CN101840377A - Data storage method based on RS (Reed-Solomon) erasure codes - Google Patents

Data storage method based on RS (Reed-Solomon) erasure codes Download PDF

Info

Publication number
CN101840377A
CN101840377A CN201010171777A CN201010171777A CN101840377A CN 101840377 A CN101840377 A CN 101840377A CN 201010171777 A CN201010171777 A CN 201010171777A CN 201010171777 A CN201010171777 A CN 201010171777A CN 101840377 A CN101840377 A CN 101840377A
Authority
CN
China
Prior art keywords
data
data block
block
rs
original
Prior art date
Application number
CN201010171777A
Other languages
Chinese (zh)
Inventor
郑燕飞
沈林峰
曹羽中
程正杰
陈克非
Original Assignee
上海交通大学
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 上海交通大学 filed Critical 上海交通大学
Priority to CN201010171777A priority Critical patent/CN101840377A/en
Publication of CN101840377A publication Critical patent/CN101840377A/en

Links

Abstract

The invention belongs to the technical field of information processing, relating to a data storage method based on RS (Reed-Solomon) erasure codes. The method comprises the following steps of: carrying out blocking processing and grouping processing on files to be stored; transmitting original data blocks of each data group to data nodes, and carrying out RS encoding by the data nodes so that each data group is increased with a plurality of redundant data blocks on the basis of the original data blocks; storing the encoded data blocks in the same data group on a plurality of machine racks, wherein the number of the data blocks in the same data group on the same machine rack does not exceed the number of the redundant data blocks; adopting the RS erasure codes to restore the data blocks when the data blocks in the stored files are damaged, and obtaining the restored data blocks. The invention saves a large amount of storage space on the basis of ensuring daily usage efficiency. The requirements for the reliability of data storage can be flexibly set according to the importance of the files. More abundant storage strategies can be provided by setting the number of backup files and encoding parameters, and the cost for data storage is greatly reduced.

Description

Date storage method based on the RS correcting and eleting codes

Technical field

What the present invention relates to is a kind of method of technical field of information processing, specifically is a kind of date storage method based on the RS correcting and eleting codes.

Background technology

The cloud storage is a kind of storage mode of serving cloud computing, refer to by functions such as cluster application, gridding technique or distributed file systems, a large amount of various dissimilar memory devices in the network are gathered collaborative work by application software, a system of data storage and Operational Visit function externally is provided jointly.

Along with the arrival of information age, the storage of magnanimity information and more and more higher to reliability requirement, carrying cost and high reliability are a pair of insoluble contradiction all the time.Can significantly improve reliability by increasing the backup number on the one hand, but memory space increases exponentially thereupon also; If want to save cost on the other hand, the backup number of restricting data is in case when the backup of a copy of it data block all damaged, whole file perhaps all can not use, and causes irremediable loss.

Through being retrieved, existing literature finds, at present basic all cloud memory technologies all are the HDFS (Hadoop Distributed File System, i.e. Hadoop distributed file system) that has adopted in a kind of similar Hadoop (a kind of open source software in the cloud computing) system.This technology mainly is that a data file is divided into big or small n piece according to the size of setting (as 64M), be with the raising reliability by full backup (for example the HDFS distributed file system among the Hadoop is 3 parts of identical copy: same each portion of frame different pieces of information node, 1 part of the back end on the other frame) again to each data block.But the shortcoming of this technology maximum is the comparison wasting space: a file will back up 2 times additional storage space, thereby has improved cost.

Summary of the invention

The objective of the invention is to overcome above shortcomings in the prior art, a kind of date storage method based on the RS correcting and eleting codes is provided.The present invention at first carries out piecemeal and grouping to all raw data of file to be stored, then each group data is carried out RS correcting and eleting codes code storage again, setting flexibly on the basis of reliability, has greatly conserve storage, to save the advantage of cost.

The present invention is achieved by the following technical solutions, the present invention includes following steps:

The first step is carried out piecemeal to file to be stored and is handled, thereby obtains N the original data block that size is identical, and these original data blocks are carried out packet transaction, obtains M data set.

Described piecemeal is handled, and is: file to be stored is divided into N original data block according to fixed size T, when last piece original data block during less than T, last piece original data block is filled redundancy, making its size just in time is T.

Described packet transaction is: K original data block as one group, organized thereby N original data block is divided into M, and wherein the number of the original data block of last original data set is less than or equal to K.

Second step, original data block block-by-block in each data set is sent to back end from user side, behind i data set end of transmission, the back end that receives these data set data sends the data that receive to a data node of management node appointment, this back end obtains all raw data block messages of this group data block, 1≤i≤M.

In the 3rd step, the back end that obtains all raw data block messages of i group data block carries out the RS coding to this data set, makes i data set from K iIndividual original data block becomes K iIndividual original data block and R iIndividual redundant data piece, 1≤i≤M, K iIt is the number of original data block in preceding i the data set of coding.

Described RS coding is successively the data of w bit in each original data block to be handled according to following formula, obtains should K iThe R of individual w bit raw data iThe redundant data of individual w bit makes up the redundant data of all w bits of j redundant data in order and to obtain j redundant data piece, 1≤j≤R i, concrete formula is:

FD=C,

Wherein: F is R i* K iVandermonde matrix, D is K in i the data set iThe K that individual w bit raw data is formed i* 1 matrix, C are R in i the data set in coding back iThe R that the redundant data of individual w bit is formed i* 1 matrix.

In the 4th step, on several frames with the data block store in the same data set behind the coding, and the data block of same data set is no more than R on the same frame iIndividual.

In the 5th step, when having data block to damage in the storage file, adopt RS correcting and eleting codes recover data blocks, the data block after being restored.

Described recover data blocks is:

1) number when the data block of damaging is less than or equal to R iThe time, obtain the original data block of w bit respectively by following formula, and then make up original data block that obtains damaging or the redundant data piece that obtains damaging by the RS coding;

A′D=E′,

Wherein: I is K i* K iUnit matrix, F is R i* K iVandermonde matrix, D is K in i the data set iThe K that individual w bit raw data piece is formed i* 1 matrix, C are R in i the data set in coding back iThe R that individual w bits of redundancy data piece is formed i* 1 matrix, A ' are the new matrixes behind the pairing row of removal corrupt data piece from A, and E ' is the new matrix behind the pairing row of removal corrupt data piece from E;

2) when the number of the data block of damaging greater than R iThe time, the data block of damage can't be recovered.

Compared with prior art, the invention has the beneficial effects as follows: saved a large amount of storage spaces, backed up 3 parts of just achieved reliabilities originally, only needed backup to get final product for 2 parts or 1 part now, thereby reduced data storage cost; The requirement of reliability can be provided with neatly according to file importance, by the setting of backup quantity and coding parameter, can provide abundant more storage policy; Comprise original all original data blocks in the data set behind the coding, can guarantee efficient used in everyday like this.

Embodiment

Below method of the present invention is further described: present embodiment is being to implement under the prerequisite with the technical solution of the present invention, provided detailed embodiment and concrete operating process, but protection scope of the present invention is not limited to following embodiment.

Embodiment

Present embodiment is used for the file of 180M is stored, and specifically may further comprise the steps:

The first step is carried out piecemeal to 180M file to be stored and is handled, and is divided into 6 equal-sized original data blocks according to fixed size 30M, and these 6 original data blocks are divided into two groups, and every group comprises 3 original data blocks.

Second step, original data block block-by-block in each data set is sent to back end from user side, behind i data set end of transmission, the back end that receives these data set data sends the data that receive to a data node of management node appointment, this back end obtains all raw data block messages of this group data block, 1≤i≤2.

In the 3rd step, the back end that obtains all raw data block messages of i group data block carries out (5,3) RS coding to this data set, makes i data set become 3 original data blocks and 2 redundant data pieces from 3 original data blocks.

RS coding described in the present embodiment is successively the data of 4 bits in each original data block to be handled according to following formula, obtain the redundant data of 24 bits of corresponding per 34 bit raw data, present embodiment obtains first redundant data of 1536 4 bits and second redundant data of 1536 4 bits altogether, first redundant data of 1536 4 bits is combined into the first redundant data piece in order, second redundant data of 1536 4 bits is combined into the second redundant data piece in order, promptly obtain the redundant data piece of two 30M respectively, concrete formula is:

FD=C, (formula one)

Wherein: F is 2 * 3 Vandermonde matrix, and 3 * 1 the matrix that D is that 34 bit raw data are formed in i the data set, C are 2 * 1 matrixes of the redundant data composition of 24 bits in i the data set in coding back.

In the present embodiment F = 111 123 .

In the 4th step, on several frames with the data block store in the same data set behind the coding, and the data block of same data set is no more than 2 on the same frame.

In the 5th step, when having data block to damage in the storage file, adopt RS correcting and eleting codes recover data blocks, the data block after being restored.

With first group of data set is example, and described recover data blocks is:

1) when the number of the data block of damaging is less than or equal to 2, obtain original data block by following formula, and then original data block that obtains damaging or the redundant data piece that obtains damaging by the RS coding;

A ' D=E ', (formula two)

Wherein: I is 2 * 2 unit matrix, F is 2 * 3 Vandermonde matrix, 3 * 1 the matrix that D is that 34 bit raw data are formed in the 1st data set of expression, 2 * 1 the matrix that C is that 24 bits of redundancy data are formed in the 1st data set behind the presentation code, A ' is the new matrix behind the pairing row of removal corrupt data piece from A, and E ' is the new matrix behind the pairing row of removal corrupt data piece from E.

Because matrix F is a Vandermonde matrix, appoint 2 row get matrix A or 1 provisional capital be linearity independently, so matrix A ' must be reversible, the value of D can utilize Gaussian elimination method to calculate from (formula two), and all like this original data blocks can utilize (formula two) to recover out successively according to each calculating 4 Bit datas.

2) when the number of the data block of damaging greater than 2 the time, the data block of damage can't be recovered.

Successfully recovered to be no more than arbitrarily in the same data set data block of 2 damages in this example, and saved storage space greatly: if having only a backup, (5 in the present embodiment then, 3) the RS coding will be saved 4/9 space (9 of three parts of needs of prior art backup), thereby save very big cost in extensive the use; And can make setting more flexible by backup parameter and coding parameter are set, reliable rank refinement more.

Claims (5)

1. the date storage method based on the RS correcting and eleting codes is characterized in that, may further comprise the steps:
The first step is carried out piecemeal to file to be stored and is handled, thereby obtains N the original data block that size is identical, and these original data blocks are carried out packet transaction, obtains M data set;
Second step, original data block block-by-block in each data set is sent to back end from user side, behind i data set end of transmission, the back end that receives these data set data sends the data that receive to a data node of management node appointment, this back end obtains all raw data block messages of this group data block, 1≤i≤M;
In the 3rd step, the back end that obtains all raw data block messages of i group data block carries out the RS coding to this data set, makes i data set from K iIndividual original data block becomes K iIndividual original data block and R iIndividual redundant data piece, K iIt is the number of original data block in preceding i the data set of coding;
In the 4th step, on several frames with the data block store in the same data set behind the coding, and the data block of same data set is no more than R on the same frame iIndividual;
In the 5th step, when having data block to damage in the storage file, adopt RS correcting and eleting codes recover data blocks, the data block after being restored.
2. the date storage method based on the RS correcting and eleting codes according to claim 1, it is characterized in that, piecemeal described in the first step is handled, be: file to be stored is divided into N original data block according to fixed size T, when last piece original data block during less than T, last piece original data block is filled redundancy, and making its size just in time is T.
3. the date storage method based on the RS correcting and eleting codes according to claim 1, it is characterized in that, packet transaction described in the first step, be: with K original data block as one group, thereby N original data block is divided into the M group, and wherein the number of the original data block of last original data set is less than or equal to K.
4. the date storage method based on the RS correcting and eleting codes according to claim 1 is characterized in that, the RS coding described in the 3rd step is successively the data of w bit in each original data block to be handled according to following formula, obtains should K iThe R of individual w bit raw data iThe redundant data of individual w bit makes up the redundant data of all w bits of j redundant data in order and to obtain j redundant data piece, 1≤j≤R i, concrete formula is:
FD=C,
Wherein: F is R i* K iVandermonde matrix, D is K in i the data set iThe K that individual w bit raw data is formed i* 1 matrix, C are R in i the data set in coding back iThe R that the redundant data of individual w bit is formed i* 1 matrix.
5. the date storage method based on the RS correcting and eleting codes according to claim 1 is characterized in that, the recover data blocks described in the 5th step is meant:
1) number when the data block of damaging is less than or equal to R iThe time, obtain the original data block of w bit respectively by following formula, and then make up original data block that obtains damaging or the redundant data piece that obtains damaging by the RS coding;
A′D=E′,
Wherein: I is K i* K iUnit matrix, F is R i* K iVandermonde matrix, D is K in i the data set iThe K that individual w bit raw data piece is formed i* 1 matrix, C are R in i the data set in coding back iThe R that individual w bits of redundancy data piece is formed i* 1 matrix, A ' are the new matrixes behind the pairing row of removal corrupt data piece from A, and E ' is the new matrix behind the pairing row of removal corrupt data piece from E;
2) when the number of the data block of damaging greater than R iThe time, the data block of damage can't be recovered.
CN201010171777A 2010-05-13 2010-05-13 Data storage method based on RS (Reed-Solomon) erasure codes CN101840377A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201010171777A CN101840377A (en) 2010-05-13 2010-05-13 Data storage method based on RS (Reed-Solomon) erasure codes

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201010171777A CN101840377A (en) 2010-05-13 2010-05-13 Data storage method based on RS (Reed-Solomon) erasure codes

Publications (1)

Publication Number Publication Date
CN101840377A true CN101840377A (en) 2010-09-22

Family

ID=42743757

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201010171777A CN101840377A (en) 2010-05-13 2010-05-13 Data storage method based on RS (Reed-Solomon) erasure codes

Country Status (1)

Country Link
CN (1) CN101840377A (en)

Cited By (24)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102364472A (en) * 2011-10-25 2012-02-29 中兴通讯股份有限公司 Data storage method and system
CN102520890A (en) * 2011-12-30 2012-06-27 北京天地云箱科技有限公司 RS (Reed-Solomon) - DRAID( D redundant array of independent disk) system based on GPUs (graphic processing units) and method for controlling data of memory devices
CN102737032A (en) * 2011-04-06 2012-10-17 腾讯科技(深圳)有限公司 Method and device for file recovery
CN102937967A (en) * 2012-10-11 2013-02-20 南京中兴新软件有限责任公司 Data redundancy realization method and device
CN102968356A (en) * 2011-12-30 2013-03-13 中国科学院上海光学精密机械研究所 Data processing method of cloud storage system
CN103561057A (en) * 2013-10-15 2014-02-05 深圳清华大学研究院 Data storage method based on distributed hash table and erasure codes
CN103605582A (en) * 2013-11-27 2014-02-26 华中科技大学 Erasure code storage and reconfiguration optimization method based on redirect-on-write
CN103645861A (en) * 2013-12-03 2014-03-19 华中科技大学 Method for reconstructing failure node in erasure code cluster
CN104052576A (en) * 2014-06-07 2014-09-17 华中科技大学 Data recovery method based on error correcting codes in cloud storage
CN104572344A (en) * 2013-10-29 2015-04-29 杭州海康威视系统技术有限公司 Backup Method and system for multi-cloud data
CN105051700A (en) * 2013-03-14 2015-11-11 苹果公司 Selection of redundant storage configuration based on available memory space
CN105357294A (en) * 2015-10-31 2016-02-24 成都华为技术有限公司 Method for data storage and cluster management node
CN105635252A (en) * 2015-12-23 2016-06-01 浪潮集团有限公司 Erasure code redundant backup strategy of Hadoop distributed file system (HDFS)
CN105677249A (en) * 2016-01-04 2016-06-15 浙江宇视科技有限公司 Data block partitioning method, device and system
CN105893169A (en) * 2016-03-31 2016-08-24 乐视控股(北京)有限公司 File storage method and system based on erasure codes
CN106302573A (en) * 2015-05-14 2017-01-04 杭州海康威视系统技术有限公司 A kind of method, system and device using erasure codes to process data
CN104035732B (en) * 2014-06-17 2017-01-18 中国人民解放军国防科学技术大学 Data placing method aiming at erasure codes
CN106527986A (en) * 2016-11-03 2017-03-22 北京百度网讯科技有限公司 Method and device for storing data
CN106603673A (en) * 2016-12-19 2017-04-26 上海交通大学 Fine-grained cloud storage scheduling method based on erasure codes
CN107357932A (en) * 2017-07-31 2017-11-17 云城(北京)数据科技有限公司 A kind of file memory method and device
CN108365956A (en) * 2018-02-12 2018-08-03 北京京东尚科信息技术有限公司 Method and apparatus for storing information
CN110389715A (en) * 2018-04-23 2019-10-29 杭州海康威视系统技术有限公司 Date storage method, storage server and cloud storage system
WO2020043203A1 (en) * 2018-08-31 2020-03-05 杭州海康威视系统技术有限公司 Erasure code-based data storage method, device, and electronic apparatus
US10644726B2 (en) 2013-10-18 2020-05-05 Universite De Nantes Method and apparatus for reconstructing a data block

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070162684A1 (en) * 2006-01-06 2007-07-12 Kalos Matthew J Apparatus and method to configure, format, and test, a data storage subsystem product
CN101448281A (en) * 2008-12-31 2009-06-03 宁波中科集成电路设计中心有限公司 Fault-tolerance method in wireless sensor network
CN101622604A (en) * 2007-09-26 2010-01-06 株式会社东芝 Semiconductor memory device and its control method

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070162684A1 (en) * 2006-01-06 2007-07-12 Kalos Matthew J Apparatus and method to configure, format, and test, a data storage subsystem product
CN101622604A (en) * 2007-09-26 2010-01-06 株式会社东芝 Semiconductor memory device and its control method
CN101448281A (en) * 2008-12-31 2009-06-03 宁波中科集成电路设计中心有限公司 Fault-tolerance method in wireless sensor network

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
杨照宏: "《分布式海量存储系统的可靠性和容错性研究》", 31 December 2007, article "分布式海量存储系统的可靠性和容错性研究" *

Cited By (35)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102737032A (en) * 2011-04-06 2012-10-17 腾讯科技(深圳)有限公司 Method and device for file recovery
CN102737032B (en) * 2011-04-06 2016-06-08 腾讯科技(深圳)有限公司 The method of file access pattern and device
CN102364472A (en) * 2011-10-25 2012-02-29 中兴通讯股份有限公司 Data storage method and system
CN102520890A (en) * 2011-12-30 2012-06-27 北京天地云箱科技有限公司 RS (Reed-Solomon) - DRAID( D redundant array of independent disk) system based on GPUs (graphic processing units) and method for controlling data of memory devices
CN102968356A (en) * 2011-12-30 2013-03-13 中国科学院上海光学精密机械研究所 Data processing method of cloud storage system
CN102520890B (en) * 2011-12-30 2014-10-22 北京天地云箱科技有限公司 RS (Reed-Solomon) - DRAID( D redundant array of independent disk) system based on GPUs (graphic processing units) and method for controlling data of memory devices
CN102937967B (en) * 2012-10-11 2018-02-27 南京中兴新软件有限责任公司 Data redundancy realization method and device
WO2014056381A1 (en) * 2012-10-11 2014-04-17 中兴通讯股份有限公司 Data redundancy implementation method and device
CN102937967A (en) * 2012-10-11 2013-02-20 南京中兴新软件有限责任公司 Data redundancy realization method and device
CN105051700A (en) * 2013-03-14 2015-11-11 苹果公司 Selection of redundant storage configuration based on available memory space
CN105051700B (en) * 2013-03-14 2017-10-20 苹果公司 Based on available memory space selection redundant storage configuration
CN103561057A (en) * 2013-10-15 2014-02-05 深圳清华大学研究院 Data storage method based on distributed hash table and erasure codes
US10644726B2 (en) 2013-10-18 2020-05-05 Universite De Nantes Method and apparatus for reconstructing a data block
CN104572344A (en) * 2013-10-29 2015-04-29 杭州海康威视系统技术有限公司 Backup Method and system for multi-cloud data
CN103605582B (en) * 2013-11-27 2017-01-25 华中科技大学 Erasure code storage and reconfiguration optimization method based on redirect-on-write
CN103605582A (en) * 2013-11-27 2014-02-26 华中科技大学 Erasure code storage and reconfiguration optimization method based on redirect-on-write
CN103645861B (en) * 2013-12-03 2016-04-13 华中科技大学 The reconstructing method of failure node in a kind of correcting and eleting codes cluster
CN103645861A (en) * 2013-12-03 2014-03-19 华中科技大学 Method for reconstructing failure node in erasure code cluster
CN104052576B (en) * 2014-06-07 2017-05-10 华中科技大学 Data recovery method based on error correcting codes in cloud storage
CN104052576A (en) * 2014-06-07 2014-09-17 华中科技大学 Data recovery method based on error correcting codes in cloud storage
CN104035732B (en) * 2014-06-17 2017-01-18 中国人民解放军国防科学技术大学 Data placing method aiming at erasure codes
CN106302573A (en) * 2015-05-14 2017-01-04 杭州海康威视系统技术有限公司 A kind of method, system and device using erasure codes to process data
CN105357294B (en) * 2015-10-31 2018-10-02 成都华为技术有限公司 A kind of method and cluster management node of storage data
CN105357294A (en) * 2015-10-31 2016-02-24 成都华为技术有限公司 Method for data storage and cluster management node
CN105635252B (en) * 2015-12-23 2019-06-25 浪潮集团有限公司 A kind of Hadoop distributed file system HDFS correcting and eleting codes redundancy backup method
CN105635252A (en) * 2015-12-23 2016-06-01 浪潮集团有限公司 Erasure code redundant backup strategy of Hadoop distributed file system (HDFS)
CN105677249B (en) * 2016-01-04 2019-01-15 浙江宇视科技有限公司 The division methods of data block, apparatus and system
CN105677249A (en) * 2016-01-04 2016-06-15 浙江宇视科技有限公司 Data block partitioning method, device and system
CN105893169A (en) * 2016-03-31 2016-08-24 乐视控股(北京)有限公司 File storage method and system based on erasure codes
CN106527986A (en) * 2016-11-03 2017-03-22 北京百度网讯科技有限公司 Method and device for storing data
CN106603673A (en) * 2016-12-19 2017-04-26 上海交通大学 Fine-grained cloud storage scheduling method based on erasure codes
CN107357932A (en) * 2017-07-31 2017-11-17 云城(北京)数据科技有限公司 A kind of file memory method and device
CN108365956A (en) * 2018-02-12 2018-08-03 北京京东尚科信息技术有限公司 Method and apparatus for storing information
CN110389715A (en) * 2018-04-23 2019-10-29 杭州海康威视系统技术有限公司 Date storage method, storage server and cloud storage system
WO2020043203A1 (en) * 2018-08-31 2020-03-05 杭州海康威视系统技术有限公司 Erasure code-based data storage method, device, and electronic apparatus

Similar Documents

Publication Publication Date Title
US10055451B2 (en) Object loss reporting in a data storage system
Xia et al. A Tale of Two Erasure Codes in {HDFS}
EP3101530B1 (en) Data distribution method, data storage method, related apparatus, and system
US9032149B1 (en) Storage system having cross node data redundancy and method and computer readable medium for same
US9983959B2 (en) Erasure coding of data within a group of storage units based on connection characteristics
Rashmi et al. Explicit construction of optimal exact regenerating codes for distributed storage
US10187088B2 (en) Cost-efficient repair for storage systems using progressive engagement
US7437652B2 (en) Correcting multiple block data loss in a storage array using a combination of a single diagonal parity group and multiple row parity groups
US9996413B2 (en) Ensuring data integrity on a dispersed storage grid
US10613933B2 (en) System and method for providing thin-provisioned block storage with multiple data protection classes
KR101270815B1 (en) In-place transformations with applications to encoding and decoding various classes of codes
US20170272100A1 (en) Distributed Storage System Data Management And Security
US20140101111A1 (en) Backup of data across network of devices
US8086937B2 (en) Method for erasure coding data across a plurality of data stores in a network
US8522073B2 (en) Distributed storage of recoverable data
Rashmi et al. A piggybacking design framework for read-and download-efficient distributed storage codes
JP4340078B2 (en) Method and apparatus for backing up a data storage device
US8219887B2 (en) Parallel Reed-Solomon RAID (RS-RAID) architecture, device, and method
CA2717549C (en) Dynamically quantifying and improving the reliability of distributed data storage systems
US7363532B2 (en) System and method for recovering from a drive failure in a storage array
Oggier et al. Coding techniques for repairability in networked distributed storage systems
Chang et al. Myriad: Cost-Effective Disaster Tolerance.
CN1230747C (en) Restoring large capacity memory to previous state responding to processing information
US20150142863A1 (en) System and methods for distributed data storage
CN102833040B (en) Decoding processing method, device and coding/decoding system

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C02 Deemed withdrawal of patent application after publication (patent law 2001)
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20100922