comprise before the data of writing are at random merged in SSD:
SSD is divided into meta-data region and data field, and meta-data region records the information of the data of writing at random in SSD, the data that data area stores is write at random.
3. according to the method for claim 2 storage data, it is characterized in that, described method further comprises: data field is cut apart at least one data block to the data block that sign has been used.
4. according to the method for claim 3 storage data, it is characterized in that, described data field is cut apart and comprised at least one data block: data field is divided into at least one data block according to the size of band in RAID5.
5. according to the method for claim 2 storage data, it is characterized in that, the information of the data of writing at random in described SSD comprises: the size of the data that storage is write at random, the data of writing are at random memory location and data memory location in RAID group of writing at random in SSD.
6. according to the method for claim 2 storage data, it is characterized in that, described method further comprises: when RAID group is write data, while writing less than band, by described, less than strip data, be cached in SSD.
7. according to the method for claim 6 storage data, it is characterized in that, describedly by being stored in data in SSD, writing RAID5 group and comprise:
When band meets full strip data, the data in described band are write in RAID group, remove the data in band described in SSD.
8. according to the method for claim 1 storage data, it is characterized in that, describedly by being stored in data in SSD, writing RAID group and comprise:
In default time threshold, be stored in the data of writing at random in SSD and do not revise, the described data of writing are at random write in RAID group.
9. according to the method for claim 1 storage data, it is characterized in that, describedly by being stored in data in SSD, writing RAID group and comprise:
To only need RCW operation, and the data that are stored in the band of data volume maximum in SSD write in RAID group.
10. according to the method for claim 1 storage data, it is characterized in that, describedly by being stored in data in SSD, writing RAID group and comprise:
When the data volume of RMW operation is identical, will in disk, there be the data of a plurality of continuous strips position to write in RAID group.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310750346.9A CN103729149A (en) | 2013-12-31 | 2013-12-31 | Data storage method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310750346.9A CN103729149A (en) | 2013-12-31 | 2013-12-31 | Data storage method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN103729149A true CN103729149A (en) | 2014-04-16 |
Family
ID=50453240
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310750346.9A Pending CN103729149A (en) | 2013-12-31 | 2013-12-31 | Data storage method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103729149A (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104503710A (en) * | 2015-01-23 | 2015-04-08 | 福州瑞芯微电子有限公司 | Method and device for increasing writing speed of nand flash |
CN105573668A (en) * | 2015-12-11 | 2016-05-11 | 浪潮(北京)电子信息产业有限公司 | Data storage method and apparatus |
WO2016141785A1 (en) * | 2015-03-11 | 2016-09-15 | 中兴通讯股份有限公司 | Processing method and apparatus for high-performance redundant array of inexpensive disks (raid) mechanism |
CN106648469A (en) * | 2016-12-29 | 2017-05-10 | 华为技术有限公司 | Method and device for processing cache data and storage controller |
CN107832018A (en) * | 2017-11-22 | 2018-03-23 | 深圳忆联信息系统有限公司 | A kind of RAID implementation and SSD |
CN107832007A (en) * | 2017-10-25 | 2018-03-23 | 记忆科技(深圳)有限公司 | A kind of method of raising SSD combination properties |
CN108874594A (en) * | 2018-06-25 | 2018-11-23 | 郑州云海信息技术有限公司 | Disk replacing options, device, equipment and storage medium in a kind of storage cluster |
CN111399765A (en) * | 2019-12-31 | 2020-07-10 | 杭州海康威视系统技术有限公司 | Data processing method and device, electronic equipment and readable storage medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101354636A (en) * | 2008-09-08 | 2009-01-28 | 创新科存储技术(深圳)有限公司 | Method and system for writing data into magnetic disk array |
CN101727299A (en) * | 2010-02-08 | 2010-06-09 | 北京同有飞骥科技有限公司 | RAID5-orientated optimal design method for writing operation in continuous data storage |
CN103049221A (en) * | 2012-12-19 | 2013-04-17 | 创新科存储技术有限公司 | Method and device for processing disk array cache memory flash |
US20130282974A1 (en) * | 2012-04-23 | 2013-10-24 | Pramod G. Joisha | Merging arrays using shiftable memory |
-
2013
- 2013-12-31 CN CN201310750346.9A patent/CN103729149A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101354636A (en) * | 2008-09-08 | 2009-01-28 | 创新科存储技术(深圳)有限公司 | Method and system for writing data into magnetic disk array |
CN101727299A (en) * | 2010-02-08 | 2010-06-09 | 北京同有飞骥科技有限公司 | RAID5-orientated optimal design method for writing operation in continuous data storage |
US20130282974A1 (en) * | 2012-04-23 | 2013-10-24 | Pramod G. Joisha | Merging arrays using shiftable memory |
CN103049221A (en) * | 2012-12-19 | 2013-04-17 | 创新科存储技术有限公司 | Method and device for processing disk array cache memory flash |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104503710B (en) * | 2015-01-23 | 2018-02-13 | 福州瑞芯微电子股份有限公司 | A kind of method and apparatus for improving flash memory writing speed |
CN104503710A (en) * | 2015-01-23 | 2015-04-08 | 福州瑞芯微电子有限公司 | Method and device for increasing writing speed of nand flash |
WO2016141785A1 (en) * | 2015-03-11 | 2016-09-15 | 中兴通讯股份有限公司 | Processing method and apparatus for high-performance redundant array of inexpensive disks (raid) mechanism |
CN105573668B (en) * | 2015-12-11 | 2018-10-12 | 浪潮(北京)电子信息产业有限公司 | A kind of date storage method and device |
CN105573668A (en) * | 2015-12-11 | 2016-05-11 | 浪潮(北京)电子信息产业有限公司 | Data storage method and apparatus |
CN106648469A (en) * | 2016-12-29 | 2017-05-10 | 华为技术有限公司 | Method and device for processing cache data and storage controller |
CN106648469B (en) * | 2016-12-29 | 2020-01-17 | 华为技术有限公司 | Cache data processing method and device and storage controller |
WO2018121455A1 (en) * | 2016-12-29 | 2018-07-05 | 华为技术有限公司 | Cached-data processing method and device, and storage controller |
CN107832007A (en) * | 2017-10-25 | 2018-03-23 | 记忆科技(深圳)有限公司 | A kind of method of raising SSD combination properties |
CN107832018A (en) * | 2017-11-22 | 2018-03-23 | 深圳忆联信息系统有限公司 | A kind of RAID implementation and SSD |
CN108874594A (en) * | 2018-06-25 | 2018-11-23 | 郑州云海信息技术有限公司 | Disk replacing options, device, equipment and storage medium in a kind of storage cluster |
CN108874594B (en) * | 2018-06-25 | 2021-10-22 | 郑州云海信息技术有限公司 | Disk replacement method, device, equipment and storage medium in storage cluster |
CN111399765A (en) * | 2019-12-31 | 2020-07-10 | 杭州海康威视系统技术有限公司 | Data processing method and device, electronic equipment and readable storage medium |
CN111399765B (en) * | 2019-12-31 | 2022-11-01 | 杭州海康威视系统技术有限公司 | Data processing method and device, electronic equipment and readable storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103729149A (en) | Data storage method | |
US10445231B2 (en) | Method and system for storage device metadata management and processing | |
US10127166B2 (en) | Data storage controller with multiple pipelines | |
CN102122235B (en) | RAID4 (redundant array of independent disks) system and data reading and writing method thereof | |
CN106445405B (en) | Data access method and device for flash memory storage | |
CN104035830A (en) | Method and device for recovering data | |
CN104679442A (en) | Method and device for improving performance of disk array | |
US9875030B2 (en) | Media write operation | |
CN101526885A (en) | Method for improving capability of redundant array of independent disk and controller therefor | |
US8856439B2 (en) | Method and device for utilizing application-level prior knowledge for selectively storing data in higher performance media | |
CN103019623A (en) | Memory disc processing method and device | |
US11093339B2 (en) | Storage utilizing a distributed cache chain and a checkpoint drive in response to a data drive corruption | |
CN104035886B (en) | Disk remaps method, apparatus and electronic equipment | |
US20090027796A1 (en) | Information recording device and control method therefor | |
CN102262657B (en) | Method and system for storing multimedia data | |
CN105786721A (en) | Memory address mapping management method and processor | |
US10901866B2 (en) | Failure detection and data recovery in a storage system | |
CN103399783A (en) | Storage method and device of mirror image documents of virtual machines | |
CN117149062A (en) | Processing method and computing device for damaged data of magnetic tape | |
CN102362263A (en) | SSD controller, and method for operating an SSD controller | |
CN103645995A (en) | Data writing method and device | |
CN113703664B (en) | Random writing rate optimization implementation method for eMMC chip | |
CN109521961B (en) | Method and system for improving read-write performance of solid state disk | |
CN103064803B (en) | A kind of data read-write method of NAND Flash storage device and device | |
CN101149946B (en) | Method for treating stream media data |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140416 |
|
RJ01 | Rejection of invention patent application after publication |