CN104679614B - A kind of Database Disaster Backup/Recover - Google Patents
A kind of Database Disaster Backup/Recover Download PDFInfo
- Publication number
- CN104679614B CN104679614B CN201510147765.2A CN201510147765A CN104679614B CN 104679614 B CN104679614 B CN 104679614B CN 201510147765 A CN201510147765 A CN 201510147765A CN 104679614 B CN104679614 B CN 104679614B
- Authority
- CN
- China
- Prior art keywords
- server
- master server
- database
- snapshot
- data
- 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.)
- Active
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention relates to database field, more particularly to a kind of Database Disaster Backup/Recover.Including master server and the snapshot for including database from server, the master server and from server；During database normal operation, master server and from server be respectively separated preset time or according to instruction to database carry out snapshot；When master server breaks down, by being operated from server generation for master server；When master server recovers data, its data is first returned into the specified snapshot before failure occurs, then the data duplication not having in the specified snapshot to master server, master server is rejoined into cluster from from server；The present invention can be the time window backed up at any time using creating for not possessing BACKUP TIME window, and do not influence the normal operation of database and application, and can solve the problems, such as the generation of daily record branch, the present invention can also rapidly be integrated into current application environment and be used to test, and do not influence current system and continue to run with.
Description
Technical field
The present invention relates to database field, more particularly to a kind of Database Disaster Backup/Recover.
Background technology
Database is the core of all IT systems, saves the partial data run in operation system, and Database Systems are set
The journal recovery of database concerns important during standby collapse.At present, the calamity of reply Database Systems equipment collapse is mostly such as device
Shown in Fig. 1, by two server groups into master server primary, from server standalone, under normal circumstances by main clothes
Business device externally provides service, receives the operations such as application program Du ﹑ Xie ﹑ Geng Gai ﹑ deletions, and standalone provides read-only operation；Such as
Shown in Fig. 2, the data of master server are copied to from server in real time, and the data of master server are received from server, and its is extensive
Multiple to arrive out of database, the data that master server follows server have certain residual quantity t ' all the time.
And when master server is delayed machine, continue offer service from server, will occurrence log branch situation, that is, from
The daily record of server is not the newest daily record of database on master server, but master server is delayed, machine reception arrives last
Brought into operation at daily record, so causing daily record that branch occurs, as shown in Figure 3.
This can not be returned in cluster after resulting in master server reparation.
Existing solution method is the backup by recovering last time, it is returned to the time of last time backup
Point, because last time BACKUP TIME point is certainly than the time point of normal mode is switched to from server old (more earlier), institute
The database after backup can be added into cluster.But it is huge the time required to backup database, such as 20T data volume, according to
Main flow optical fiber 8Gb is calculated at present, 20T=20*1024*1024M=20971520M, 8Gb=8Gb/8=1GB=1024MB, by
In optical fiber packet header bag tail and transmission efficiency factor, the speed 1024*0.8=819.2MB really transmitted is also converted to, so
Recovery time ideally needs 20*1024*1024/819.2 about to need 7.1 hour recovery time.Such case for
It is unacceptable for key business.
The content of the invention
The problem of it is an object of the invention to overcome current Database Disaster Backup/Recover data recovery to need overlong time, there is provided
A kind of extremely short Database Disaster Backup/Recover of recovery time, including master server with from server, the master server and from service
Include the snapshot of database in device.
During database normal operation, master server and from server be respectively separated preset time or according to instruction to database
Carry out snapshot.
When master server breaks down, by being operated from server generation for master server.
When master server recovers data, its data is first returned into the specified snapshot before failure occurs, then from from service
The data duplication not having in the specified snapshot to master server, master server is rejoined into cluster in device.
Further, during database normal operation, from server data be master server in data by optical fiber or with
Too net replicates in real time.
Further, during database normal operation, master server externally receives reading and writing, changes and deletion action；From service
Device only externally provides read-only operation.
Further, when master server is not normally functioning, immediately by externally providing reading and writing from server, changing and delete
Division operation.
Preferably, the snapshot is to be stored using ZFS file system or vxfs file system.
Compared with prior art, beneficial effects of the present invention：
Due to taking snapping technique to be backed up for master and slave server, therefore the present invention can be by arbitrarily setting snapshot
Interval time is the application for not possessing BACKUP TIME window, creates the time window backed up at any time, and do not influence
Database and the normal operation of application.Simultaneously when database is recovered, by way of selecting to recover to specified snapshot
It is quick to solve the problems, such as the generation of daily record branch.The present invention can be achieved only to do smaller change (increase to traditional database calamity for equipment
Database snapshot) current application environment both can be rapidly integrated into for testing, and do not influence current system and continue to run with.
Brief description of the drawings：
Fig. 1 is traditional disaster recovery and backup systems schematic diagram.
Fig. 2 is traditional disaster recovery and backup systems principal and subordinate's server transaction data storage schematic diagram.
Fig. 3 be traditional disaster recovery and backup systems delay progress data recovery after machine when occurrence log branch schematic diagram.
Fig. 4 is Database Disaster Backup/Recover schematic diagram provided by the invention.
Principal and subordinate's snapshot server schematic diagram in Fig. 5 embodiment of the present invention 1.
Fig. 6 is that master server recovers schematic diagram data in the embodiment of the present invention 1.
Principal and subordinate's snapshot server schematic diagram in Fig. 7 embodiment of the present invention 2.
Fig. 8 is that master server recovers schematic diagram data in the embodiment of the present invention 2.
Embodiment
Below in conjunction with the accompanying drawings and specific embodiment the present invention is described in further detail.But this should not be interpreted as to this
The scope for inventing above-mentioned theme is only limitted to following embodiment, all to belong to the present invention based on the technology that present invention is realized
Scope.
Embodiment 1：It is an object of the invention to overcome current Database Disaster Backup/Recover data recovery to need overlong time
Problem, there is provided a kind of extremely short Database Disaster Backup/Recover of recovery time, as shown in figure 4, including master server with from server, institute
State master server and include the snapshot of database from server；(it is noted that in this area, host node, main service
Device refers both to same technological concept, similar, is also same technological concept from server, standby server, from node) master server
With being linked together from optional between server by optical fiber or Ethernet, the snapshot of database may be selected to use ZFS files
System or vxfs file system.
During database normal operation, data are that data are multiple in real time by optical fiber or Ethernet in master server from server
System；Now, master server externally receives reading and writing to database, changed and deletion action；Only externally provided from server read-only
Operation.
Master server and it is respectively separated preset time from server or itself data-base content is carried out according to instruction fast
According to, it is necessary to illustrate, being conducted due to data from master server to from server needs the regular hour, therefore with for the moment
Carve, master server with from server data certain residual quantity t ' (as shown in Figure 2) all the time be present, therefore, even if master server
With carrying out snapshot to its data in synchronization all the time from server, the snapshot contents of its synchronization also differ；Therefore,
Master server and the snapshot interval time from server, the snapshot of master server can be set according to being actually needed respectively completely
Interval time can be identical with the snapshot interval time from server, can also it is entirely different (i.e. master server may be selected with from
Server carries out snapshot in the same time all the time, also can carry out snapshot in different time).
In the present embodiment, master server and snapshot (i.e. main service is carried out respectively from server using identical interval time
Device and from server all the time the same time carry out snapshot), set server normal operation when, master server, from server 4
The snapshot of individual different time is respectively T1, T2, T3, T4 and T1 ', T2 ', T3 ', T4 ', now, master server snapshot as shown in Figure 5
With residual quantity t ' be present from snapshot server content, i.e., for snapshot T1 and T1 ', T2 and the T2 ' of synchronization, T3 and T3 ',
Residual quantity t ' be present in T4 and T4 '.When master server breaks down (machine of such as delaying) after snapshot T4, from a moment occurs by failure,
It is operated from server generation for master server；When i.e. master server is not normally functioning, immediately by externally being provided from server
Database reading and writing, change and deletion action, and still carry out snapshot according to former default interval time from server and produce fast
According to T5 ' (as shown in Figure 6).
And, it is necessary to rejoin cluster after master server reparation, now need first to carry out data recovery to main service, recover
Step is as follows：
(1) master server its data is first returned into the specified snapshot before failure occurs, because failure occurs in snapshot
After T4, any one in T1, T2, T3, T4 may be selected in theory and recovered, but because the Snapshot time that selection recovers more leans on
Before, then the data recovery time needed is longer, and select recover Snapshot time too close to fault time when, such as return to T4
When, due to the presence of data residual quantity t ' between T4 and T4 ', it is impossible to eliminate daily record branch problem.Therefore selected in the present embodiment extensive
Arrive snapshot T2 again, that is, when ensureing snapshot data content that selection recovers earlier than master server failure from server in data
Appearance.
(2) by master server snapshot data compared with the snapshot data from server, and will from server T2 it is fast
Master server is returned according to the data duplication after the moment, master server rejoins cluster.
It is mainly that master server returns to using the data recovery time needed for Database Disaster Backup/Recover provided by the invention
After specifying snapshot (being snapshot T2 in the present embodiment), master server replicates this from from server and specifies caused number after snapshot
According to；Compared with traditional disaster recovery and backup systems need whole database data carrying out duplication, after only replicating before failure sometime
Data obviously spend the time much less.
Embodiment 2：The present embodiment is with the difference of embodiment 1, master server with it is each pre- according to itself from server
If time carry out snapshot, as shown in fig. 7, within the same period, the snapshot that sets master server be respectively T1, T2, T3,
T4, T5, it is T1 ', T2 ', T3 ', T4 ' from the snapshot of server.
In the present embodiment, and now needed first to main service, it is necessary to when rejoining cluster after master server fault restoration
Data recovery is carried out, recovering step is as follows：
(1) master server its data is first returned into the specified snapshot before failure occurs, it is assumed that master server is in snapshot
Broken down after T5, to eliminate daily record branch problem, when ensureing the snapshot data content of selection recovery earlier than master server failure
The data content from server, therefore master server is recovered to snapshot T3 (as shown in Figure 8) in the present embodiment.
(2) by master server snapshot data compared with the snapshot data from server, and will from server T3 it is fast
Master server is returned according to the data duplication after the moment, master server rejoins cluster.
Claims (5)
- A kind of 1. Database Disaster Backup/Recover, it is characterised in that including master server with from server, the master server and from clothes Include the snapshot of database in business device；During database normal operation, master server and from server be respectively separated preset time or according to instruction to database carry out Snapshot；The snapshot interval time of master server is identical with the snapshot interval time from server, or different；When master server breaks down, by being operated from server generation for master server；When master server recovers data, its data is first returned into the specified snapshot before failure occurs, then from from server The data duplication not having in the specified snapshot to master server, master server is rejoined into cluster.
- 2. Database Disaster Backup/Recover as claimed in claim 1, it is characterised in that during database normal operation, from server Data are that data are replicated in real time by optical fiber or Ethernet in master server.
- 3. Database Disaster Backup/Recover as claimed in claim 1, it is characterised in that during database normal operation, master server pair It is outer to receive reading and writing, change and deletion action；Read-only operation is only externally provided from server.
- 4. Database Disaster Backup/Recover as claimed in claim 1, it is characterised in that when master server is not normally functioning, immediately By externally providing reading and writing from server, changing and deletion action.
- 5. Database Disaster Backup/Recover as claimed in claim 1, it is characterised in that the snapshot be using ZFS file system or Vxfs file system stores.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510147765.2A CN104679614B (en) | 2015-03-31 | 2015-03-31 | A kind of Database Disaster Backup/Recover |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510147765.2A CN104679614B (en) | 2015-03-31 | 2015-03-31 | A kind of Database Disaster Backup/Recover |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104679614A CN104679614A (en) | 2015-06-03 |
| CN104679614B true CN104679614B (en) | 2018-03-27 |
Family
ID=53314704
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510147765.2A Active CN104679614B (en) | 2015-03-31 | 2015-03-31 | A kind of Database Disaster Backup/Recover |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104679614B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105117165A (en) * | 2015-07-17 | 2015-12-02 | 浪潮(北京)电子信息产业有限公司 | Dual-controller snapshot taking-over and switching-back method and apparatus |
| CN109783277B (en) * | 2015-11-18 | 2023-12-29 | 华为技术有限公司 | Disaster recovery end, production end and data recovery method between disaster recovery end and production end |
| CN106919473B (en) * | 2015-12-28 | 2020-10-30 | 创新先进技术有限公司 | Data disaster backup system and service processing method |
| CN105955989B (en) * | 2015-12-31 | 2020-12-22 | 华云数据控股集团有限公司 | Method for establishing master server and slave server of cloud platform database |
| CN107015883B (en) * | 2016-01-28 | 2020-07-17 | 中国移动通信集团上海有限公司 | Dynamic data backup method and device |
| CN107657055A (en) * | 2017-10-18 | 2018-02-02 | 国网福建省电力有限公司 | A kind of back-up system method of data synchronization based on distribution main website |
| CN108416040B (en) * | 2018-03-14 | 2020-10-27 | 上海达梦数据库有限公司 | Database repair method and device, terminal equipment and storage medium |
| CN111209139A (en) * | 2019-12-29 | 2020-05-29 | 航天信息股份有限公司 | Disaster recovery method and device based on version control management system |
| CN112445780A (en) | 2020-11-27 | 2021-03-05 | 上海上讯信息技术股份有限公司 | Master-slave architecture deployment method and device based on snapshot |
| CN112286733B (en) * | 2020-12-23 | 2021-04-06 | 深圳市科力锐科技有限公司 | Method, device and equipment for determining backup data recovery time and storage medium |
Family Cites Families (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5101492A (en) * | 1989-11-03 | 1992-03-31 | Compaq Computer Corporation | Data redundancy and recovery protection |
| CN1642030A (en) * | 2004-01-05 | 2005-07-20 | 华为技术有限公司 | Method for realizing network management duoble computer disaster-tolerance back-up |
| CN100353331C (en) * | 2006-03-10 | 2007-12-05 | 四川大学 | Long-distance data fast restoring method of network information system |
| CN100591022C (en) * | 2007-11-29 | 2010-02-17 | 上海华为技术有限公司 | Method and apparatus for implementing preactivation of batch configuration |
| CN101668048B (en) * | 2009-09-28 | 2012-08-22 | 武汉大学 | Data protection method and system for supporting any time data access and modification |
| CN102710752B (en) * | 2012-05-13 | 2015-12-02 | 徐州易存信息技术有限公司 | Calamity is for storage system |
| CN103914359B (en) * | 2012-12-31 | 2017-09-26 | 中国移动通信集团浙江有限公司 | A kind of data reconstruction method and device |
| CN203149553U (en) * | 2013-03-28 | 2013-08-21 | 中国大唐集团财务有限公司 | Remote disaster recovery system with data safety verification function |
| CN103226502B (en) * | 2013-05-21 | 2015-08-19 | 中国工商银行股份有限公司 | A kind of data calamity is for control system and data reconstruction method |
-
2015
- 2015-03-31 CN CN201510147765.2A patent/CN104679614B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN104679614A (en) | 2015-06-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104679614B (en) | A kind of Database Disaster Backup/Recover | |
| CN101316184B (en) | Disaster tolerance switching method, system and device | |
| CN102891849B (en) | Service data synchronization method, data recovery method, data recovery device and network device | |
| CN110209726A (en) | Distributed experiment & measurement system system, method of data synchronization and storage medium | |
| CN105069160A (en) | Autonomous controllable database based high-availability method and architecture | |
| US20140258241A1 (en) | Zero and near-zero data loss database backup and recovery | |
| CN106815097A (en) | Database disaster tolerance system and method | |
| CN101739313B (en) | Method for protecting and restoring continuous data | |
| WO2018098972A1 (en) | Log recovery method, storage device and storage node | |
| CN105955836A (en) | Cold-hot backup automatic walkthrough multifunction system | |
| CN105302667B (en) | High reliability data backup and restoration methods based on aggregated structure | |
| CN109189860A (en) | A kind of active and standby increment synchronization method of MySQL based on Kubernetes system | |
| CN102214128A (en) | Repurposable recovery environment | |
| CN109032849A (en) | Hot backup system, heat backup method and computer equipment | |
| CN106484565A (en) | Method of data synchronization between multiple data centers and relevant device | |
| CN107229540A (en) | A kind of database restoring method and system based on time point | |
| CN101714169A (en) | Method and system for transmitting data on multiple nodes | |
| JP6511739B2 (en) | Redundant system and redundant method | |
| CN110209526A (en) | A kind of accumulation layer synchronization system and storage medium | |
| CN105956139A (en) | Method for synchronizing database in two machines | |
| CN105956207A (en) | Binlog-based configurable mysql database real-time synchronization method | |
| CN109347655B (en) | Network data based fault recovery system and method and storage medium | |
| WO2018076696A1 (en) | Data synchronization method and out-of-band management device | |
| CN106021020B (en) | A kind of dual-active data station application data hanging method and device | |
| CN112364104A (en) | Distributed database capacity expansion method, distributed database system and computer readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |