CN102082830A - Unstable network-oriented distributed file storage method based on quality perception - Google Patents
Unstable network-oriented distributed file storage method based on quality perception Download PDFInfo
- Publication number
- CN102082830A CN102082830A CN2011100205949A CN201110020594A CN102082830A CN 102082830 A CN102082830 A CN 102082830A CN 2011100205949 A CN2011100205949 A CN 2011100205949A CN 201110020594 A CN201110020594 A CN 201110020594A CN 102082830 A CN102082830 A CN 102082830A
- Authority
- CN
- China
- Prior art keywords
- node
- client
- volume
- file
- title
- 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
Images
Abstract
The invention discloses a distributed file storage method based on quality perception. In the method, a node contribution weight calculation formula and a node redundancy backup policy based on recovery volume generation and synthesis are introduced, and a stable cluster storage mode-oriented distributed file system Hadoop is improved to adapt to a large-scale unstable network. The method has the advantages that: the traditional distributed storage mode mirroring redundancy backup policy is improved, and the storage efficiency of the traditional distributed file storage system is improved; checkout and recovery calculation is introduced, and the novel file storage service system processes various redundancy blocks equally, so the parallel transmission rate is improved and the file macro transmission rate is greatly improved; and the network stability tolerance is improved and scattered idle storage and network bandwidth resources on the internet are fully utilized.
Description
Technical field
The present invention relates to the distributed document storage problem, relate in particular to a kind of distributed document storage means based on the quality perception towards unstable network.
Background technology
Hadoop[1] be the Distributed Calculation open source software that the Apache foundation is released, main sub-project comprises: MapReduce[2-3]---the calculating cluster of a large-scale distributed data processing software frame collection; HDFS---a distributed file system [4].Hadoop mainly supports exploitation by Yahoo company, and has confirmed its reliability service [5] under the stabilizing network environment through about 200 the stable node operation tests of Yahoo.International many renowned companies, as Facebook, Amazon and Last.fm company, all utilize that Hadoop stores, activity such as calculating and data mining.Hadoop has become one of the first-selected support programs of structure data center of enterprise [6-7].
Hadoop was designed to be fit to operate on the common hardware (commodity hardware) in its early stage of development, and its distributed file system HDFS can be deployed on the cheap machine.Yet, because at the beginning of the Hadoop design is to be deployed in each node of metastable group system, therefore, if directly Hadoop is deployed in the huge wide area Internet network of instability highly and node quantity, then there are shortcomings such as the memory space utilance is low, network dynamic adaptable difference, cause and to provide reliable and high-quality networked stores service to the user.
The feature of Internet network is compared to the difference of traditional distributed network:
(1) node quantity is huge, and than tens of scales to thousands of nodes of traditional distributed network, the Internet network can reach the scale of tens thousand of nodes easily;
(2) the node height is autonomous and uncontrollable, and node can enter or exit network easily;
(3) node in addition may be for some reason to network swindle, spurious information.
List of references:
[1]?http://hadoop.apache.org/,?2010.
[2]?J.Dean?and?S.Ghemawat.?Mapreduce:?a exible?data?processing?tool.?Commun.?ACM,?53(1):72–77,?2010.
[3]?J.Dean?and?S.Ghemawat.?Mapreduce:?simpli?ed?data?processing?on?large?clusters.?In?SDI’04:?Proceedings?of?the?6th?Symposium?on?Operating?Systems?Design?&?Implementation,?pages10–10,?2004.
[4]?Garhan?Attebury,?Andrew?Baranovski,?Ken?Bloom,?et?al.?Hadoop?Distributed?File?System?for?the?Grid//Proc.?of?2009?IEEE?Nuclear?Science?Symposium?Conference?Record,?NSS/MIC?2009.?Orlando,?FL,?United?states,?October?25-31?2009:1056-1061.
[5]?O.?O’Malley?and?A.C.?Murthy.?Winning?a?60?second?dash?with?a?yellow?elephant.?Technical?report,?Yahoo!,?2009.
[6]?Lana?Abadie,?Paolo?Badino,?Jean-Philippe?Baud,et?al.?Storage?Resource?Managers:?Recent?international?experience?on?requirements?and?multiple?co-operating?implementations//Proc.?of?24th?IEEE?Conference?on?Mass?Storage?Systems?and?Technologies,?MSST?2007,?San?Diego,
[7]?Cornelli?F,?Damiani?E,?Capitani?S?D.?Choosing?Reputable?Servents?in?a?P2P?Network//Proc.?of?the?11th?International?World?Wide?Web?Confererce.?Honolulu,?Hawaii,?USA,?2002:376-386。
Summary of the invention
The purpose of this invention is to provide a kind of distributed document storage means based on the quality perception towards unstable network.
The step based on the distributed document storage means of quality perception towards unstable network is as follows:
1) any client's node in the distributed system joins in the distributed system, and to title node bulletin client node disk residual capacity, available network bandwidth and processor performance index, the title node to client's node regularly the transmission performance test bag determine to enliven in the network transmission delay and the transmission speed of client's node;
2) the title node periodically calculates each according to aftermentioned node contribution weights computing formula and enlivens client's node contribution weights, and the title node detects all notice informations that enliven client's node and transmission history record, identification cheating and malicious act;
3) file to be stored is uploaded in any client's node request in the distributed system, the title node adopts aftermentioned recovery volume generation method to cut apart file to be stored, the file to be stored agllutination closes and enlivens client's node contribution weights and aftermentioned node selection strategy, determine the client's node and the redundant digit of each volume of carrying, the title node transmits corresponding file volume to the carrying client node of determining;
4) storage file that any client's node request is downloaded in the distributed system, the title node finds and enlivens the association volume that client's node is stored the storage file of request download, the client's node that enlivens of the association volume of the storage file that request is downloaded transmits related volume to the title node, after the title node is collected enough association volumes, adopt aftermentioned to recover the synthetic original of volume method, behind the success composite document, the title node is asked the storage file of download to the transmission of client's node.
Described step 2) the node contribution weights computing formula in is:
Wherein:
W n ---the node warp
nContribution weights after the data transfer task;
α---constant;
RB i ---this node
iThe blocks of files number that actual success is transmitted in the subtask;
TB i ---this node
iThe blocks of files number of plan storage in the subtask;
t i ---the
iSubtask finishes this node needed time of model data piece of transmission that the back statistics draws, the s/64M of unit;
C---constant;
Sp---current this node remaining space;
Sp a ---the remaining space of each node maximum in the current system;
Sp Min ---the remaining space of each node minimum in system during initialization;
Sp Max ---the remaining space of each node maximum in system during initialization.
Recovery volume generating algorithm in the described step 3) is:
Recover the generation of volume:
Wherein:
i---the input file reference numeral;
j---the output file reference numeral;
Input (i)---the
iIndividual input file is an original document;
Output (j)---the
jIndividual output file promptly recovers volume.
Recovery in the described step 4) intervolves into algorithm:
Supposing has original volume
A, B, C, D, EAnd recover to roll up
X, Y, Z, original volume
B, C, DLose, then utilize the formula of the file that remaining roll recovery goes out to lose as follows:
Wherein: matrix
Q---the matrix that existing volume is formed, promptly
A, E, X, Y, Z;
Matrix
M---recover Jacobian matrix, each row represent an original volume, places one 1 at corresponding line of that row of existing original volume representative, the particular location of going should with original volume at matrix
QIn the position consistent; In that delegation that each representative recovers to roll up, place
i (j-1)
The beneficial effect that the present invention compared with prior art has:
1, the advantage that the invention discloses based on the distributed document storage means of quality perception is: System Hardware Requirement is low, it is little to build cost.
Common distributed file system all is to operate in the stable cluster environment of high speed generally at present, and hardware configuration and network environment are had very high requirement.And distributed document storage means of the present invention acts on Internet Wide Area Network environment, the composition of client's node is not confined to group system of a high price, the PC of common, inexpensive also can become reliable and stable client's node, therefore system to build cost little, system resource is rich and easy to get.
2, the advantage that the invention discloses based on the distributed document storage means of quality perception is: file storage more fast, reliable, node is selected targetedly.
The present invention has introduced the node selection strategy based on the storage quality, and adding can be estimated the weights of node contribution, carries out node according to the weights size again and selects, and makes stores service more targeted.Comprised in the contribution weights of the node definition simultaneously the taking all factors into consideration of factors such as node bandwidth, memory space, line duration, system can store by the high node of preoption value.Therefore, this method can provide quicker, safe and reliable file storage service to the user.In the middle of the application of reality, select the strategy of node at random than the Hadoop system, the node storage policy of native system has significantly improved service rate and quality.
3, the advantage that the invention discloses based on the distributed document storage means of quality perception is: backup policy is safer, and it is fewer to take resource.
Backup policy of the present invention is based on recovering the volume technology, and with respect to other traditional redundant mirror back-up strategies, this strategy has remarkable advantages on the disk utilance.For the initial data volume of client file, the maximum number of the recovery that can produce volume is and the identical number of initial data volume that promptly the file redundancy is 100% to the maximum.This means, with respect to traditional redundancy backup strategy, for example present most popular HDFS file system, the file storage of native system has dwindled 1/3 at least, this is very important advantage for the main service object of distributed file system---super large file.
4, the advantage that the invention discloses based on the distributed document storage means of quality perception is: data maintenance is easier.
Because recover the characteristic of volume, system does not need the content of focused data piece, only need the quantity of service data volume, as long as the book number in the system reaches the required minimum requirements of recovery file, can guarantee the integrality of customer data.So aspect data maintenance, the carrying out that native system can be more prone to.
Description of drawings
The present invention will be further described below in conjunction with drawings and Examples.
Fig. 1 is that the credible distributed document stores service platform framework based on redundancy and quality is summarized under the dynamic network environment;
Fig. 2 is this distributed document stores service platform storage node contribution weights calculation flow chart;
Fig. 3 is the storage and the download service flow process of this distributed document stores service platform;
Fig. 4 is the node contribution weights change curve of node3 in 17 continuous storing process.
Embodiment
The step based on the distributed document storage means of quality perception towards unstable network is as follows:
1) any client's node in the distributed system joins in the distributed system, and to title node bulletin client node disk residual capacity, available network bandwidth and processor performance index, the title node to client's node regularly the transmission performance test bag determine to enliven in the network transmission delay and the transmission speed of client's node;
2) the title node periodically calculates each according to aftermentioned node contribution weights computing formula and enlivens client's node contribution weights, and the title node detects all notice informations that enliven client's node and transmission history record, identification cheating and malicious act;
3) file to be stored is uploaded in any client's node request in the distributed system, the title node adopts aftermentioned recovery volume generation method to cut apart file to be stored, the file to be stored agllutination closes and enlivens client's node contribution weights and aftermentioned node selection strategy, determine the client's node and the redundant digit of each volume of carrying, the title node transmits corresponding file volume to the carrying client node of determining;
4) storage file that any client's node request is downloaded in the distributed system, the title node finds and enlivens the association volume that client's node is stored the storage file of request download, the client's node that enlivens of the association volume of the storage file that request is downloaded transmits related volume to the title node, after the title node is collected enough association volumes, adopt aftermentioned to recover the synthetic original of volume method, behind the success composite document, the title node is asked the storage file of download to the transmission of client's node.
Described step 2) the node contribution weights computing formula in is:
Wherein:
W n ---the node warp
nContribution weights after the data transfer task;
α---constant;
RB i ---this node
iThe blocks of files number that actual success is transmitted in the subtask;
TB i ---this node
iThe blocks of files number of plan storage in the subtask;
t i ---the
iSubtask finishes this node needed time of model data piece of transmission that the back statistics draws, the s/64M of unit;
C---constant;
Sp---current this node remaining space;
Sp a ---the remaining space of each node maximum in the current system;
Sp Min ---the remaining space of each node minimum in system during initialization;
Sp Max ---the remaining space of each node maximum in system during initialization.
Recovery volume generating algorithm in the described step 3) is:
Recover the generation of volume:
Wherein:
i---the input file reference numeral;
j---the output file reference numeral;
Input (i)---the
iIndividual input file is an original document;
Output (j)---the
jIndividual output file promptly recovers volume.
Recovery in the described step 4) intervolves into algorithm:
Supposing has original volume
A, B, C, D, EAnd recover to roll up
X, Y, Z, original volume
B, C, DLose, then utilize the formula of the file that remaining roll recovery goes out to lose as follows:
Wherein: matrix
Q---the matrix that existing volume is formed, promptly
A, E, X, Y, Z;
Matrix
M---recover Jacobian matrix, each row represent an original volume, places one 1 at corresponding line of that row of existing original volume representative, the particular location of going should with original volume at matrix
QIn the position consistent; In that delegation that each representative recovers to roll up, place
i (j-1)
Embodiment:
1, recovering the volume generating algorithm is:
If file
A, with file
ABe divided into 5 equal volumes (not containing concluding volume), be respectively
A1, A2, A3, A4, A5
According to the generating algorithm of recovering volume, utilize above-mentioned 5 book to generate 3 and recover volume
R1, R2, R3Wherein:
2, recover to intervolve into algorithm and be:
If because of certain client's node off-line, cause configuration file in the system
ABook
A2, A3, A4Lose.At this moment, 2 initial data volumes have been remained in the system
A1, A5, three are recovered volume
R1, R2, R3, the volume sum is no less than configuration file
AInitial data volume sum.Therefore the book of losing can be recovered.
Obliterated data recovery process:, set up matrix according to recovering the roll recovery formula.
The dematrix equation gets:
The book that can obtain losing thus
A2, A3, A4
Utilize the legacy data volume
A1, A5, and recover the book of coming out
A2, A3, A4, go back original
A
3, node contribution weights computing formula is:
Setting up departments has 6 client's nodes in the system, and the weights of each client's node are 1 when system initialization, with one of them client's node n ode3 as investigating object.At first add up the remanence disk space of each node
Sp, be respectively { 9.71,15.3,19.75,19.75,14.91,11.23 }
GB, calculate
。
At this moment, another one client node will be uploaded one 1
GBFile.File uploads to the title node by Web Portal.The title node carries out bundling to file, obtains 16 file data volumes altogether, and utilizes these 16 file data volumes to generate 14 and recover volume, and 30 volumes are arranged altogether.Calculate the volume number that each client's node should carry, the book number that wherein should carry on the node3 is
Wherein:
Sum---total file volume number;
W3---the current node contribution weights of node3;
Wi---current node
iNode contribution weights.
This time after the end of file transfer, the title node scans every evaluation index of node3, and wherein node3 goes up successful data block transmitted number
RB=5, the unit data piece transmission time
T=4.496
s, node3 this moment the remanence disk space
Sp=19.43
GB, the weights that calculate node3 according to node contribution weights computing formula are:
The connection point manager (CP manager) database on the newname node more, each client's node contribution weights records following { 1.56,1.17,1.41,1.26,1.03,0.91 }.
If have client's node to upload onesize file this moment again, according to the node contribution weights after upgrading, node3 is last should data carried by data volume number to be:
After the end of file transfer, the title node is update system connection point manager (CP manager) database once more, repeats with this, periodically the storage quality of node3 is upgraded evaluation.Accompanying drawing 4 is the node contribution weights change curve of node3 in 17 continuous storing process.
Claims (4)
1. distributed document storage means based on the quality perception towards unstable network is characterized in that its step is as follows:
1) any client's node in the distributed system joins in the distributed system, and to title node bulletin client node disk residual capacity, available network bandwidth and processor performance index, the title node to client's node regularly the transmission performance test bag determine to enliven in the network transmission delay and the transmission speed of client's node;
2) the title node periodically calculates each according to aftermentioned node contribution weights computing formula and enlivens client's node contribution weights, and the title node detects all notice informations that enliven client's node and transmission history record, identification cheating and malicious act;
3) file to be stored is uploaded in any client's node request in the distributed system, the title node adopts aftermentioned recovery volume generation method to cut apart file to be stored, the file to be stored agllutination closes and enlivens client's node contribution weights and aftermentioned node selection strategy, determine the client's node and the redundant digit of each volume of carrying, the title node transmits corresponding file volume to the carrying client node of determining;
4) storage file that any client's node request is downloaded in the distributed system, the title node finds and enlivens the association volume that client's node is stored the storage file of request download, the client's node that enlivens of the association volume of the storage file that request is downloaded transmits related volume to the title node, after the title node is collected enough association volumes, adopt aftermentioned to recover the synthetic original of volume method, behind the success composite document, the title node is asked the storage file of download to the transmission of client's node.
2. a kind of distributed document storage means based on the quality perception towards unstable network according to claim 1 is characterized in that described step 2) in node contribution weights computing formula be:
Wherein:
W n ---the node warp
nContribution weights after the data transfer task;
α---constant;
RB i ---this node
iThe blocks of files number that actual success is transmitted in the subtask;
TB i ---this node
iThe blocks of files number of plan storage in the subtask;
t i ---the
iSubtask finishes this node needed time of model data piece of transmission that the back statistics draws, unit
S/64M
C---constant;
Sp---current this node remaining space;
Sp a ---the remaining space of each node maximum in the current system;
Sp Min ---the remaining space of each node minimum in system during initialization;
Sp Max ---the remaining space of each node maximum in system during initialization.
3. a kind of distributed document storage means based on the quality perception towards unstable network according to claim 1 is characterized in that the recovery volume generating algorithm in the described step 3) is:
Recover the generation of volume:
Wherein:
i---the input file reference numeral;
j---the output file reference numeral;
Input (i)---the
iIndividual input file is an original document;
Output (j)---the
jIndividual output file promptly recovers volume.
4. a kind of distributed document storage means based on the quality perception towards unstable network according to claim 1 is characterized in that recovery in the described step 4) intervolves into algorithm and is:
Supposing has original volume
A, B, C, D, EAnd recover to roll up
X, Y, Z, original volume
B, C, DLose, then utilize the formula of the file that remaining roll recovery goes out to lose as follows:
Wherein: matrix
Q---the matrix that existing volume is formed, promptly
A, E, X, Y, Z;
Matrix
M---recover Jacobian matrix, each row represent an original volume, places one 1 at corresponding line of that row of existing original volume representative, the particular location of going should with original volume at matrix
QIn the position consistent; In that delegation that each representative recovers to roll up, place
i (j-1)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201110020594 CN102082830B (en) | 2011-01-18 | 2011-01-18 | Unstable network-oriented distributed file storage method based on quality perception |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201110020594 CN102082830B (en) | 2011-01-18 | 2011-01-18 | Unstable network-oriented distributed file storage method based on quality perception |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102082830A true CN102082830A (en) | 2011-06-01 |
| CN102082830B CN102082830B (en) | 2013-03-06 |
Family
ID=44088579
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201110020594 Expired - Fee Related CN102082830B (en) | 2011-01-18 | 2011-01-18 | Unstable network-oriented distributed file storage method based on quality perception |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102082830B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102419766A (en) * | 2011-11-01 | 2012-04-18 | 西安电子科技大学 | Data redundancy and file operation methods based on Hadoop distributed file system (HDFS) |
| CN102521218A (en) * | 2011-12-15 | 2012-06-27 | 方正国际软件有限公司 | File combining method and file combining device |
| CN103186554A (en) * | 2011-12-28 | 2013-07-03 | 阿里巴巴集团控股有限公司 | Distributed data mirroring method and data storage node |
| CN103635882A (en) * | 2011-06-29 | 2014-03-12 | 微软公司 | Controlling network utilization |
| CN104142871A (en) * | 2013-05-10 | 2014-11-12 | 中国电信股份有限公司 | Data backup method and device and distributed file system |
| CN106294585A (en) * | 2016-07-28 | 2017-01-04 | 四川新环佳科技发展有限公司 | A kind of storage method under cloud computing platform |
| CN108366094A (en) * | 2018-01-16 | 2018-08-03 | 厦门集微科技有限公司 | A kind of cloud storage platform management method |
| CN108875035A (en) * | 2018-06-25 | 2018-11-23 | 郑州云海信息技术有限公司 | The date storage method and relevant device of distributed file system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101009567A (en) * | 2006-01-26 | 2007-08-01 | 西门子通信技术（北京）有限公司 | A method and system utilizing peer-to-peer network entity to provide the network service |
| CN101446976A (en) * | 2008-12-26 | 2009-06-03 | 中兴通讯股份有限公司 | File storage method in distributed file system |
| US20090232015A1 (en) * | 2008-02-01 | 2009-09-17 | Peter Domschitz | Traffic control device, packet-based network and method for controlling traffic in a packet-based network |
| CN101944124A (en) * | 2010-09-21 | 2011-01-12 | 卓望数码技术（深圳）有限公司 | Distributed file system management method, device and corresponding file system |
-
2011
- 2011-01-18 CN CN 201110020594 patent/CN102082830B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101009567A (en) * | 2006-01-26 | 2007-08-01 | 西门子通信技术（北京）有限公司 | A method and system utilizing peer-to-peer network entity to provide the network service |
| US20090232015A1 (en) * | 2008-02-01 | 2009-09-17 | Peter Domschitz | Traffic control device, packet-based network and method for controlling traffic in a packet-based network |
| CN101446976A (en) * | 2008-12-26 | 2009-06-03 | 中兴通讯股份有限公司 | File storage method in distributed file system |
| CN101944124A (en) * | 2010-09-21 | 2011-01-12 | 卓望数码技术（深圳）有限公司 | Distributed file system management method, device and corresponding file system |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10013281B2 (en) | 2011-06-29 | 2018-07-03 | Microsoft Technology Licensing, Llc | Controlling network utilization |
| CN103635882A (en) * | 2011-06-29 | 2014-03-12 | 微软公司 | Controlling network utilization |
| CN102419766A (en) * | 2011-11-01 | 2012-04-18 | 西安电子科技大学 | Data redundancy and file operation methods based on Hadoop distributed file system (HDFS) |
| CN102521218B (en) * | 2011-12-15 | 2014-07-23 | 方正国际软件有限公司 | File combining method and file combining device |
| CN102521218A (en) * | 2011-12-15 | 2012-06-27 | 方正国际软件有限公司 | File combining method and file combining device |
| CN103186554B (en) * | 2011-12-28 | 2016-11-23 | 阿里巴巴集团控股有限公司 | Distributed data mirror method and storage back end |
| CN103186554A (en) * | 2011-12-28 | 2013-07-03 | 阿里巴巴集团控股有限公司 | Distributed data mirroring method and data storage node |
| CN104142871B (en) * | 2013-05-10 | 2017-05-24 | 中国电信股份有限公司 | Data backup method and device and distributed file system |
| CN104142871A (en) * | 2013-05-10 | 2014-11-12 | 中国电信股份有限公司 | Data backup method and device and distributed file system |
| CN106294585A (en) * | 2016-07-28 | 2017-01-04 | 四川新环佳科技发展有限公司 | A kind of storage method under cloud computing platform |
| CN106294585B (en) * | 2016-07-28 | 2019-10-18 | 上海倍增智能科技有限公司 | A kind of storage method under cloud computing platform |
| CN108366094A (en) * | 2018-01-16 | 2018-08-03 | 厦门集微科技有限公司 | A kind of cloud storage platform management method |
| CN108366094B (en) * | 2018-01-16 | 2021-05-07 | 厦门集微科技有限公司 | Cloud storage platform management method |
| CN108875035A (en) * | 2018-06-25 | 2018-11-23 | 郑州云海信息技术有限公司 | The date storage method and relevant device of distributed file system |
| CN108875035B (en) * | 2018-06-25 | 2022-02-18 | 郑州云海信息技术有限公司 | Data storage method of distributed file system and related equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102082830B (en) | 2013-03-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102082830B (en) | Unstable network-oriented distributed file storage method based on quality perception | |
| US10346431B1 (en) | System and method for automated run-tme scaling of cloud-based data store | |
| TWI549080B (en) | The method, system and device for sending information of category information | |
| Joshi et al. | Coding for fast content download | |
| KR102082355B1 (en) | Processing Techniques for Large Network Data | |
| CN104378423B (en) | Metadata cluster distributed memory system and reading, the method for write-in storage data | |
| JP6126099B2 (en) | Marketplace for timely event data distribution | |
| CN102055650B (en) | Load balance method and system and management server | |
| CN106815254B (en) | Data processing method and device | |
| CN107193909A (en) | Data processing method and system | |
| CN111459986B (en) | Data computing system and method | |
| CN103067297B (en) | A kind of dynamic load balancing method based on resource consumption prediction and device | |
| CN105138615A (en) | Method and system for building big data distributed log | |
| CN103761309A (en) | Operation data processing method and system | |
| JP2014531092A (en) | Distributed data stream processing method and system | |
| CN102360320A (en) | Terminal backup object sharing and recovery method based on cloud architecture | |
| CN104539681A (en) | Distributed GIS accelerating system and GIS service processing method | |
| CN101594256A (en) | Disaster recovery method, device and system | |
| CN102413185A (en) | Cloud platform automatic updating system based on BT technology | |
| CN106060173A (en) | Cluster type photographing data storage system based on cloud storage | |
| CN102508735A (en) | Method for backing up and restoring terminal system based on cloud architecture | |
| CN109981416A (en) | A kind of block chain method for testing performance and system | |
| CN102387179A (en) | Distributed file system and nodes, saving method and saving control method thereof | |
| CN102495772A (en) | Characteristic-based terminal program cloud backup and recovery methods | |
| CN103281400A (en) | Data segmenting, coding and recovering method used for cloud storage gateway |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20130306 Termination date: 20140118 |