CN100397803C - N+1 duplicates data real-time synchronising method - Google Patents
N+1 duplicates data real-time synchronising method Download PDFInfo
- Publication number
- CN100397803C CN100397803C CNB2005100182316A CN200510018231A CN100397803C CN 100397803 C CN100397803 C CN 100397803C CN B2005100182316 A CNB2005100182316 A CN B2005100182316A CN 200510018231 A CN200510018231 A CN 200510018231A CN 100397803 C CN100397803 C CN 100397803C
- Authority
- CN
- China
- Prior art keywords
- data
- synchrodata
- backup
- synchronization
- mainboard
- 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.)
- Expired - Fee Related
Links
Images
Abstract
The present invention relates to a real-time synchronization method for N+1 backup data. The present invention is characterized in that N is lager than or equal to 1, a structure of one backup and N main pieces is adopted, the dynamic synchronization relation between the main pieces and the backup is reflected by a synchronization linked list, and the logic identification of a backup board and Socket information are recorded by the synchronization linked list. A main board asynchronously sends synchronization data, namely a data modification set and a timestamp derived from updated data to the backup board in real time according to the synchronization linked list, and records the synchronization data in an external storage file. The timestamp for marking a data updating sequence is expressed by the combination of the sequence number of the external storage file which comprises the synchronization data and the offset in the external storage file. The backup board is in charge of receiving the synchronization data and checking the integrality of the synchronization data. The backup board backups the synchronization data according to the timestamp of the synchronization data, or actively sends a request to a main board for the rear backup data of the lost synchronization data. The function that one board card provides a real-time data backup service for a plurality of board cards is realized. The backup board is allowed to be added into a backup system at any time, and the extensibility of the backup system is realized. The present invention also realizes a rapid synchronization method for main and backup out of synchronization by the way.
Description
Technical field
The present invention relates to the data high reliability of the communications field, a kind of N+1 backed up data real-time synchronization method and device thereof are provided, realize that an integrated circuit board provides the service of backing up in realtime of data for a plurality of integrated circuit boards.
Background technology
In the communications field,, some key equipments are adopted mode, not losing of assurance data and not the interrupting of business of 1+1 backup in order to improve the high reliability of system.A common integrated circuit board has separate processor, internal memory, network interface and storage subsystem.If N piece integrated circuit board is arranged, need to dispose N piece backup plate so, the system cost of this 1+1 backup mode is very high.If active and standby integrated circuit board is positioned at machine frame inside, because the number of machine frame slot is limited, the configuration mode of this 2N also is difficult to realize.The utilization of the resource of 1+1 backup mode is also insufficient in addition, and under asymmetric backup mode, the idle ratio of integrated circuit board is 1/1.
Several data is disclosed in the prior art synchronously or the technology of backup.Usually utilize these technology,, and then realize high reliability and load balancing so that realize the data on the active and standby plate are realized reaching real-time consistency synchronously.Introduce the data synchronizing function in addition, can not grievous injury master data base access efficiency.
Chinese patent literature 02103019.7 discloses the method for active and standby plate real time data synchronization in the communication system, in the active and standby data buffer zone that is provided with; When the mainboard data modification, in its buffering area, write the data that mainboard was revised; Buffering area is handled data; When buffer area data volume reaches one regularly, the real-time synchronized process of mainboard read data and send to the real-time synchronized process of slave board from buffering area; Upgrade the slave board database, and operating result is returned to mainboard; If operating result is synchronous success, corresponding data in the delete buffer then, otherwise resend data, in buffering area, be recorded as sky.
There are some problems in this method for synchronous, at first is the restriction in sync buffering district.If active and standby out of alignment time is longer, mainboard needs the buffer memory mass data, even causes overflowing of buffering area.Next is an efficiency.Only upgrading the slave board database, and operating result is returned to mainboard, simultaneous operation is just finished, so mainboard must wait for the renewal operation of slave board, and this can bring delay to data sync.Also has safety issue.The synchronization principles of this method is that the database manipulation of mainboard is reformed at slave board, reaches data consistency.This way not only has the execution efficiency, the more important thing is the dangerous property of possibility, because the database manipulation that mistake is carried out on mainboard can propagate on the slave board equally, makes slave board be difficult to the effect of performance data disaster tolerance.There is limitation in the transmission method of this synchrodata at last.Use retransmission mechanism to send synchrodata according to the order of buffering area queuing, lack flexibility, slave board can only be passive accepts synchrodata, not have the ability of active error correction.
Chinese patent literature 02111898.1 discloses the method for data synchronization in a kind of main stock redundance system, this method is when active and standby during with the node triggering synchronous, the master obtains function with node by data, the poll master uses each object of each module in the node, obtaining needs data in synchronization, again with all data arrangement at certain region of memory, compress then and be sent to secondary node, after secondary node receives synchrodata, carry out the inverse process of aforesaid operations, thereby realized the data sync of primary, spare node.This method has realized the unified management of synchrodata, but its shortcoming also is tangible.At first the efficient of poll is lower.The method of unified management can influence each module real-time property of synchronizing in addition.
If system does not allow slave board to add at any time, and do not support active and standby asynchronous starting, then the extensibility of system is not known where to begin.System should handle the various combined situation of active and standby boot sequence, guarantees the correctness of data sync.
Except the data sync under the normal condition, also exist because faults such as networking cause occurring between slave board and the mainboard situation of active and standby step-out, if can not realize active and standby synchronous again fast, will influence the efficient and the real-time of data sync.
Summary of the invention
The objective of the invention is provides N+1 backed up data real-time synchronization method in order to overcome the expensive problem of 1+1 data backup high reliability scheme, realizes that a node provides the service of backing up in realtime of data for a plurality of nodes.The present invention allows backup node to join at any time in the standby system, supports active and standby asynchronous starting, and the system that makes is with good expansibility.The present invention has also realized the fast synchronization method under a kind of active and standby step loss condition.
Technical scheme of the present invention is: N+1 backed up data real-time synchronization method, it is characterized in that N 〉=1, employing one is equipped with N master's backup architecture, uses active and standby dynamic synchronized relation of synchronous chained list reflection, synchronously logical identifier, the socket information of chain table record slave board; Mainboard according to synchronous chained list in real time with synchrodata---be data modification collection and the timestamp that Data Update derives, asynchronous transmission is to slave board, and storage file record synchrodata outside; The timestamp that identification data upgrades order uses the sequence number of synchrodata place file and the combination of side-play amount hereof to represent; Slave board is responsible for receiving isochronous data and is detected the integrality of synchrodata, according to timestamp Backup Data or active Backup Data after the synchrodata that the mainboard request is lost of synchrodata.
Aforesaid N+1 backed up data real-time synchronization method is characterized in that adopting the following step:
The step of taking on the mainboard comprises:
Synchronous chained list of all mainboard initialization of step 1 is used to write down integrated circuit board and attribute thereof; The chained list initial condition is set to sky synchronously.
Step 2 starts independently watcher thread synchronously, confirms synchronization request; Each database uses same watcher thread, and watcher thread is confirmed different data synchronization request by database-name.
Step 3 writes down logical identifier, the socket information of slave board in synchronous chained list, determine the forwarding position of synchrodata; Start independently Synchronous Processing thread, be responsible for the reception and the processing of synchrodata.
Step 4 is checked the backup object of synchronous chained list before the pre-submission of the renewal of mainboard data; If exist then execution in step 5; The definition of pre-submit state is that local data base upgrades to be finished substantially, just also not final submission the to.
What each mainboard of step 5 was real-time sends to slave board with synchrodata; If the detection link failure is then being nullified slave board in the chained list synchronously; Every synchrodata is safeguarded a timestamp simultaneously; Connectionless protocol is used in communication, guarantees the correctness of the synchrodata that the opposite end receives, but does not need to guarantee the order and the integrality of data.In application layer process synchrodata lose, out-of-sequence situation.
Mainboard does not need the execution result of waiting for that slave board is synchronous, reduce active and standby mutual, improved data synchronization efficient.
Because active and standby plate has adopted more new database of different modes, has improved the access efficiency and the fault-tolerant ability of database.
Step 6 writes down synchrodata on local file, and submits Data Update to; Mainboard is the inactive file of deletion regularly.If memory space is lower than certain threshold value, then, delete inactive sync data file according to file logic sequence number size, reclaim memory space; If slave board need be visited the file of having deleted, then step-out is too of a specified duration for slave board and mainboard, and slave board need restart the slave board database, carries out database recovery process.
Usually the space of external memory is much larger than the memory space of core buffer, is large enough to hold the data that are not synchronized to slave board after the active and standby long-time step-out, prevented losing of synchrodata.
Step 7 circulation execution in step 4-6 is up to receiving active and standby switching or database close message.
The step of taking on the slave board comprises:
Step 1 starts several synchronization request threads, and application is also determined to concern with the data sync of each mainboard; The synchronization request thread need use parameters such as data configuration, network site that configuration file obtains each mainboard to initiate connection request, after mainboard is confirmed, connects thread and withdraws from automatically; Above process all is asynchronous carrying out, with guaranteed efficiency.
Step 2 record mainboard and attribute thereof start independently Synchronous Processing thread, are responsible for the reception and the processing of synchrodata.The step of taking on this step and the mainboard 3 is similar, is responsible for accepting, handling the synchrodata of opposite end.
Step 3 receiving isochronous data is checked the timestamp of synchrodata, if equal current required timestamp, and new database more then; If greater than the timestamp of current required synchrodata, then in this synchrodata of temporary file buffer memory, check whether temporary file has required synchrodata, there is then more new database; If there is no then to the required synchrodata of mainboard request, and new database more.Except the review time stabs, slave board is also checked the integrality of synchrodata, and to the synchrodata that receives according to timestamp ordering; If detect the situation that the synchrodata genetic sequence is disorderly or lose, slave board initiatively sends required synchrodata to the mainboard request.Has initiatively error correcting capability.
Step 4 is calculated the timestamp of up-to-date synchrodata, and checks whether temporary file has available synchrodata; There is then more new database, otherwise turns to step 3.
Step 5 circulation execution in step 3-4 is up to receiving active and standby switching or database close message.
Above step 3 has realized the fast synchronization method under a kind of active and standby step loss condition.The step-out decision condition is the timestamp of the timestamp of the synchrodata received on the slave board greater than the required synchrodata in this locality, and this declaratives synchrodata has been lost, and need obtain from mainboard.Even the timestamp of the synchrodata of receiving does not abandon greater than the timestamp of the required synchrodata in this locality, but be cached in the local temporary file yet, utilize the synchrodata and the network bandwidth to greatest extent, the active and standby step-out data of concurrent processing.If slave board has been caught up with the Data Update frequency of mainboard, then withdraw from the processing of active and standby step-out.If do not support active and standby asynchronous starting, then allow slave board to join standby system at any time and just do not know where to begin.The various built-up sequences that must guarantee active and standby startup can not influence the synchronous correctness of standby system.
Support active and standby asynchronous starting, specifically take following steps:
Before step 1 database starts, check the state of integrated circuit board.
If the state that step 2 database will be set is for being equipped with execution in step 5.
Step 3 checks whether the state of this plate is main, otherwise wait reports an error up to overtime.
Step 4 setting data storehouse state is main, returns.
Step 5 checks whether the state of all associated panel is main.Otherwise wait for and reporting an error up to overtime.
Step 6 setting data storehouse state returns for being equipped with.
Advantage of the present invention is as follows:
1. reduced the cost of highly reliable system, the utilance height of resource is easy to expansion.In the system of N+1 mechanism, a backup of every N piece plate configuration plate, the idle ratio of resource is 1: N.
2. allow backup node to join at any time in the standby system, support active and standby asynchronous starting, realize the extensibility of system.
3. handled the situation that occurs active and standby step-out when the networking is unusual, realized that a kind of backstage carries out the synchronous method of rapid data automatically.
4. improve the high availability of database.If any veneer breaks down, slave board is responsible for recovering the data, services of fault integrated circuit board.
5. do not need extra hardware device, have certain versatility.
Though detailed open and illustrated and person of skill in the art will appreciate that the present invention: need not to break away from spirit of the present invention and protection range, just can make appear on the various forms with details on change.
Description of drawings
Fig. 1, the structure chart of N+1 back-up device of the present invention.
Embodiment
Below embodiments of the invention are described in further detail.
Important integrated circuit board for communication equipment adopts the heat of 1+1 to be equipped with mode basically.Lower to those high reliability request, the equipment that strict cost control is arranged, need between redundancy, grade of service expectation and the system cost and the complexity of equipment, obtain a balance, the N+1 backup is a more suitably solution, and the present invention provides the problem of the service of backing up in realtime of data in order to solve a node for a plurality of nodes.
Fig. 1 is the structural representation of N+1 back-up device.This device is formed by Ethernet is interconnected by N+1 integrated circuit board.N piece integrated circuit board has separate processor, storage subsystem, internal memory and Network Interface Unit separately, and one or more databases operate on the Main Processing Unit.Have only a slave board to be responsible for receiving synchrodata, and keep and the data realtime uniform of each mainboard from N piece mainboard.Integrated circuit board can be positioned at identical or machine frame independently, physically has loose coupling.
Database on the mainboard has relative independentability, and slave board needs initiatively to set up correlation to the mainboard registration, and slave board breaks down can not influence the normal data manipulation of mainboard.Data flow on the mainboard is forwarded on the slave board by the synchronization link of setting up, and realizes many-to-one data backup.There are mapping relations in database unified management on the slave board, its subdata base and each mainboard database, and this mapping relations can be asynchronous foundation, have concurrency and advantage efficiently.Subdata base uses different working regions, guarantees the subdata base non-interference.
Below with active and standby in real time synchronously in active and standby reciprocal process be example, this method is described;
Step 1, the N of configuration backup system is 2, promptly has two mainboards and a slave board.IP is set at 90.0.1 successively, 90.0.2 and 90.0.3.Datamation district on the mainboard 1 is/dev/A1, and the data logical identifier is A1, and the datamation district on the mainboard 2 is/dev/A2 that the data logical identifier is A2.The data logical identifier of slave board is A3, has two data service areas and mainboard correspondence.
Step 2, the database application of independent startup mainboard 1,2.
Step 3, the database application of startup slave board, then slave board connects mainboard 1 and mainboard 2 respectively, sets up active and standby synchronized relation.The synchronous chained list of the corresponding renewal of active and standby plate reflects current synchronized relation.
Data Update takes place and forms synchrodata in step 4, mainboard 1.Synchrodata comprises message headers such as timestamp and data length.The synchronous chained list of mainboard inspection has slave board member A3, and the IP under it is 90.0.0.3, and then mainboard uses udp protocol, synchrodata is sent to the integrated circuit board at 90.0.0.3 place.Additional record on the sync data file of local memory device then.If sync data file surpasses certain size such as 5MB, a then newly-built sync data file, and increase its logic sequence number automatically.Mainboard 1 also can be made regular check on disk space, deletes inactive sync data file.
Step 5, slave board receives the synchrodata from mainboard 1, and checks the timestamp of synchronization message.The synchrodata timestamp that receives is the current required timestamp of slave board, then according to the modification set content of synchrodata, upgrades the slave board database, simultaneously relatively on the slave board/position of dev/A1, additional record on sync data file.Slave board also will be checked the size of sync data file, and carries out identical file operation.
Step 6 if Data Update has also taken place mainboard 2, is used identical step.But after slave board receives synchrodata, use service area/dev/A2 on the slave board, guarantee can not disturb between synchrodata.
Step 7, circulation execution in step 4～6, the Data Update on the mainboard 1,2 just have been synchronized to slave board in real time.
If fault has appearred in the communication link of mainboard 1 and slave board, continued about 30 seconds.The synchrodata in this period will all be lost, and active and standby step-out has taken place, and the treatment step of active and standby step-out is as follows:
Step 1, the communication link of mainboard 1 and slave board recover normal, and mainboard continues to send to slave board with current synchrodata.
Step 2, slave board have been received the record of synchrodata, also will check the timestamp of synchrodata, find the timestamp 50 of the timestamp 100 of reception greater than current needs.Slave board is the buffer memory synchrodata at first, and checks whether the temporary file of slave board has the complete synchrodata that equals timestamp 50.The result does not inquire and satisfies condition, and slave board is lost synchrodata 50 to 99 according to chained list synchronously to the 90.0.1 request.
Step 3, mainboard 1 has received the request of the synchrodata of slave board, extracts synchrodata 50 to 99 from sync data file, and sends to slave board 90.0.3.
Step 4, slave board receives the synchrodata 50 to 99 of losing, and calculates up-to-date timestamp.
During the treatment step 1 to 4, slave board receives the most preceding synchrodata simultaneously, and is cached in the temporary file.After active and standby step-out was finished dealing with, slave board also will be checked the synchrodata in the temporary file, utilized more new database of temporary file, and then waited for the transmission of synchrodata.
The present invention has carried out application implementation on the database of embedded environment, reached its intended purposes.Realized that an equipment is that multiple devices carry out real-time data backup, stable, the data synchronization efficiency height does not need extra hardware device, has certain versatility.The present invention has strengthened the high reliability of system, and quick-recovery faulty board data soon when networking or single board default occurring realize active and standby switching, finishes from the fault location to the switching usually at tens of milliseconds between hundreds of milliseconds.
Claims (3)
1.N+1 the backed up data real-time synchronization method is characterized in that N 〉=1, employing one is equipped with N master's backup architecture, uses active and standby dynamic synchronized relation of synchronous chained list reflection, synchronously logical identifier, the socket information of chain table record slave board; Mainboard is data modification collection and the timestamp that Data Update derives with synchrodata according to synchronous chained list in real time one by one, and asynchronous transmission is to slave board, and storage file writes down synchrodata outside; The timestamp that identification data upgrades order uses the sequence number of synchrodata place file and the combination of side-play amount hereof to represent; Slave board is responsible for receiving isochronous data and is detected the integrality of synchrodata, according to timestamp Backup Data or active Backup Data after the synchrodata that the mainboard request is lost of synchrodata.
2. N+1 backed up data as claimed in claim 1 is synchronous method in real time, it is characterized in that active and standby plate takes following steps:
The step of taking on the mainboard comprises:
Synchronous chained list of step 1 mainboard initialization is used to write down integrated circuit board and attribute thereof;
Step 2 starts independently watcher thread synchronously, confirms synchronization request; Each database uses same watcher thread, and watcher thread is confirmed different data synchronization request by database-name;
Step 3 writes down logical identifier, the socket information of slave board in synchronous chained list, determine the forwarding position of synchrodata; Start independently Synchronous Processing thread, be responsible for the reception and the processing of synchrodata;
Step 4 is checked the backup object of synchronous chained list before the pre-submission of the renewal of mainboard data; If exist then execution in step 5;
Step 5 sends to slave board with synchrodata; If send failure, then nullifying the slave board object in the chained list synchronously; While, every synchrodata was safeguarded the timestamp of an identification renewal order; Connectionless protocol is used in data communication, guarantees that the synchrodata that the opposite end receives is correct, does not need to guarantee the order and the integrality of synchrodata;
Step 6 writes down synchrodata on local file, and submits Data Update to; Mainboard is the inactive sync data file of deletion regularly:
Step 7 circulation execution in step 4-6 is up to receiving active and standby switching or database close message;
The step that slave board is taked comprises:
Step 1 starts several synchronization request threads, and application is also determined to concern with the data sync of each mainboard;
Step 2 record mainboard and attribute thereof start independently Synchronous Processing thread, are responsible for the reception and the processing of synchrodata;
Step 3 receiving isochronous data is checked the timestamp of synchrodata, if equal current required timestamp, and new database more then; If greater than the timestamp of current required synchrodata, then in this synchrodata of temporary file buffer memory, check whether temporary file has required synchrodata, there is then more new database; If there is no then to the required synchrodata of mainboard request, and new database more; Except the review time stabs, also check the integrality of synchrodata, and to the synchrodata that receives according to timestamp ordering;
Step 4 is calculated the timestamp of up-to-date synchrodata, and checks whether temporary file has available synchrodata; There is then more new database, otherwise turns to step 3;
Step 5 circulation execution in step 3-4 is up to receiving active and standby switching or database close message.
3. N+1 backed up data as claimed in claim 1 or 2 is synchronous method in real time, it is characterized in that supporting active and standby asynchronous starting, specifically takes following steps:
Before step 1 database starts, check the state of integrated circuit board;
If the state that step 2 database will be set is for being equipped with execution in step 5;
Step 3 checks whether the state of this plate is main, otherwise wait reports an error up to overtime;
Step 4 setting data storehouse state is main, returns;
Step 5 checks whether the state of all associated panel is main; Otherwise wait for and reporting an error up to overtime;
Step 6 setting data storehouse state returns for being equipped with.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100182316A CN100397803C (en) | 2005-02-01 | 2005-02-01 | N+1 duplicates data real-time synchronising method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100182316A CN100397803C (en) | 2005-02-01 | 2005-02-01 | N+1 duplicates data real-time synchronising method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1677887A CN1677887A (en) | 2005-10-05 |
| CN100397803C true CN100397803C (en) | 2008-06-25 |
Family
ID=35050221
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2005100182316A Expired - Fee Related CN100397803C (en) | 2005-02-01 | 2005-02-01 | N+1 duplicates data real-time synchronising method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100397803C (en) |
Families Citing this family (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100357902C (en) * | 2005-12-20 | 2007-12-26 | 华为技术有限公司 | Data back-up method of plug-in card for business processing unit board |
| CN100465911C (en) * | 2006-04-13 | 2009-03-04 | 华为技术有限公司 | Backpu method |
| CN101197657B (en) * | 2007-11-12 | 2011-05-25 | 中兴通讯股份有限公司 | Data synchronization method of distributed system single panel |
| CN102447542B (en) * | 2010-10-09 | 2014-12-10 | 中兴通讯股份有限公司 | Difference self-recognizing method and system for configuration data of network equipment |
| WO2011116728A2 (en) * | 2011-04-29 | 2011-09-29 | 华为技术有限公司 | Data processing method and device for communication device |
| CN103164384B (en) * | 2011-12-15 | 2016-05-18 | 中国银联股份有限公司 | The synchronization realizing method of multi-computer system shared drive and system thereof |
| CN102567535A (en) * | 2011-12-30 | 2012-07-11 | 北京握奇数据系统有限公司 | Method and device for synchronizing file data in multi-file system smart card |
| CN104066107A (en) * | 2013-03-22 | 2014-09-24 | 中兴通讯股份有限公司 | Method and system for business disaster tolerance |
| CN103684720B (en) * | 2014-01-06 | 2017-12-19 | 迈普通信技术股份有限公司 | A kind of system of selection of active and standby service unit and device |
| CN105991301A (en) * | 2015-03-13 | 2016-10-05 | 杭州迪普科技有限公司 | Message synchronization method and device |
| CN107203433A (en) * | 2017-06-13 | 2017-09-26 | 微梦创科网络科技（中国）有限公司 | The multiple inter-program data synchronous method and device of a kind of shared drive |
| CN112242197A (en) * | 2019-07-17 | 2021-01-19 | 北京爱诺斯科技有限公司 | Vision monitoring method based on big data fitting |
| CN110719326B (en) * | 2019-10-08 | 2021-11-12 | 中国电子科技集团公司第二十八研究所 | Data synchronization method in high-availability software system adapting to low-quality communication link |
| CN111241200B (en) * | 2020-01-10 | 2024-02-20 | 浙江华创视讯科技有限公司 | Master-slave synchronous processing method and device based on SQLite database |
| CN111400263A (en) * | 2020-03-16 | 2020-07-10 | 上海英方软件股份有限公司 | Monitoring switch-back method and device based on file change |
| CN114416718A (en) * | 2022-03-28 | 2022-04-29 | 西安热工研究院有限公司 | Method, system, electronic device and storage medium for synchronizing historical data of database |
| CN115858688B (en) * | 2023-03-03 | 2023-04-18 | 北京阿玛西换热设备制造有限公司 | Remote data synchronization method and system for database |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1437348A (en) * | 2002-02-04 | 2003-08-20 | 深圳市中兴通讯股份有限公司 | Real-time synchronizing method for data in both main and spare board in communication system |
| US6625623B1 (en) * | 1999-12-16 | 2003-09-23 | Livevault Corporation | Systems and methods for backing up data files |
| CN1561022A (en) * | 2004-03-04 | 2005-01-05 | 中兴通讯股份有限公司 | Method of internal data base main-spare synchronous |
-
2005
- 2005-02-01 CN CNB2005100182316A patent/CN100397803C/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6625623B1 (en) * | 1999-12-16 | 2003-09-23 | Livevault Corporation | Systems and methods for backing up data files |
| CN1437348A (en) * | 2002-02-04 | 2003-08-20 | 深圳市中兴通讯股份有限公司 | Real-time synchronizing method for data in both main and spare board in communication system |
| CN1561022A (en) * | 2004-03-04 | 2005-01-05 | 中兴通讯股份有限公司 | Method of internal data base main-spare synchronous |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1677887A (en) | 2005-10-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100397803C (en) | N+1 duplicates data real-time synchronising method | |
| EP3493471B1 (en) | Data disaster recovery method, apparatus and system | |
| US7565572B2 (en) | Method for rolling back from snapshot with log | |
| US9916113B2 (en) | System and method for mirroring data | |
| US7694177B2 (en) | Method and system for resynchronizing data between a primary and mirror data storage system | |
| US6941327B2 (en) | Apparatus and method for database synchronization in a duplex system | |
| CN104486319B (en) | Suitable for the configuration file real-time synchronization method and its system of high-availability system | |
| CN106685691A (en) | Efficient configuration update method and system of terminal device in Internet of things system | |
| CN101751415B (en) | Metadata service system, metadata synchronized method and writing server updating method | |
| CN107623703B (en) | Synchronization method, device and system for Global Transaction Identifier (GTID) | |
| US7797571B2 (en) | System, method and circuit for mirroring data | |
| CN110601903A (en) | Data processing method and device based on message queue middleware | |
| CN102937955A (en) | Main memory database achieving method based on My structured query language (SQL) double storage engines | |
| CN111478796B (en) | Cluster capacity expansion exception handling method for AI platform | |
| US8862707B2 (en) | Method and apparatus for managing data of operation system | |
| EP1577776B1 (en) | Method and apparatus for data synchronization in a distributed data base system | |
| CN102045187B (en) | Method and equipment for realizing HA (high-availability) system with checkpoints | |
| CN110990190A (en) | Distributed file lock fault processing method, system, terminal and storage medium | |
| US20100274758A1 (en) | Data processing method, computer, and data processing program | |
| CN113722396B (en) | Method and equipment for switching main and standby services of data synchronous receiving end | |
| CN107357800A (en) | A kind of database High Availabitity zero loses solution method | |
| CN112783694B (en) | Long-distance disaster recovery method for high-availability Redis | |
| CN103049348B (en) | The method that under multiserver environment, data fault-tolerant stores | |
| WO2021115043A1 (en) | Distributed database system and data disaster backup drilling method | |
| EP2887592A1 (en) | Enum-dns disaster recovery method and system in ims network |
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 | ||
| C19 | Lapse of patent right due to non-payment of the annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |