CN1831810A - Method and middleware for data base duplicating using middleware in group communication system - Google Patents

Method and middleware for data base duplicating using middleware in group communication system Download PDF

Info

Publication number
CN1831810A
CN1831810A CN 200510055215 CN200510055215A CN1831810A CN 1831810 A CN1831810 A CN 1831810A CN 200510055215 CN200510055215 CN 200510055215 CN 200510055215 A CN200510055215 A CN 200510055215A CN 1831810 A CN1831810 A CN 1831810A
Authority
CN
China
Prior art keywords
affairs
database
middleware
execution
database storing
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
Application number
CN 200510055215
Other languages
Chinese (zh)
Other versions
CN100377156C (en
Inventor
王勇
焦丽梅
王启荣
李小华
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huawei Technologies Co Ltd
Original Assignee
Institute of Computing Technology of CAS
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 Institute of Computing Technology of CAS filed Critical Institute of Computing Technology of CAS
Priority to CNB2005100552154A priority Critical patent/CN100377156C/en
Publication of CN1831810A publication Critical patent/CN1831810A/en
Application granted granted Critical
Publication of CN100377156C publication Critical patent/CN100377156C/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Abstract

A method of utilizing intermediate component to carry out copy of databank in group communication system includes steps of local transaction execution, transaction broadcast, transaction sequence arrangement, sequence disorder judgment, sequence - set transaction execution and databank back - write. The said intermediate component consists of modules corresponding to steps in said method.

Description

Utilize middleware to carry out the method and the middleware of database replication in the group communication system
Technical field
The present invention relates to Database Systems, specifically, the present invention relates to the method and system that utilizes the middleware fulfillment database to duplicate.
Background technology
Duplicating is to copy one group of data the technology of a plurality of data sources to from a data source, is a piece of data is published to effective means on a plurality of memory nodes.Database replication is that a piece of data is published on the multiple servers, thereby makes different server users can share this piece of data in the scope of authority permission.Particularly, preserve identical database on each memory node, utilize reproduction technology can guarantee that the database that is distributed in different memory nodes upgrades automatically synchronously, thereby guarantee the consistance of data for full coping database.
Utilizing group communication system to carry out that sync database duplicates is a kind of new technology of database replication, and research is at home and abroad all arranged.Group communication system is meant the notion according to the Lamport logical timer, on the basis of false synchronization (Virtual Synchrony) model or expansion false synchronization model (Extended Virtual Synchrony), realize a communication middle layer of fault-tolerant processing under the distributed computing environment.Group communication system is safeguarded a process group, and various multi-casting communication modes are provided, and the function that data security transmits and group membership management is served is arranged.Application program uses it can realize fault-tolerant under the distributed environment easily, the duplicating and the dynamic management of system of data and key service.Group communication system mainly comprises following function: member management, be used for detection node adding, withdraw from; Various communication modes, as FIFO: guarantee the message from same node, the order that the take over party receives and the order of transmission are consistent; TOTAL-ORDER: a plurality of nodes when broadcasting simultaneously in group guarantee that the order that these message arrive all nodes is consistent; SAFE: when division took place network, the EVS model was followed in the reception of message.
Utilize group communication system to carry out the implementation that sync database duplicates and mainly contain two kinds, be respectively: revise DBMS (data base management system (DBMS)) mode and middleware mode.The former refers to revise DBMS, increases replication protocol, supports internally to duplicate; Middleware among the latter is meant one deck system software that is increased between traditional client and server, it can be integrated into a logic integral body with the resource and the service of the server of a plurality of isomorphisms or isomery, thereby provide a unified service interface to the client, simplify user's operation, the constructs database middleware is a kind of effective means that realizes distributed system on original unit Database Systems.Adopting the middleware mode is by intercepting and capturing and analyze the data stream of client with server interaction, realizing transparent database replication.
People such as Kemme revise open source code, the basis of DBMS-PostgreSQL on, realized synchronization replication realize owing to duplicate in that DBMS is inner, therefore can obtain more performance.The said method document that specifically sees reference: B.Kemme and G.Alonso.Don ' t be lazy, be consistent:Postgres-R, A new wayto implement Database Replication.In Proc.Of Inf.Conf.On Very Large Database, VLDB ' 00,2000.This method is the database replication operation towards affairs, but following shortcoming is arranged:
1: must know the source code of DBMS, realize complicated;
2: when client access was frequent, the reversed ratio of affairs was big, caused throughput to descend.
3: be difficult to realize duplicating between the xenogenesis DBMS.This mainly be because: must in various DBMS, increase identical replication protocol, need between xenogenesis DBMS, carry out the conversion of data layout.If replication protocol has been revised, to revise various DBMS accordingly.
4: adopt the mode of receiver-driven to realize partly duplicating.Whether receiver-driven gives all member node with each information broadcast, selected to handle by the recipient.This mode has increased the traffic, has had a strong impact on the extensibility of system.
People such as Amir adopt the mode of middleware, have solved to exist network to cut apart the problem of duplicating under the situation on the wide area network.The visible list of references of the realization of this method: Y.Amir, C.Danilov, M.Miskin-Amir, J.Stanton, and C.Tutu.Practical Wide Area Database Replication.Technical Report CNDS-2002-1, JohnsHopkins University, 2002.But the copy mode that this method provided is an oriented manipulation, that is to say that duplicates operates liking single renewal at every turn, whenever duplicating an operation all will be to all database nodes broadcasting once, for duplicating of affairs that comprise many statements, then correspondingly need broadcasting repeatedly.Like this, all will broadcast once owing to whenever duplicate an operation, so communication overhead is big; For affairs, because each operation all will be synchronously once, the execution time of affairs is elongated, and the transaction response time is slack-off, has also reduced the throughput of system.
The implementation of being duplicated by above-mentioned two kinds of existing sync databases as seen, above-mentioned two kinds of methods have defective and advantage separately, if can gather the strong point of two kinds of methods, maximize favourable factors and minimize unfavourable ones, a kind of database replication middleware towards affairs is provided, just can realizes synchronization replication easily and fast, effectively database.
Summary of the invention
One of purpose of the present invention provides the method for utilizing middleware to carry out database replication in a kind of group communication system, this method can seating surface to the database replication of affairs and guarantee the stric consistency of data; Another object of the present invention provides the middleware that carries out database replication in a kind of group communication system, this middleware can seating surface to the database replication of affairs and guarantee the stric consistency of data.
To achieve these goals, the invention provides the method for utilizing middleware to carry out database replication in a kind of group communication system, be used on a plurality of database storing nodes between the duplicating of database, to guarantee the consistance of database on each node; This method comprises:
The local matter execution in step: the database storing node receives the transactions requests of sending from client application and carries out this affairs at local node;
Affairs broadcasting step: the affairs that will run succeeded in database storing node this locality are broadcast to all database storing nodes by group communication system;
The affairs ordered steps: group communication system utilizes its TOTAL-ORDER communication mechanism that all affairs by its broadcasting are carried out sequencing, forms a transaction queues;
Out of order determining step: the affairs that the database storing node has been carried out in the local matter execution in step have a corresponding position in described transaction queues, and whether judge had before this correspondence position in the described transaction queues and the afoul affairs of the described affairs of having carried out;
Sequencing affairs execution in step: carry out described transaction queues according to the judged result of out of order determining step; Do not have when conflict when judging, then carry out other affairs except that the affairs of having carried out in the transaction queues in proper order according to the affairs in the described transaction queues; When judgement has conflict, then cancel the described affairs of having carried out, carry out transaction queues then;
Database writes back step: when on all database storing nodes all successful execution after the same affairs in the described transaction queues, each database storing node upgrades its data-base content according to the execution result of these affairs.
The present invention also provides the middleware that utilizes middleware to carry out database replication in a kind of group communication system, and this middleware is installed on the database storing node, and can with the data base management system (DBMS) communication on client application and the database storing node.This middleware comprises:
The local matter execution module is used to intercept and capture the transactions requests that is sent to the database storing node from client application, and this transactions requests is sent to the data base management system (DBMS) execution, and execution result is back to client application;
The affairs broadcast module, the affairs that are used for running succeeded in database storing node this locality are broadcast to all database storing nodes by group communication system;
Out of order judge module, the affairs that the database storing node has been carried out in the local matter execution module have a corresponding position in described transaction queues, this module is used for judging whether described transaction queues had and the afoul affairs of the described affairs of having carried out before this correspondence position;
Sequencing affairs execution module is used for carrying out described transaction queues according to the judged result of out of order judge module; Do not have when conflict when judging, then carry out other affairs except that the affairs of having carried out in the transaction queues in proper order according to the affairs in the described transaction queues; When judgement has conflict, then cancel the described affairs of having carried out, carry out transaction queues then;
Database writes back module, be used for when on all database storing nodes all successful execution after the same affairs of described transaction queues, each database storing node upgrades its data-base content according to the execution result of these affairs.
The present invention is that unit broadcasts database node with affairs when carrying out database replication, greatly reduces the communication overhead of system, has improved the response time of affairs, and can guarantee the stric consistency of database on the different nodes.
Description of drawings
Fig. 1 utilizes middleware to carry out the system schematic of database replication in the conventional group communication system;
Fig. 2 is the composition synoptic diagram of the present invention's middleware of being used for database replication;
Fig. 3 is the process flow diagram of database copy method of the present invention.
Embodiment
Below in conjunction with the drawings and specific embodiments the present invention is described in further detail.
In Database Systems, generally include a plurality of database storing nodes,, a complete database identical with other node content is arranged all on any one node for full replicated database system.Connect by network between these database storing nodes, in the present invention, communicate by group communication system between the database storing node.Fig. 1 shows the conventional middleware that utilizes and carries out the system of database replication in group communication system, usually comprise a plurality of database storing nodes in this system, Fig. 1 only exemplarily shows node 1 and 2 two data library storage of node node, all comprise a database that content is identical in each node 1 and 2, and also include on node 1 and 2 and be used for the data base management system (DBMS) that to operate database.All also have a middleware on the node 1 and 2, this middleware can be intercepted and captured the data stream between client application and the data base management system (DBMS) and this data stream analyzed and operate, to realize the transparent database replication of data base management system.As shown in Figure 1, in group communication system, a group communication system is arranged all on each database storing node 1 and 2, node 1 and 2 utilizes group communication system to carry out data and transmission of Information by network 3.
Middleware on the database storing node is realized by a specific program that normally in one embodiment of the invention, the structure of this middleware as shown in Figure 2.In Fig. 2, this middleware comprises that local matter execution module, affairs broadcast module, out of order judge module, sequencing affairs execution module and database write back module.Wherein, the affairs broadcast module preferably includes semantic analysis unit and affairs recomposition unit.Below in conjunction with Fig. 1 and Fig. 2, describe the present invention according to database replication process flow diagram shown in Figure 3.
Step 10 is the local matter execution in step, in this step, local matter execution module in the middleware is intercepted and captured the transactions requests that is sent to the database storing node from client application, and this transactions requests is sent to data base management system (DBMS) carries out, and execution result is back to client application.Usually, affairs comprise a plurality of database manipulations, and these database manipulations are generally SQL statement.An exemplary transactions requests is as follows:
Begin ∥ begins
The operation of statement S1 ∥ read data;
Statement S2 ∥ data writing operation;
Statement S3 ∥ data writing operation;
Commit ∥ submits to
This transactions requests comprises three data base manipulation statements of S1～S3, and with the end of a submiting command (commit) as affairs.The database storing node normally receives and carries out this transactions requests in interactively mode, that is to say, operation of the every execution of database storing node all can be back to client with operating result, and send next action statement by client, send the end of submiting command until client as transactions requests.For example in aforesaid exemplary transaction request, client application send statement S1, database storing node are back to client with the result after carrying out this operation S1, then by client application send statement S2 again, and the like.In this process, the failure of any one operation in carrying out affairs all will be cancelled these affairs, and returns a failed message to client, does not for example carry out the needed data of operation when finding on the database storing node, then cancels these affairs; Perhaps, the database storing node receives the rollback order (Rollback) that client application itself sends in this process, also can cause cancelling of these affairs.
Should be noted that, in local matter execution in step 10, the database storing node is when each operation that the execution client is sent, only the content in the database is called in internal memory or the buffer memory and operate, and operating result returned client, even but this is operating as and upgrades operation, promptly relate to the operation of revising data-base content, the also content in the new database more not in this step.
Step 20 is affairs broadcasting step, in this step, the transactions requests of sending from client application is broadcast to all database storing node with these affairs by group communication system by the affairs broadcast module in the middleware after memory node this locality runs succeeded.Here, the affairs broadcast module is preferably only broadcasted away the new affairs of those operation compositions that relate to renewal in the affairs, to reduce communication overhead.In one embodiment, the affairs broadcast module comprises semantic analysis unit and affairs recomposition unit, wherein, the transactions requests that the semantic analysis element analysis sends from client application, therefrom select those action statement that relate to renewal, the action statement that be will be referred to upgrade by the affairs recomposition unit is organized into affairs and broadcasts away then.With the aforementioned affairs of being made up of three data base manipulation statements of S1～S3 is example, and the semantic analysis unit carries out semantic analysis to statement S1～S3, picks out data writing operation statement S2 and S3; The affairs recomposition unit is organized into following affairs with statement S2 and S3 and broadcasts away:
Begin ∥ begins
Statement S2 ∥ data writing operation;
Statement S3 ∥ data writing operation;
Commit ∥ submits to.
Step 30 is the affairs ordered steps, and in this step, group communication system utilizes its TOTAL-ORDER communication mechanism that all affairs by its broadcasting are carried out sequencing, forms a transaction queues.In Database Systems based on group communication system, in a period of time, usually a plurality of affairs that have the database storing node broadcasts to go out are propagated in network, TOTAL-ORDER communication mechanism in the group communication system can carry out sequencing to these affairs, makes a plurality of affairs enter into the database storing node with identical order.
Step 40 is out of order determining step, in this step, the out of order judge module of the middleware on the database storing node judge the affairs of in the local matter execution in step, having carried out whether meet with transaction queues in affairs have and conflict.As previously mentioned, in local matter execution in step 10, the local matter execution module of the middleware on data library storage node is intercepted and captured to be sent to the transactions requests of database storing node and to be sent to data base management system (DBMS) from client application and is carried out; And in affairs broadcasting step 20 and affairs ordered steps 30, by the affairs broadcast module in the middleware all database storing nodes are broadcasted and sent in order to these affairs by group communication system behind the group communication system sequencing again after this transactions requests runs succeeded.Like this, for a data library storage node, certain or some affairs in the transaction queues of sequencing are carried out in local matter execution in step 10, and the affairs that this has been carried out also are in a certain correspondence position in the transaction queues simultaneously.Thereby whether this database storing node need be judged in transaction queues, be in the affairs correspondence position affairs of having carried out before and conflict mutually with the affairs that this has been carried out.
Step 50 is sequencing affairs execution in step, and in this step, sequencing affairs execution module is carried out transaction queues according to the judged result of out of order judge module.The judged result of out of order judge module comprises two kinds, and a kind of not conflict does not so then need to repeat these affairs again, and continues to carry out other affairs in the transaction queues; A kind of is that conflict is arranged, and then needs transaction retraction that this has been carried out, just the execution result of these affairs in internal memory or the buffer memory is back to the state before these affairs are carried out, and then carries out transaction queues again.The sequencing affairs execution module of middleware all sends the message that runs succeeded to other all database storing nodes on data library storage node after the affairs in preferably every successful execution transaction queues when carrying out transaction queues.
Step 60 writes back step for database, in this step, when on all database storing nodes all successful execution after the same affairs in the transaction queues, write back module by the database of the middleware on each node and upgrade its data-base content according to the execution result of these affairs, all affairs in executing transaction queues are also finished the renewal of database.Database writes back module can judge whether all same affairs in the successful execution transaction queues of all nodes according to the message that runs succeeded that other node sends.Database writes back module as long as find to have a node not have a certain affairs in can the successful execution transaction queues, does not just carry out the renewal of database, and this transaction rollback (rollback) and its are deleted from transaction queues.Like this, can guarantee that the database of each node can carry out identical renewal, thereby guarantee the consistance of database on the different nodes.

Claims (10)

1, the method for utilizing middleware to carry out database replication in a kind of group communication system, be used on a plurality of database storing nodes between the duplicating of database, to guarantee the consistance of database on each node; This method comprises:
The local matter execution in step: the database storing node receives the transactions requests of sending from client application and carries out this affairs at local node;
Affairs broadcasting step: the affairs that will run succeeded in database storing node this locality are broadcast to all database storing nodes by group communication system;
The affairs ordered steps: group communication system utilizes its TOTAL-ORDER communication mechanism that all affairs by its broadcasting are carried out sequencing, forms a transaction queues;
Out of order determining step: the affairs that the database storing node has been carried out in the local matter execution in step have a corresponding position in described transaction queues, and whether judge had before this correspondence position in the described transaction queues and the afoul affairs of the described affairs of having carried out;
Sequencing affairs execution in step: carry out described transaction queues according to the judged result of out of order determining step; Do not have when conflict when judging, then carry out other affairs except that the affairs of having carried out in the transaction queues in proper order according to the affairs in the described transaction queues; When judgement has conflict, then cancel the described affairs of having carried out, carry out transaction queues then;
Database writes back step: when on all database storing nodes all successful execution after the same affairs in the described transaction queues, each database storing node upgrades its data-base content according to the execution result of these affairs.
2, utilize middleware to carry out the method for database replication in the group communication system according to claim 1, it is characterized in that, in the local matter execution in step, described memory node is carried out the transactions requests that client application sends with alternant way, all operating result is back to client application for each operation that comprises in the transactions requests, but new database content more not.
3, utilize middleware to carry out the method for database replication in the group communication system according to claim 1, it is characterized in that, in the local matter execution in step, when a transactions requests is carried out failure or receive the rollback order of client application transmission, cancel the execution of these affairs.
4, utilize middleware to carry out the method for database replication in the group communication system according to claim 1, it is characterized in that, in affairs broadcasting step, only the operational group that comprises database update in the affairs is made into affairs and broadcasts.
5, utilize middleware to carry out the method for database replication in the group communication system according to claim 1, it is characterized in that, in sequencing affairs execution in step, all send the message that runs succeeded after the affairs in the described transaction queues of the every successful execution of database storing node to other database storing node.
6, utilize middleware to carry out the method for database replication in the group communication system according to claim 5, it is characterized in that, write back in the step at database, the message that runs succeeded that the database storing node sends by each node judges whether all same affairs of the described transaction queues of successful execution of all database storing nodes.
7, utilize middleware to carry out the middleware of database replication in a kind of group communication system, this middleware is installed on the database storing node, and can with the data base management system (DBMS) communication on client application and the database storing node, and it is characterized in that this middleware comprises:
The local matter execution module is used to intercept and capture the transactions requests that is sent to the database storing node from client application, and this transactions requests is sent to the data base management system (DBMS) execution, and execution result is back to client application;
The affairs broadcast module, the affairs that are used for running succeeded in database storing node this locality are broadcast to all database storing nodes by group communication system;
Out of order judge module, the affairs that the database storing node has been carried out in the local matter execution module have a corresponding position in described transaction queues, this module is used for judging whether described transaction queues had and the afoul affairs of the described affairs of having carried out before this correspondence position;
Sequencing affairs execution module is used for carrying out described transaction queues according to the judged result of out of order judge module; Do not have when conflict when judging, then carry out other affairs except that the affairs of having carried out in the transaction queues in proper order according to the affairs in the described transaction queues; When judgement has conflict, then cancel the described affairs of having carried out, carry out transaction queues then;
Database writes back module, be used for when on all database storing nodes all successful execution after the same affairs of described transaction queues, each database storing node upgrades its data-base content according to the execution result of these affairs.
8, utilize middleware to carry out the middleware of database replication in the group communication system according to claim 7, it is characterized in that, described affairs broadcast module comprises:
The semantic analysis unit is used to analyze the transactions requests of sending from client application, therefrom selects the operation that relates to renewal;
The affairs recomposition unit is used for that the described action statement that relates to renewal is organized into affairs and broadcasts.
9, utilize middleware to carry out the middleware of database replication in the group communication system according to claim 7, it is characterized in that, all send the message that runs succeeded after the affairs in the described transaction queues of the described every successful execution of sequencing affairs execution module to other database storing node.
10, utilize middleware to carry out the middleware of database replication in the group communication system according to claim 9, it is characterized in that described database writes back the message that runs succeeded that module sends according to other database storing node and judges whether all same affairs of the described transaction queues of successful execution of all database storing nodes.
CNB2005100552154A 2005-03-08 2005-03-16 Method and middleware for data base duplicating using middleware in group communication system Expired - Fee Related CN100377156C (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CNB2005100552154A CN100377156C (en) 2005-03-08 2005-03-16 Method and middleware for data base duplicating using middleware in group communication system

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
CN200510051362.4 2005-03-08
CN200510051362 2005-03-08
CNB2005100552154A CN100377156C (en) 2005-03-08 2005-03-16 Method and middleware for data base duplicating using middleware in group communication system

Publications (2)

Publication Number Publication Date
CN1831810A true CN1831810A (en) 2006-09-13
CN100377156C CN100377156C (en) 2008-03-26

Family

ID=36994115

Family Applications (1)

Application Number Title Priority Date Filing Date
CNB2005100552154A Expired - Fee Related CN100377156C (en) 2005-03-08 2005-03-16 Method and middleware for data base duplicating using middleware in group communication system

Country Status (1)

Country Link
CN (1) CN100377156C (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102063502A (en) * 2011-01-06 2011-05-18 天津神舟通用数据技术有限公司 Method for realizing synchronization of data in heterogeneous database
CN101840348B (en) * 2009-12-24 2013-02-20 陕西移数通电讯有限公司 Database middleware system
CN103345502A (en) * 2013-07-01 2013-10-09 曙光信息产业(北京)有限公司 Transaction processing method and system of distributed type database
CN104679841A (en) * 2015-02-11 2015-06-03 北京京东尚科信息技术有限公司 Consumption terminal data flow copying method and system
CN105512266A (en) * 2015-12-03 2016-04-20 曙光信息产业(北京)有限公司 Method and device for achieving operational consistency of distributed database
CN106415536A (en) * 2014-03-10 2017-02-15 甲骨文国际公司 Instantaneous unplug of pluggable database from one container database and plug into another container database

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5995972A (en) * 1996-12-30 1999-11-30 J.D. Edwards World Source Company System and method for retrieving data
AR028459A1 (en) * 2000-05-18 2003-05-07 Nagravision Sa METHOD OF DISTRIBUTED DATABASE MANAGEMENT
CN1185596C (en) * 2001-12-21 2005-01-19 鸿富锦精密工业（深圳）有限公司 Distributed case management system and its method

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101840348B (en) * 2009-12-24 2013-02-20 陕西移数通电讯有限公司 Database middleware system
CN102063502A (en) * 2011-01-06 2011-05-18 天津神舟通用数据技术有限公司 Method for realizing synchronization of data in heterogeneous database
CN102063502B (en) * 2011-01-06 2013-07-31 天津神舟通用数据技术有限公司 Method for realizing synchronization of data in heterogeneous database
CN103345502A (en) * 2013-07-01 2013-10-09 曙光信息产业(北京)有限公司 Transaction processing method and system of distributed type database
CN103345502B (en) * 2013-07-01 2017-04-26 曙光信息产业(北京)有限公司 Transaction processing method and system of distributed type database
CN106415536A (en) * 2014-03-10 2017-02-15 甲骨文国际公司 Instantaneous unplug of pluggable database from one container database and plug into another container database
CN106415536B (en) * 2014-03-10 2019-12-17 甲骨文国际公司 method and system for pluggable database transmission between database management systems
CN104679841A (en) * 2015-02-11 2015-06-03 北京京东尚科信息技术有限公司 Consumption terminal data flow copying method and system
CN104679841B (en) * 2015-02-11 2018-06-08 北京京东尚科信息技术有限公司 A kind of consumption terminal data flow replication method and system
CN105512266A (en) * 2015-12-03 2016-04-20 曙光信息产业(北京)有限公司 Method and device for achieving operational consistency of distributed database

Also Published As

Publication number Publication date
CN100377156C (en) 2008-03-26

Similar Documents

Publication Publication Date Title
US8041735B1 (en) Distributed file system and method
USRE47106E1 (en) High-performance log-based processing
JP4557975B2 (en) Reassign ownership in a non-shared database system
US9501550B2 (en) OLAP query processing method oriented to database and HADOOP hybrid platform
US20050160315A1 (en) Geographically distributed clusters
CA2550614C (en) Cluster database with remote data mirroring
WO2013155752A1 (en) Database and hadoop hybrid platform-oriented olap query processing method
US20050289152A1 (en) Method and apparatus for implementing a file system
US20050071336A1 (en) Systems and methods for logging and recovering updates to data structures
US20040210673A1 (en) Messaging between heterogeneous clients of a storage area network
KR20060117505A (en) A recovery method using extendible hashing based cluster log in a shared-nothing spatial database cluster
CN1831810A (en) Method and middleware for data base duplicating using middleware in group communication system
CA2512312A1 (en) Metadata based file switch and switched file system
Rodrigues et al. The GlobData fault-tolerant replicated distributed object database
Byun et al. Resilient data management for replicated mobile database systems
CN1835607A (en) Short message two-staged netgate based on PC server and service environment
CN112286889B (en) Wide area network-oriented metadata copy synchronization method for distributed file system
Siddha A Persistent Snapshot Device Driver for Linux
JP4664931B2 (en) Partitioning database ownership between different database servers to control access to the database
Triantafillou et al. Multiclass replicated data management: exploiting replication to improve efficiency
Ji Availability, scalability and cost-effectiveness of cluster-based Internet infrastructures
Delis et al. Database architectures
Jenq Performance measurement, modelling, and evaluation of integrated concurrency control and recovery algorithms in distributed database systems
Byun et al. Fault-tolerant quorum consensus scheme for replication control in mobile distributed database systems: FTQC
Jin et al. A distributed persistent object store for scalable service

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
ASS Succession or assignment of patent right

Free format text: FORMER OWNER: INSTITUTE OF COMPUTING TECHNOLOGY, CHINESE ACADEMY OF SCIENCES

Effective date: 20130530

C41 Transfer of patent application or patent right or utility model
COR Change of bibliographic data

Free format text: CORRECT: ADDRESS; FROM: 100080 HAIDIAN, BEIJING TO: 518129 SHENZHEN, GUANGDONG PROVINCE

TR01 Transfer of patent right

Effective date of registration: 20130530

Address after: 518129 Bantian HUAWEI headquarters office building, Longgang District, Guangdong, Shenzhen

Patentee after: For Technology Co.,Ltd.

Address before: 100080 Haidian District, Zhongguancun Academy of Sciences, South Road, No. 6, No.

Patentee before: Institute of Computing Technology, Chinese Academy of Sciences

C56 Change in the name or address of the patentee
CP01 Change in the name or title of a patent holder

Address after: 518129 Bantian HUAWEI headquarters office building, Longgang District, Guangdong, Shenzhen

Patentee after: HUAWEI TECHNOLOGIES Co.,Ltd.

Address before: 518129 Bantian HUAWEI headquarters office building, Longgang District, Guangdong, Shenzhen

Patentee before: For Technology Co.,Ltd.

CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20080326

CF01 Termination of patent right due to non-payment of annual fee