CN110069365A - Manage the method and corresponding device, computer readable storage medium of database - Google Patents
Manage the method and corresponding device, computer readable storage medium of database Download PDFInfo
- Publication number
- CN110069365A CN110069365A CN201910353959.6A CN201910353959A CN110069365A CN 110069365 A CN110069365 A CN 110069365A CN 201910353959 A CN201910353959 A CN 201910353959A CN 110069365 A CN110069365 A CN 110069365A
- Authority
- CN
- China
- Prior art keywords
- database
- management server
- primary
- primary database
- slave
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/14—Error detection or correction of the data by redundancy in operation
- G06F11/1402—Saving, restoring, recovering or retrying
- G06F11/1446—Point-in-time backing up or restoration of persistent data
- G06F11/1448—Management of the data involved in backup or backup restore
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/14—Error detection or correction of the data by redundancy in operation
- G06F11/1402—Saving, restoring, recovering or retrying
- G06F11/1446—Point-in-time backing up or restoration of persistent data
- G06F11/1458—Management of the backup or restore process
- G06F11/1464—Management of the backup or restore process for networked environments
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/14—Error detection or correction of the data by redundancy in operation
- G06F11/1402—Saving, restoring, recovering or retrying
- G06F11/1446—Point-in-time backing up or restoration of persistent data
- G06F11/1458—Management of the backup or restore process
- G06F11/1466—Management of the backup or restore process to make the backup process non-disruptive
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/27—Replication, distribution or synchronisation of data between databases or within a distributed database system; Distributed database system architectures therefor
Abstract
Present disclose provides a kind of methods and corresponding server, computer readable storage medium for managing database.This method comprises: receiving the running state information of primary database;The operating status of the primary database is determined according to the running state information of the primary database;When the operating status for determining the primary database is that the primary database breaks down, from least two select from database one from database as new primary database;And to sending the information about new primary database from the corresponding management server proxy module of database from other in database with described at least two respectively, by with it is described other from the corresponding management server proxy module control of database it is described other from database from new primary database Backup Data.
Description
Technical field
This disclosure relates to database field, and in particular, to a kind of method for managing database and corresponding device,
Computer readable storage medium.
Background technique
Infrastructure of the database as storing data, is widely used in the various products of internet area.For having
Data volume is big, products to characteristics such as response delay requirement height, business complexity, in order to support the load of product business, especially
Instantaneous superelevation load, it has been proposed that use data-base cluster.Data-base cluster may include a primary database and one
Or it is multiple from database, wherein primary database is responsible for the write-in of data, be responsible for the reading of data from database, and from database
It can be from primary database Backup Data (or being referred to as synchrodata or replicate data).In data-base cluster, work as master
When database breaks down, it can will be promoted to primary database from database, and to reduce the downtime of business, keep database
The high degree of availability of service.
In order to guarantee the data consistency of each database in data-base cluster, the O&M of data-base cluster is improved,
Some general data base management systems have been proposed.For example, including using extensive data base management system at present
Keepalived system, cloud Database Systems and primary database high-availability system (Master High Availability,
MHA).However, there are various defects for these existing data base management systems, for example, cannot support the extension from database with
Meet that the read-write demand of superelevation load, self-healing ability is not strong and can not manage that a large amount of data-base cluster, there are safety issues
It is unable to satisfy the conjunction rule requirement etc. of financial scenario.
Summary of the invention
For this purpose, the method and corresponding device, computer-readable storage that present disclose provides a kind of for managing database
Medium.
According to one aspect of the disclosure, a kind of method for managing database is provided.The described method includes: receiving
The running state information of primary database;The operation shape of the primary database is determined according to the running state information of the primary database
State;When the operating status for determining the primary database is that the primary database breaks down, from least two from database
Select one from database as new primary database;And to respectively with described at least two from other in database from number
Send the information about new primary database according to the corresponding management server proxy module in library, by with it is described other from database
The control of corresponding management server proxy module it is described other from database from new primary database Backup Data.
According to another aspect of the present disclosure, it provides a kind of for managing the device of database.Described device includes: to receive
Unit is configured as receiving the running state information of primary database;Determination unit is configured as the fortune according to the primary database
Row status information determines the operating status of the primary database;Selecting unit is configured as working as described in determination unit determination
The operating status of primary database is the primary database when breaking down, from least two select from database one from data
Library is as new primary database;And transmission unit, it is configured as to respectively with described at least two from other in database
Send the information about new primary database from the corresponding management server proxy module of database, by with it is described other from number
According to the control of the corresponding management server proxy module in library it is described other from database from new primary database Backup Data.
According to the disclosure example, receiving unit is configured as from management server corresponding with the primary database
Proxy module receives the running state information of the primary database.
According to the disclosure example, receiving unit be configured as from described at least two from database at least
One receives the running state information of the primary database from the corresponding management server proxy module of database.
According to the disclosure example, selecting unit is configured as determining candidate's from described at least two from database
From database;And selected from the slave database of the candidate one from database as new primary database.
According to the disclosure example, selecting unit be configured to determine that described at least two from database its affairs
Mark be managed server acquisition slave database, wherein the management server for manage the primary database and it is described extremely
Lack two from database;Whether it is greater than described at least two from the quantity of database from the quantity of database determined by judgement
Predetermined ratio;When it is identified from the quantity of database be greater than described at least two from the predetermined ratio of the quantity of database when,
It is determined as candidate slave database from database for identified;And selected from the slave database of the candidate one from number
According to library as new primary database.
According to the disclosure example, each there is its priority for serving as primary database from database;Selecting unit
It is configured as the priority according at least to each of slave database of the candidate from database, from the slave data of the candidate
Selected in library one from database as new primary database.
According to the disclosure example, selecting unit be configured as according to each of slave database of the candidate from
The priority and Transaction Identifier of database, selected from the slave database of the candidate one from database as new master data
Library.
According to the disclosure example, selecting unit is configured as having highest in the slave database for judging the candidate
Whether the slave database of priority has maximum transaction mark;And when the slave database with highest priority has most major issue
When business mark, using the slave database with highest priority as new primary database.
According to the disclosure example, selecting unit is configured as working as, and there is the slave database of highest priority not have
When maximum transaction identifies, pass through management server proxy module corresponding with the slave database with highest priority, control tool
There is the slave database of highest priority from the slave database back-up data identified with maximum transaction, so that having highest preferential
The slave database of grade is identified with maximum transaction;And using the slave database with highest priority as new primary database.
According to the disclosure example, transmission unit is configured as taking to management corresponding with the new primary database
Device proxy module of being engaged in sends configuration information, wherein the configuration information indicate the new primary database and it is described other from data
Replicanism between library, so that management server proxy module corresponding with the new primary database controls the new master
Database uses the replicanism.
According to the disclosure example, transmission unit is configured as to management server corresponding with the primary database
Proxy module sends restoration information, so that management server proxy module corresponding with the primary database repairs the master data
Library.
According to another aspect of the present disclosure, it provides a kind of for managing the device of database.The management server packet
Include processor;And memory, wherein computer-readable code, the computer-readable code are stored in the memory
When being run by the processor, the above method is executed.
According to another aspect of the present disclosure, a kind of computer readable storage medium is provided, instruction is stored thereon with, it is described
Instruction by processor when being run, so that the processor executes the above method.
By disclosed method and corresponding device, computer readable storage medium, when primary database breaks down,
One can be selected from database as new primary database, to ensure the High Availabitity of database service from multiple.In addition, logical
Disclosed method and corresponding device, computer readable storage medium are crossed, management server acts on behalf of mould by management server
Block operates database, avoids relying on SSH and operates to database, meets the compliance requirement of financial scenario.
Detailed description of the invention
The embodiment of the present disclosure is described in more detail in conjunction with the accompanying drawings, the above-mentioned and other purpose of the disclosure,
Feature and advantage will be apparent.Attached drawing is used to provide to further understand the embodiment of the present disclosure, and constitutes explanation
A part of book is used to explain the disclosure together with the embodiment of the present disclosure, does not constitute the limitation to the disclosure.In the accompanying drawings,
Identical reference label typically represents same parts or step.
Fig. 1 shows the configuration diagram of keepalived system.
Fig. 2 shows the configuration diagrams of cloud Database Systems.
Fig. 3 shows the configuration diagram of MHA system.
Fig. 4 A is the configuration diagram according to the data base management system of the embodiment of the present disclosure.
Fig. 4 B is the schematic diagram of the database in the data base management system according to the embodiment of the present disclosure.
Fig. 5 is the flow chart according to the method for the embodiment of the present disclosure executed by management server.
Fig. 6 is an illustrative detailed process of the framework according to shown in the application drawing 4A of the embodiment of the present disclosure.
Fig. 7 is the framework signal according to the realization management server High Availabitity and system library High Availabitity of the embodiment of the present disclosure
Figure.
Fig. 8 shows the structural schematic diagram of the management server according to the embodiment of the present disclosure.
Fig. 9 shows the framework of the computer equipment according to the embodiment of the present disclosure.
Specific embodiment
In order to enable the purposes, technical schemes and advantages of the disclosure become apparent, root is described in detail below with reference to accompanying drawings
According to the example embodiment of the disclosure.In the accompanying drawings, identical reference label indicates identical element from beginning to end.It is understood that
The embodiments described herein is merely illustrative, and is not necessarily to be construed as limiting the scope of the present disclosure.
Firstly, -3 describing existing keepalived system, cloud Database Systems and MHA system respectively referring to Fig.1.Figure
1 shows the configuration diagram of keepalived system.As shown in Figure 1, keepalived system includes a primary database 110
With one from database 120.When primary database 110 breaks down, new master data can be promoted to from database 120
Library.Since keepalived system is only supported one main one from can not be extended to from database, therefore, be not suitable for database
The application scenarios that example is more or database loads are more.In addition, in keepalived system, when primary database breaks down
When, primary database does not have any self-healing ability.
Cloud Database Systems have more advantage relative to keepalived system.Specifically, cloud Database Systems are one
Main one on the basis of, increases one or more read-only from database, is extended, provides to from database to realize
Better reading extended capability.In addition, the database self-healing component of a set of complexity is devised in cloud Database Systems, so as to
Make primary database self-healing when primary database breaks down.
Fig. 2 shows the configuration diagrams of cloud Database Systems.As shown in Fig. 2, cloud Database Systems include a main number
According to 210, one, library from database 220 and three it is read-only from database 230,240 and 250 (with dotted line and main number i.e. in Fig. 2
The database connected according to library 210).When primary database 210 breaks down, master data only can be promoted to from database 220
Library.In addition, when primary database 210 breaks down, primary database 210 can be with self-healing, and after self-healing, primary database 210
It can be used as from the slave database of database 220 and from from 220 Backup Data of database.
Although cloud Database Systems overcome the defect of keepalived system, it is substantially or one one master one
From database high-availability system.There are certain defects for cloud Database Systems.For example, if master data in cloud Database Systems
Library and from database simultaneously break down when, can not ensure the High Availabitity of database service.In addition, due to cloud Database Systems
Limitation is forced, therefore, the read-only limited amount from database in cloud Database Systems, thus can not be according to demand to from data
Library is extended.In addition, in cloud Database Systems, the read-only computer room locating for the database can not be customized, to be unable to satisfy
Individual demand of certain business to the geographical location of computer room.
Primary database high-availability system (Master High Availability, MHA) is the number of real one master and multiple slaves
According to library high-availability system.Fig. 3 shows the configuration diagram of existing MHA system.As shown in figure 3, MHA system includes a master
Database 310, four is from database 320,330,340 and 350.When primary database 310 breaks down, from database 320,
330, any one of 340 and 350 primary database can be promoted to.
However, MHA system is to be realized by script, and script can not support a series of complex of database self-healing to grasp
Make, therefore, the database in MHA system does not have self-healing ability.In addition, being configured in MHA system for managing database
Managing device.The managing device carries out database by safety shell protocol (Secure SHell protocol, SSH)
Operation.For example, the managing device can be by the primary database in SSH timing probe cluster, and works as and detect primary database
When breaking down, which can select one and to be promoted to primary database from database from multiple, and lead to
Cross SSH indicate it is remaining from database from new primary database Backup Data.However, SSH has defect in secure context, such as
SSH can suffer from " go-between " attack, therefore, be forbidden to use SSH under financial scenario.
There is a problem of in view of existing data base management system it is various, the present disclosure proposes one kind for one master it is more
From data base management system specifically propose a kind of method for managing the database in one master and multiple slaves and corresponding
Server, computer readable storage medium.By the data base management system of the disclosure, existing data base administration system is overcome
The defect of system.
In the following, describing the framework of the data base management system of the disclosure referring to Fig. 4 A-4B.Fig. 4 A is according to disclosure reality
Apply the configuration diagram of the data base management system of example.As shown in Figure 4 A, which includes management server
410, primary database 420, from database 430 to 450, management server proxy module 410-2 corresponding with primary database 420,
And it is acted on behalf of from the corresponding management server proxy module 410-3 of database 430, with from the corresponding management server of database 440
Module 410-4 and with from the corresponding management server proxy module 410-5 of database 450.
In the framework shown in Fig. 4 A, each database can be deployed at some node in a manner of application program, example
Such as, it is deployed on some server.For example, each database can be deployed in a server in a manner of existing MySQL
On.In addition, management server proxy module corresponding with each database can also be deployed in some in a manner of application program
At node.In addition, each database and management server proxy module corresponding with each database can be deployed in identical section
At point, it can also be deployed at different nodes.In addition, primary database 420 and can be deployed in each other from database 430 to 450
At different nodes.
In addition, management server 410 passes through management server proxy module 410-2 respectively in the framework shown in Fig. 4 A
It is operated to 410-5 to primary database 420 and from database 430 to 450, to realize to database 420 and from database 430
To 450 management.Specifically, management server 410 is used to control the operation to carry out to database.When management server 410
When operating to certain database, management server 410 can be to management server proxy module corresponding with the database
Send instruction, then, management server proxy module corresponding with the database according to the instruction received to the database into
Row operation.For example, management server proxy module corresponding with the database can according to the instruction received, pass through SQL language
Speech operates the database.Here " operation " may include the switching of database, modification, creation, deletion, installation, again
One or more of open, unload, starting duplication, setting replicanism etc..By framework shown in Fig. 4 A, when primary database is sent out
When raw failure, one can be selected from database as new primary database, to ensure that the height of database service can from multiple
With.In addition, when primary database breaks down, which has self-healing ability, so as to effective by this framework
Ground reduces manually-operated number, improves O&M efficiency.In addition, avoiding relying on SSH by this framework and being carried out to database
Operation meets the compliance requirement of financial scenario.
In addition, data base management system shown in Fig. 4 A can also include system library 410-1, so as to management server 410
Instance metadata is stored, thus the information of persisted instances.
In addition, data base management system shown in Fig. 4 A can also include that one or more is read-only from database, to realize
To the extension from database, extended capability is preferably read to provide.It includes one that Fig. 4 A, which shows the data base management system,
It is read-only from database, i.e., from database 460.In addition, the framework can also include with from the corresponding management server of database 460
Proxy module 410-6.
In addition, in the framework shown in Fig. 4 A, it is each to be located in a different geographical location from database.For example, from number
The first city can be located at according to library 430 to 450, the second city can be located at from database 460.Disposing in this way is advantageous in that,
When the business in the second city wants the data of reading database cluster, the slave database in the second city need to be only read, without reading
The slave database in the first city, across city access when so as to avoid business reading database reduce from the second city to
Across the city network delay in one city.
Fig. 4 B is the schematic diagram of the database in the data base management system according to the embodiment of the present disclosure.As shown in Figure 4 B,
The data base management system has five databases, respectively primary database 420, from database 430 to 450 and read-only from data
Library 460, wherein be located at the first city (as shown in the white box in Fig. 4 B) from database 430 to 450, it is read-only from database 460
Positioned at the second city (as shown in the grey box in Fig. 4 B).Primary database 420 is responsible for the write-in of data, from database 430,440,
450 and the read-only reading for being responsible for data from database 460, and from database 430,440,450 and it is read-only can from database 460
With from 420 Backup Data of primary database.When primary database 420 breaks down, from any in database 430,440 and 450
One can be promoted to primary database, and read-only cannot be promoted to primary database from database 460.For example, working as master data
When library 420 is broken down, it is promoted to primary database from database 440, then is responsible for the write-in of data from database 440, from number
According to library 430,450 and the read-only reading for being responsible for data from database 460, and from database 430,450 and read-only from database
460 can be from from 440 Backup Data of database.
It will be appreciated that although being illustrated in figures 4A and 4 B a management server, a system library, a master data
Library and four are from database, but this is only schematical.The framework may include more management servers, and/or more
System library, and/or more primary databases, and/or less or more from database.
The method of the execution of the management server as shown in Fig. 4 A is described below in conjunction with Fig. 5.Specifically, Fig. 5 will be combined
To describe the method that the master-slave database of the execution of the management server as shown in Fig. 4 A switches.Fig. 5 is according to the embodiment of the present disclosure
By management server execute method 500 flow chart.In method 500, management server is for managing a master data
Library and at least two is from database.Here " at least two from database " can not include read-only from database.For example, scheming
In the example of 4A, " at least two from database " may include from database 430 to 450.
As shown in figure 5, management server receives the running state information of primary database in step S501.Primary database
Running state information can indicate the operating status of primary database.The operating status of primary database can not sent out for primary database
Raw failure.Alternatively, the operating status of primary database can also break down for primary database.
According to the disclosure example, when deployment primary database and management server corresponding with primary database act on behalf of mould
When the node of block does not break down, management server can be received from management server proxy module corresponding with primary database and be led
The running state information of database.In this example, management server can be periodically from management corresponding with primary database
The running state information of server agent module reception primary database.Alternatively, management server can aperiodically from
Management server proxy module corresponding with primary database receives the running state information of primary database.
Primary database periodically is received from management server proxy module corresponding with primary database in management server
Running state information situation in, management server can periodically from management server corresponding with primary database act on behalf of
Module receives the heartbeat message of primary database, and wherein the heartbeat message of the primary database includes the operating status letter of primary database
Breath.
For example, the heartbeat mechanism between management server and primary database can be preset.It is main under the heartbeat mechanism
Database can be used as the client of management server and periodically report the heartbeat message of primary database to management server,
And the heartbeat message of the primary database includes the running state information of primary database.Specifically, primary database can be according to certainly
The operating status of body periodically generates the running state information of primary database, and includes by the running state information of primary database
In the heartbeat message of primary database.Then, management server proxy module corresponding with primary database can be to management service
The heartbeat message of device transmission primary database.Correspondingly, management server can be from management server generation corresponding with primary database
The heartbeat message that module receives primary database is managed, to receive the running state information of primary database.
Master data aperiodically is received from management server proxy module corresponding with primary database in management server
In the situation of the running state information in library, management server can be to management server proxy module wheel corresponding with primary database
Ask the running state information of primary database.
However, when the nodes break down of deployment primary database and management server proxy module corresponding with primary database
When, management server possibly can not receive the operation shape of primary database from management server proxy module corresponding with primary database
State information.Therefore, management server can preset waiting time (for example, 24 seconds), as can not be from when management server
Management server will wait when management server proxy module corresponding with primary database receives the running state information of primary database
To time.When management server can not still connect from management server proxy module corresponding with primary database within the waiting time
When receiving the running state information of primary database, deployment primary database and management server generation corresponding with primary database can be determined
Failure has occurred in the node of reason module, then management server can be believed by the operating status for receiving primary database from database
Breath.It is had received conversely, working as the management server within the waiting time from management server proxy module corresponding with primary database
When the running state information of primary database, management server need not be by the operating status for receiving primary database from database
Information.
Be described below management server by from database come receive primary database running state information example.Root
According to an example of the disclosure, management server can be from corresponding from least one of database from database at least two
Management server proxy module receive primary database running state information.In this example, management server can be with the period
Property receive main number from least one of database from the corresponding management server proxy module of database from at least two
According to the running state information in library.Alternatively, management server can aperiodically from at least two from database
At least one receives the running state information of primary database from the corresponding management server proxy module of database.
Management server periodically from at least two from least one of database from the corresponding pipe of database
Manage server agent module receive primary database running state information situation in, management server can periodically from
At least two receive at least one from number from least one of database from the corresponding management server proxy module of database
According to the heartbeat message in library, wherein this at least one from the heartbeat message of database include primary database running state information.
For example, management server can be preset and from the heartbeat mechanism between database.Under the heartbeat mechanism, from
Database can be used as the client of management server and periodically report heartbeat message from database to management server,
It and should include the running state information of primary database from the heartbeat message of database.It specifically, can be periodical from database
Ground detects the operating status of primary database, is then believed according to the operating status that the operating status of primary database generates primary database
Breath, and the running state information of primary database is included in from the heartbeat message of database.Then, corresponding with from database
Management server proxy module can send the heartbeat message from database to management server.Correspondingly, management server can
With from the heartbeat message from database is received from the corresponding management server proxy module of database, to receive primary database
Running state information.
In addition, from the heartbeat message of database can also include the running state information from database.Specifically, from data
Library can periodically generate the running state information from database according to the operating status of itself, and will be from the operation of database
Status information is included in from the heartbeat message of database.Then, with can from the corresponding management server proxy module of database
To send the heartbeat message from database to management server.Correspondingly, management server can be from corresponding with from database
Management server proxy module receives the heartbeat message from database, to receive the running state information from database.
In management server aperiodically from corresponding from least one of database from database at least two
Management server proxy module receives in the situation of the running state information of primary database, and management server can be to at least two
It is a from least one of database from the operating status of the corresponding management server proxy module poll primary database of database
Information.For example, management server can take to at least two from least one of database from the corresponding management of database
The running state information for device proxy module request primary database of being engaged in.Then, master can be detected in response to the request from database
Then the operating status of database generates bishop according to the running state information in library according to the operating status of primary database.Then, with
The running state information of primary database can be sent to management server from the corresponding management server proxy module of database.
Fig. 5 is returned to, in step S502, management server is according to the determination of the running state information of the primary database
The operating status of primary database.When the running state information of primary database instruction primary database does not break down, management service
Device can determine that the operating status of primary database is that primary database does not break down.Alternatively, when the operation shape of primary database
When state information instruction primary database breaks down, management server can determine the operating status of primary database for primary database hair
Raw failure.
Then, in step S503, when management server determines that the operating status of the primary database is the master data
When library is broken down, management server from described at least two select from database one from database as new master data
Library.For example, management server can determine candidate slave databases from least two from database, then from it is described it is candidate from
Selected in database one from database as new primary database.Here " candidate from database " can be one or more
It is a from database.
Be described below management server determined from least two from database one of candidate slave database it is exemplary
Process.
According to the disclosure example, firstly, management server can determine at least two from database its affairs
Identify the slave database obtained by the management server.Specifically, management server can preset acquisition Transaction Identifier
Time-out time (for example, 5 seconds).Management server can determine at least two from database its Transaction Identifier in the time-out
The interior slave database obtained by the management server.
For example, management server has got the affairs mark from database 430 in time-out time in the example of Fig. 4 A
Know and from the Transaction Identifier of database 440 without getting the Transaction Identifier from database 450, then management server can be with
Determine that slave database that its Transaction Identifier is obtained in the time-out time by the management server is from database 430 and from number
According to library 440.
" Transaction Identifier " described above may include the identification information for the affairs that primary database executed and from data
The identification information for the affairs that library executed.For example, Transaction Identifier can be global transaction identifier (Global Transaction
IDentifiers, GTID).Management server can determine thing that primary database executed according to from the Transaction Identifier of database
Business and the affairs executed from database, so that it is determined that should be from the consistency of the data of the data and primary database of database
(or being registration).For example, when the affairs that primary database executed are identical with the affairs executed from database, pipe
Reason server can determine that the data from database are identical as the data of primary database.When affairs that primary database executed and
When the affairs part executed from database is identical, management server can determine the data and primary database from database
Data portion is identical.
In addition, at least two from database its Transaction Identifier be not managed server acquisition slave database,
Management server can delete these from database, for example, these are kicked out of data-base cluster from database from data-base cluster.
For example, in the example of Fig. 4 A, management server got in time-out time from the Transaction Identifier of database 430 and from
The Transaction Identifier of database 440 is without getting the Transaction Identifier from database 450, then management server can will be from data
Kick out of data-base cluster in library 450.
Management server determined at least two from database its Transaction Identifier obtained by the management server
After database, management server may determine that whether the identified quantity from database is greater than described at least two from number
According to the predetermined ratio (for example, half) of the quantity in library.It is greater than described at least two from the quantity of database when identified
When from the predetermined ratio of the quantity of database, management server can be determined as candidate slave data from database for identified
Library.When it is identified from the quantity of database be less than described at least two from the predetermined ratio of the quantity of database when, management clothes
Business device will not be determined as candidate slave database from database for identified.
It from database is from database 430 and from database determined by management server for example, in the example of Fig. 4 A
440.From the quantity of database be 2 as determined by management server, at least two from the quantity of database be 3, and institute
The quantity of determining slave database is greater than the half of at least two quantity from database, then management server can will be from
Database 430 and the slave database for being determined as candidate from database 440.
It from database is from database 430 determined by management server in another example in the example of Fig. 4 A.Due to pipe
Managing determined by server from the quantity of database is 1, at least two from the quantity of database be 3, and it is identified from number
According to the quantity in library less than at least two from the half of the quantity of database, then management server will not will be from database 430
It is determined as candidate slave database.
After management server determines candidate slave database from least two from database, management server can be from
Selected in the slave database of the candidate one from database as new primary database.According to the disclosure example, often
It is a to can have its priority for serving as primary database from database.For example, management server can be in advance to each from data
Its priority for serving as primary database of lab setting.For example, management server can give in advance from database in the example of Fig. 4 A
430 its priority for serving as primary database of setting are 1, are 2 to its priority for serving as primary database is arranged from database 440,
And to from database 450 be arranged its priority for serving as primary database be 3.
In this example, management server can be according at least to each of slave database of the candidate from database
Priority, selected from the slave database of the candidate one from database as new primary database.For example, management server
It can be according to each of candidate slave database from the priority and Transaction Identifier of database, from the slave database of the candidate
It is middle select one from database as new primary database.
Management server is considered when selecting new primary database from both the priority of database and Transaction Identifier, is
It is necessary to.This is because, in the ideal case, the slave database with highest priority may have maximum transaction mark,
That is the registration of the data of the data and primary database of the slave database with highest priority is highest.Therefore, work as management
For server using the slave database with highest priority as when new primary database, other can be from highest from database
The slave database back-up data of priority, to guarantee the data consistency of data-base cluster.However, under non-ideal conditions,
Slave database with highest priority may not have maximum transaction and identify, i.e. the number of the slave database with highest priority
Registration according to the data with primary database is not highest.Therefore, when management server is by the slave number with highest priority
When according to library as new primary database, although other can be from the slave DB Backup number with highest priority from database
According to, but do not ensure that the data consistency of data-base cluster.Therefore, management server can be according to the slave number of the candidate
According to each of library from the priority and Transaction Identifier of database, selection has highest preferential from the slave database of the candidate
The slave database of grade is as new primary database, to guarantee the data consistency of data-base cluster.
Priority and affairs of the management server according to each of candidate slave database from database is described below
Mark, selects an exemplary flow from database as new primary database from the slave database of the candidate.
Management server may determine that whether the slave database in candidate slave database with highest priority has most
Big Transaction Identifier.According to the disclosure example, when there is the slave database with highest priority maximum transaction to identify,
Management server can be using the slave database with highest priority as new primary database.
For example, in the example of Fig. 4 A, it is candidate from database be from database 430 and from database 440, and from number
It is 1 according to the priority that primary database is served as in library 430, the priority for serving as primary database from database 440 is 2, from database 440
With maximum transaction identify, then management server can by from database 440 as new primary database.
According to another example of the disclosure, when the slave database with highest priority, which does not have maximum transaction, to be identified,
Management server can control the slave database with highest priority and identify with maximum transaction, then will have highest preferential
The slave database of grade is as new primary database, to guarantee the data consistency of data-base cluster.
Specifically, when the slave database with highest priority, which does not have maximum transaction, to be identified, management server can be with
By management server proxy module corresponding with the slave database with highest priority, control have highest priority from
Database is from the slave database back-up data identified with maximum transaction, so that the slave database with highest priority has
Maximum transaction mark.For example, management server can when the slave database with highest priority does not have maximum transaction and identifies
To send to management server proxy module corresponding with the slave database with highest priority about with maximum transaction mark
The information for the slave database known, so as to management server proxy module corresponding with the slave database with highest priority control
Slave database with highest priority is from the slave database back-up data identified with maximum transaction.When with highest priority
Slave database after the slave database back-up data identified with maximum transaction, with highest priority slave database energy
Enough there is maximum transaction mark, then management server is using the slave database with highest priority as new primary database.
For example, in the example of Fig. 4 A, it is candidate from database be from database 430 and from database 440, and from number
It is 1 according to the priority that primary database is served as in library 430, the priority for serving as primary database from database 440 is 2, from database 430
It is identified with maximum transaction, then management server can the corresponding management server proxy module 410-4 of Xiang Yucong database 440
It sends about the information from database 430, so that management server proxy module 410-4 control is from database 440 from from data
430 Backup Data of library.After from database 440 from from 430 Backup Data of database, there can be maximum from database 440
Transaction Identifier, then management server can be by the slave database 440 after Backup Data as new primary database.
In addition, as also described above, primary database is responsible for the write-in of data, it is responsible for the reading of data from database,
And it can be from primary database replicate data from database.In order to forbid that data are written to from database, management server can be with
Attribute (such as super read-only (super read only) option) is replicated to from data lab setting, is held with limitation to from database
Row write operation.For example, management server 410 can give from database 430 to 450 and be respectively provided with super in the example of Fig. 4 A
Read only option.
In the disclosure, when management server using some from database as when new primary database, can first remove
From the duplication attribute of database.For example, in the example of Fig. 4 A, when management server by from database 440 as new master
When database, the super read only option from database 440 can be first removed.
Return Fig. 5, in step S504, management server to respectively with described at least two from database other from
The corresponding management server proxy module of database sends the information about new primary database, by with it is described other from data
Library corresponding management server proxy module control it is described other from database from new primary database Backup Data.
For example, in the example of Fig. 4 A, management server by from database 440 as new primary database after, pipe
Managing server can the corresponding management server proxy module 410-3 of Xiang Yucong database 430, corresponding with from database 450
Management server proxy module 410-5 is sent about the information from database 440.Management server is acted on behalf of by management server
Module 410-3 can control from database 430 from from 440 Backup Data of database and management server proxy module 410-5
Control is from database 450 from from 440 Backup Data of database.
" information about new primary database " described above can be the identification information of new primary database, example
Such as, the static ip address (or fixed ip address) of new primary database.The quiet of new primary database can be passed through from database
State IP address, from new primary database Backup Data.For example, in the example of Fig. 4 A, management server will be from database 440
After new primary database, from database 430 and from database 450 can by the static ip address from database 440,
From from 440 Backup Data of database.
After step S504, method 500 can also include step S505: management server can to the new master
The corresponding management server proxy module of database sends configuration information, wherein the configuration information indicates the new master data
Library and it is described other from the replicanism between database, so as to management server corresponding with new primary database agency
Module controls the new primary database using the replicanism.
For example, in the example of Fig. 4 A, management server by from database 440 as new primary database after, pipe
Manage server can the corresponding management server proxy module 410-4 of Xiang Yucong database 440 send configuration information, wherein this is matched
The instruction of confidence breath as the slave database 440 of new primary database and from database 430, from the duplicator between database 450
System, the slave database 440 so as to management server proxy module 410-4 control as new primary database use the duplicator
System.
" replicanism " described above may include complete synchronous replicanism, asynchronous replication mechanism or partly synchronize multiple
Any one of making mechanism etc..In situation of the primary database using complete synchronous replicanism, primary database can according to
The write operation of family instruction execution data, and after having executed the write operation of data, it is not to return to confirm to user at once
Message, but wait it is all from database after primary database Backup Data again to user return confirmation message.In master data
Using in the situation of asynchronous replication mechanism, primary database can execute the write operation of data according to user instructions in library, and
After the write operation for having executed data, confirmation message is returned to user at once, it is standby from database from primary database without waiting
Confirmation message is returned to user again after part data.In situation of the primary database using half synchronous replicanism, primary database can
It is not at once to user to execute the write operation of data according to user instructions, and after having executed the write operation of data
Confirmation message is returned, but waits part from database (can be arranged by parameter ack count from the quantity of database) from master
Confirmation message is returned to user again after database back-up data.
In the disclosure, in new primary database and the feelings from the replicanism between database for half synchronous replicanism
In shape, new primary database can execute the write operation of data according to user instructions, and in the write-in behaviour for having executed data
After work, confirmation message is not returned to user not instead of at once, one or more is waited to back up from database from new primary database
Confirmation message is returned to user again after data, to guarantee the data consistency of new data-base cluster.It is synchronous by using half
Replicanism, though new data-base cluster is hit again, such as the primary database delay machine in new data-base cluster, but
Since the data of the slave database in new data-base cluster and primary database are consistent, in new data-base cluster
Slave database still ensure that the High Availabitity of database.Therefore, by using half synchronous replicanism, data base set can be assigned
The ability of the anti-two-hit of group.
In addition, in the disclosure, management server can be to each data base set group configuration dynamic IP (for example, floating
IP), externally to provide the unified IP of service using the dynamic IP as data-base cluster.User can be accessed by the dynamic IP
Data-base cluster, such as data are written to data-base cluster by the dynamic IP.In this case, after step s 505,
Method 500 can also include step S506: the dynamic IP of data-base cluster can be tied to new master data by management server
Library.In this way, data can be written to data-base cluster by dynamic IP in user, without knowing in data-base cluster
Which database be primary database, to improve the Experience Degree of user.
For example, in the example of Fig. 4 A, management server by from database 440 as new primary database before, pipe
The Floating IP address of data-base cluster can be tied to primary database 410 by reason server.It will make from database 440 in management server
After new primary database, in step S506, the Floating IP address of data-base cluster can be tied to from number by management server
According to library 440.
By the above-mentioned method executed by management server, after the primary database delay machine in data-base cluster, management clothes
Some can be promoted to new primary database from database by business device, so that data-base cluster, which reverts to service, can use shape
State provides service to continue as user.However, the present disclosure is not limited thereto.Management server can also repair the master data of delay machine
Library.According to the disclosure example, after step S506, method 500 can also include step S507: management server can
To send restoration information to management server proxy module corresponding with the primary database, so as to corresponding with the primary database
Management server proxy module repair the primary database.Here " restoration information " may include the instruction of restarting of database
And the information about new primary database.
For example, in the example of Fig. 4 A, after 420 delay machine of primary database, management server can to primary database
420 corresponding management server proxy module 410-2 send restoration information, so that management server proxy module 410-2 is controlled
Primary database 420 processed closes and restarts and control the primary database after restarting successfully 420 from new primary database Backup Data,
So that management server proxy module 410-2 repairs primary database 420.
In this example, data-base cluster can be added in the primary database after reparation by management server.For example, management clothes
Be engaged in device can by the primary database 420 after reparation as the slave database from database 440 (being promoted to primary database) and
Data-base cluster is added.Alternatively, management server can delete the primary database after reparation from data-base cluster.
In addition, in this example, management server can carry out the verification of data consistency to the primary database after reparation,
Quadratic search is carried out, to guarantee the data consistency of data-base cluster.In addition, management server can also be by database instance
It is set as available mode, in order to provide database service.
In addition, in the disclosure, management server can show notification information to the operator of management server, so as to pipe
The operator of reason server knows the operating status of data-base cluster.For example, in the example of Fig. 4 A, when primary database 420 is delayed
When machine, management server can be to operator's display alarm information of management server, so that the operator of management server obtains
Know 420 delay machine of primary database.In another example when the primary database delay machine in data-base cluster and management server can not be from data
When library is promoted to new primary database, management server can to operator's display alarm information of management server, so as to
Operator's notice of management server can not carry out master-slave swap, to set maintenance state for data-base cluster.In another example
In the example of Fig. 4 A, after 420 delay machine of primary database and after being repaired, management server can be to the operation of management server
Person's display alarm information, so that the operator of management server knows that primary database 420 has been repaired.
By the method for the embodiment of the present disclosure, when primary database breaks down, can be selected from database from multiple
One as new primary database, to ensure the High Availabitity of database service.In addition, by the method for the embodiment of the present disclosure, pipe
Reason server operates database by management server proxy module, avoids relying on SSH and operates to database,
Meet the compliance requirement of financial scenario.In addition, being realized by the method for the embodiment of the present disclosure to the expansion from database
Exhibition preferably reads extended capability to provide.
In the following, the illustrative detailed process that framework shown in application drawing 4A will be described in conjunction with Fig. 6.Fig. 6 is root
One illustrative detailed process of framework shown in the application drawing 4A according to the embodiment of the present disclosure.As shown in fig. 6, in step S1
In, management server by primary database and the heartbeat reported from database find primary database delay machine (one main three from feelings
It must be first to find primary database delay machine under condition).Then, in step s 2, management server is attempted from all from database
Obtain GTID.Then, in step s3, management server determines whether correctly to obtain the slave database of more than half
GTID, and the slave database for not obtaining its GTID correctly can be kicked out of into data-base cluster.Then, in step s 4, manage
Reason server judgement has whether the slave database of highest priority has maximum GTID.When the slave number with highest priority
According to library have maximum GTID when, in step s 5, management server by with highest priority slave database promoted based on
Database, and remove the super read only option from database.When the slave database with highest priority does not have
When maximum GTID, in step s 12, slave database of the management server control with highest priority, which connects, has maximum
The slave database of GTID executes step S5 with replicate data after the completion of duplication.Then, in step s 6, management server
It controls other and connects new primary database from database and replicated.Then, in the step s 7, management server can be by the height
Availability database switchs to synchronous duplication, and Floating IP address is tied to new primary database, so that database service restores available.
Then, in step s 8, management server can repair old primary database, specifically, pass through kill instruction and instruction of restarting
Restart old primary database, restarts old primary database successfully.Then, in step s 9, management server controls old master
Database connects new primary database with replicate data.Then, in step slo, if necessary, management server can be with
Modify ack count parameter.Then, in step s 11, management server can set available mode for database instance.
In addition, when management server, which fails, executes step S3~S7, management server can alarm in Fig. 6,
To set maintenance state for database.In addition, in Fig. 6, after management server executes step S9~S10, if data
Library cluster is one main one from then management server can alarm and set maintenance state, otherwise, management server for database
It can also alarm and kick out of old primary database and set maintenance state for old primary database after data cluster, to prevent two
Secondary strike.
Embodiments described above is primary database in management server management data-base cluster and/or from database
To realize the embodiment of database High Availabitity.However the present disclosure is not limited thereto.According to another embodiment of the present disclosure, it may be implemented
The High Availabitity of management server.In this embodiment it is possible to dispose multiple management servers.Although deploying multiple management services
Device, still, at a time, only one management server is used to manage primary database in data-base cluster and/or from data
Library.In the disclosure, it can be taken by the primary database in current time management data-base cluster and/or from the management of database
Business device is known as work management server.
According to the disclosure example, when work management server fail, other management servers can fill
When work management server.Specifically, the management module for managing multiple management servers can be disposed.Other management services
Device can serve as work management server by means of management module.Here " management module " can be existing
ZooKeeper application program.
For example, management module can select one from other management servers when work management server fail
It is a to serve as work management server.Specifically, management server can according to the mark of each management server (for example,
ID), one is selected to serve as work management server from multiple management servers.For example, management server can be from multiple
There is the management server of minimum ID to serve as work management server for selection in management server.In this example, work as work
When management server breaks down, management module can select the management server with minimum ID from remaining management server
To serve as work management server.
In another example a management service can not be selected by management module when work management server fail
Device serves as work management server, but other management servers are automatically attempted to serve as work management server.Specifically,
Management module can notify work management server fail to give other management servers, then, other management servers
It can attempt to serve as work management server.For example, management module can detecte the operating status of work management server, and
When work management server fail, which can notify work management server fail to manage to other
Server is managed, then, other management servers can be attempted to serve as work management server.Other management servers attempt at
When for work management server, each management server in other management servers can pass through respective identification information and determine
Whether itself can serve as work management server.For example, a management server in other management servers may determine that
Whether the ID of itself is less than the ID of the remaining management server in other management servers, if it is, the management server can
To serve as work management server.
By the embodiment, in work management server fail, work can be served as by another management server
Make management server, with continue manage data-base cluster in primary database and/or from database, to realize management service
The High Availabitity of device.
According to the another embodiment of the disclosure, the High Availabitity of system library may be implemented.In this embodiment it is possible to configure more
A system library.Although being configured with multiple system libraries, at a time, management server only uses in multiple system libraries
One system library.
According to the disclosure example, IP address can be configured to each system library in multiple system libraries.This
In situation, management server can select a system library to make according to the IP address of each system library from multiple system libraries
With.For example, management server can be selected from multiple system libraries the system library with minimum IP address come using.In the example
In, when the current system library that management server uses is broken down, management server can select tool from remaining system library
Have the system library of minimum IP address come using.
By the embodiment, when current system library breaks down, another system library is can be used in management server
Instance metadata is stored, to realize the High Availabitity of system library.
In the following, the configuration diagram that realization management server High Availabitity and system library High Availabitity will be described in conjunction with Fig. 7.Figure
7 be the configuration diagram of the realization management server High Availabitity and system library High Availabitity according to the embodiment of the present disclosure.Such as Fig. 7 institute
Show, five management servers, respectively management server 1 to 5 can be disposed, wherein management server 1 to 3 can be located at deep
Ditch between fields, management server 4 to 5 can be located at Shanghai.Furthermore, it is possible to dispose ZooKeeper in each management server.In addition,
Five system libraries, respectively metadatabase (mata batabase) 1 to 5 can be disposed.Furthermore, it is possible to dispose clothes to be managed
The data-base cluster of business device management, which includes database MySQL 1 to 6.Furthermore, it is possible to dispose and database
The corresponding management server proxy module Agent of MySQL 1 to 6.
As shown in fig. 7, management server 3 manages database MySQL 1 to 6 using system library 1.When management server 3
When breaking down, ZooKeeper can detecte management server 3 and break down, and management server 3 to that event can occur
The message informing of barrier is to management server 1,2,4 and 5.Then, a management server in management server 1,2,4 and 5 can
To take over management server 3, Lai Guanli database MySQL 1 to 6.In addition, when system library 1 breaks down, management server 3
In system library 2 to 5 system library can be used to manage database MySQL 1 to 6.
In addition, the operator of management server can pass through web application programming interface (Web in Fig. 7
Application Programming Interface, Web API) carry out operate administrative service device.
Hereinafter, being described referring to Fig. 8 according to the embodiment of the present disclosure for managing the device of database.The device can be
Management server.Fig. 8 shows the structural schematic diagram of the management server 800 according to the embodiment of the present disclosure.Due to management service
The function of device 800 is identical as the details of method described above with reference to Fig. 5, therefore herein for simplicity, omits
To the detailed description of identical content.As shown in figure 8, management server 800 includes: receiving unit 810, it is configured as receiving main number
According to the running state information in library;Determination unit 820 is configured as according to the determination of the running state information of the primary database
The operating status of primary database;Selecting unit 830 is configured as determining the operation shape of the primary database when the determination unit
State is the primary database when breaking down, from least two select from database one from database as new master data
Library;And transmission unit 840, be configured as to respectively with described at least two from database other are corresponding from database
Management server proxy module sends information about new primary database, by with it is described other from the corresponding management of database
Server agent module control it is described other from database from new primary database Backup Data.Other than this four units,
Management server 800 can also include other component, however, since these components are unrelated with the content of the embodiment of the present disclosure, because
This omits its diagram and description herein.Here " at least two from database " can not include read-only from database.For example,
In the example of Fig. 4 A, " at least two from database " may include from database 430 to 450.
In the disclosure, the running state information of primary database can indicate the operating status of primary database.Primary database
Operating status can not break down for primary database.Alternatively, the operating status of primary database can also be master data
It breaks down in library.
According to the disclosure example, when deployment primary database and management server corresponding with primary database act on behalf of mould
When the node of block does not break down, receiving unit 810 can be received from management server proxy module corresponding with primary database
The running state information of primary database.In this example, receiving unit 810 can be periodically from pipe corresponding with primary database
Manage the running state information that server agent module receives primary database.Alternatively, receiving unit 810 can be aperiodically
The running state information of primary database is received from management server proxy module corresponding with primary database.
Primary database periodically is received from management server proxy module corresponding with primary database in receiving unit 810
Running state information situation in, receiving unit 810 can be periodically from management server generation corresponding with primary database
The heartbeat message that module receives primary database is managed, wherein the heartbeat message of the primary database includes the operating status letter of primary database
Breath.
For example, the heartbeat mechanism between management server and primary database can be preset.It is main under the heartbeat mechanism
Database can be used as the client of management server and periodically report the heartbeat message of primary database to management server,
And the heartbeat message of the primary database includes the running state information of primary database.Specifically, primary database can be according to certainly
The operating status of body periodically generates the running state information of primary database, and includes by the running state information of primary database
In the heartbeat message of primary database.Then, management server proxy module corresponding with primary database can be to management service
The heartbeat message of device transmission primary database.Correspondingly, receiving unit 810 can be from management server generation corresponding with primary database
The heartbeat message that module receives primary database is managed, to receive the running state information of primary database.
Master data aperiodically is received from management server proxy module corresponding with primary database in receiving unit 810
In the situation of the running state information in library, receiving unit 810 can be to management server proxy module corresponding with primary database
The running state information of poll primary database.
However, when the nodes break down of deployment primary database and management server proxy module corresponding with primary database
When, receiving unit 810 possibly can not receive the operation of primary database from management server proxy module corresponding with primary database
Status information.Therefore, receiving unit 810 can preset waiting time (for example, 24 seconds), as when 810 nothing of receiving unit
Receiving unit 810 when method receives the running state information of primary database from management server proxy module corresponding with primary database
The time to be waited.When receiving unit 810 can not still be acted on behalf of from management server corresponding with primary database within the waiting time
When module receives the running state information of primary database, deployment primary database and management corresponding with primary database clothes can be determined
Failure has occurred in the node of business device proxy module, then receiving unit 810 can be by the fortune for receiving primary database from database
Row status information.Conversely, when within the waiting time receiving unit 810 from management server proxy module corresponding with primary database
When having received the running state information of primary database, receiving unit 810 need not receive primary database by from database
Running state information.
Be described below receiving unit 810 by from database come receive primary database running state information example.
According to the disclosure example, receiving unit 810 can from at least two from least one of database from database
Corresponding management server proxy module receives the running state information of primary database.In this example, receiving unit 810 can be with
Periodically received from at least two from least one of database from the corresponding management server proxy module of database
The running state information of primary database.Alternatively, receiving unit 810 can aperiodically from at least two from database
At least one of the running state information of primary database is received from the corresponding management server proxy module of database.
Receiving unit 810 periodically from at least two from least one of database from the corresponding pipe of database
Manage server agent module receive primary database running state information situation in, receiving unit 810 can periodically from
With at least two from least one of database from the corresponding management server proxy module of database receive at least one from
The heartbeat message of database, wherein this at least one from the heartbeat message of database include primary database running state information.
For example, management server can be preset and from the heartbeat mechanism between database.Under the heartbeat mechanism, from
Database can be used as the client of management server and periodically report heartbeat message from database to management server,
It and should include the running state information of primary database from the heartbeat message of database.It specifically, can be periodical from database
Ground detects the operating status of primary database, is then believed according to the operating status that the operating status of primary database generates primary database
Breath, and the running state information of primary database is included in from the heartbeat message of database.Then, corresponding with from database
Management server proxy module can send the heartbeat message from database to management server.Correspondingly, receiving unit 810 can
With from the heartbeat message from database is received from the corresponding management server proxy module of database, to receive primary database
Running state information.
In addition, from the heartbeat message of database can also include the running state information from database.Specifically, from data
Library can periodically generate the running state information from database according to the operating status of itself, and will be from the operation of database
Status information is included in from the heartbeat message of database.Then, with can from the corresponding management server proxy module of database
To send the heartbeat message from database to management server.Correspondingly, receiving unit 810 can be from corresponding with from database
Management server proxy module receives the heartbeat message from database, to receive the running state information from database.
In receiving unit 810 aperiodically from corresponding from least one of database from database at least two
Management server proxy module receive primary database running state information situation in, receiving unit 810 can to at least
Two from least one of database from the operation shape of the corresponding management server proxy module poll primary database of database
State information.For example, receiving unit 810 can to at least two from least one of database from the corresponding management of database
The running state information of server agent module request primary database.Then, it can be detected from database in response to the request
Then the operating status of primary database generates the running state information of primary database according to the operating status of primary database.Then,
Believe with the operating status that can send primary database to receiving unit 810 from the corresponding management server proxy module of database
Breath.
In addition, in the disclosure, when the running state information of primary database instruction primary database does not break down, determining
Unit 820 can determine that the operating status of primary database is that primary database does not break down.Alternatively, when the fortune of primary database
Row status information instruction primary database is when breaking down, and determination unit 820 can determine number based on the operating status of primary database
It breaks down according to library.
In addition, in the disclosure, when determination unit 820 determines that the operating status of the primary database is the primary database
When breaking down, selecting unit 830 from described at least two select from database one from database as new master data
Library.For example, selecting unit 830 can determine candidate slave database from least two from database, then from the candidate's
Selected from database one from database as new primary database.Here " candidate from database " can be one or
It is multiple from database.
Be described below selecting unit 830 determined from least two from database one of candidate slave database it is exemplary
Process.
According to the disclosure example, firstly, selecting unit 830 can determine at least two from database its affairs
Identify the slave database obtained by the management server.Specifically, selecting unit 830 can preset acquisition Transaction Identifier
Time-out time (for example, 5 seconds).Selecting unit 830 can determine at least two from database its Transaction Identifier in the time-out
The slave database obtained in time by the management server.
For example, management server has got the affairs mark from database 430 in time-out time in the example of Fig. 4 A
Know and from the Transaction Identifier of database 440 without getting the Transaction Identifier from database 450, then selecting unit 830 can
With determine the slave database that is obtained in the time-out time by the management server of its Transaction Identifier be from database 430 and from
Database 440.
" Transaction Identifier " described above may include the identification information for the affairs that primary database executed and from data
The identification information for the affairs that library executed.For example, Transaction Identifier can be global transaction identifier (Global Transaction
IDentifiers, GTID).Management server can determine thing that primary database executed according to from the Transaction Identifier of database
Business and the affairs executed from database, so that it is determined that should be from the consistency of the data of the data and primary database of database
(or being registration).For example, when the affairs that primary database executed are identical with the affairs executed from database, pipe
Reason server can determine that the data from database are identical as the data of primary database.When affairs that primary database executed and
When the affairs part executed from database is identical, management server can determine the data and primary database from database
Data portion is identical.
In addition, at least two from database its Transaction Identifier be not managed server acquisition slave database,
Management server can delete these from database, for example, these are kicked out of data-base cluster from database from data-base cluster.
For example, in the example of Fig. 4 A, management server got in time-out time from the Transaction Identifier of database 430 and from
The Transaction Identifier of database 440 is without getting the Transaction Identifier from database 450, then selecting unit 830 can will be from number
Data-base cluster is kicked out of according to library 450.
Selecting unit 830 determined at least two from database its Transaction Identifier obtained by the management server
After database, selecting unit 830 may determine that the identified quantity from database whether be greater than described at least two from
The predetermined ratio (for example, half) of the quantity of database.It is greater than described at least two from the quantity of database when identified
It is a from the predetermined ratio of the quantity of database when, selecting unit 830 can by it is identified from database be determined as it is candidate from
Database.When it is identified from the quantity of database be less than described at least two from the predetermined ratio of the quantity of database when, choosing
Candidate slave database will not be determined as from database for identified by selecting unit 830.
It from database is from database 430 and from database determined by management server for example, in the example of Fig. 4 A
440.From the quantity of database be 2 as determined by management server, at least two from the quantity of database be 3, and institute
The quantity of determining slave database is greater than the half of at least two quantity from database, then selecting unit 830 can incite somebody to action
It is determined as candidate slave database from database 430 and from database 440.
It from database is from database 430 determined by management server in another example in the example of Fig. 4 A.Due to pipe
Managing determined by server from the quantity of database is 1, at least two from the quantity of database be 3, and it is identified from number
According to the quantity in library less than at least two from the half of the quantity of database, then selecting unit 830 will not will be from database 430
It is determined as candidate slave database.
After selecting unit 830 determines candidate slave database from least two from database, selecting unit 830 can be with
Selected from the slave database of the candidate one from database as new primary database.According to the disclosure example,
Each it can have its priority for serving as primary database from database.For example, management server can be in advance to each from number
Its priority for serving as primary database according to lab setting.For example, management server can give in advance from data in the example of Fig. 4 A
It is 1 that its priority for serving as primary database, which is arranged, in library 430, is to its priority for serving as primary database is arranged from database 440
2, and to from database 450 be arranged its priority for serving as primary database be 3.
In this example, selecting unit 830 can be according at least to each of slave database of the candidate from database
Priority, selected from the slave database of the candidate one from database as new primary database.For example, selecting unit
830 can be according to each of candidate slave database from the priority and Transaction Identifier of database, from the slave number of the candidate
According to selected in library one from database as new primary database.
Management server is considered when selecting new primary database from both the priority of database and Transaction Identifier, is
It is necessary to.This is because, in the ideal case, the slave database with highest priority may have maximum transaction mark,
That is the registration of the data of the data and primary database of the slave database with highest priority is highest.Therefore, work as management
For server using the slave database with highest priority as when new primary database, other can be from highest from database
The slave database back-up data of priority, to guarantee the data consistency of data-base cluster.However, under non-ideal conditions,
Slave database with highest priority may not have maximum transaction and identify, i.e. the number of the slave database with highest priority
Registration according to the data with primary database is not highest.Therefore, when management server is by the slave number with highest priority
When according to library as new primary database, although other can be from the slave DB Backup number with highest priority from database
According to, but do not ensure that the data consistency of data-base cluster.Therefore, management server can be according to the slave number of the candidate
According to each of library from the priority and Transaction Identifier of database, selection has highest preferential from the slave database of the candidate
The slave database of grade is as new primary database, to guarantee the data consistency of data-base cluster.
Priority and affairs of the selecting unit 830 according to each of candidate slave database from database is described below
Mark, selects an exemplary flow from database as new primary database from the slave database of the candidate.
Selecting unit 830 may determine that whether the slave database in candidate slave database with highest priority has most
Big Transaction Identifier.According to the disclosure example, when there is the slave database with highest priority maximum transaction to identify,
Selecting unit 830 can be using the slave database with highest priority as new primary database.
For example, in the example of Fig. 4 A, it is candidate from database be from database 430 and from database 440, and from number
It is 1 according to the priority that primary database is served as in library 430, the priority for serving as primary database from database 440 is 2, from database 440
With maximum transaction identify, then selecting unit 830 can by from database 440 as new primary database.
According to another example of the disclosure, when the slave database with highest priority, which does not have maximum transaction, to be identified,
Selecting unit 830 can control the slave database with highest priority and identify with maximum transaction, then will have highest excellent
The slave database of first grade is as new primary database, to guarantee the data consistency of data-base cluster.
Specifically, when the slave database with highest priority, which does not have maximum transaction, to be identified, selecting unit 830 can be with
By management server proxy module corresponding with the slave database with highest priority, control have highest priority from
Database is from the slave database back-up data identified with maximum transaction, so that the slave database with highest priority has
Maximum transaction mark.For example, when the slave database with highest priority does not have maximum transaction and identifies, selecting unit 830
It can send to management server proxy module corresponding with the slave database with highest priority about with maximum transaction
The information of the slave database of mark, so as to management server proxy module control corresponding with the slave database with highest priority
Fixture has the slave database of highest priority from the slave database back-up data identified with maximum transaction.When preferential with highest
The slave database of grade is after the slave database back-up data identified with maximum transaction, the slave database with highest priority
There can be maximum transaction mark, then selecting unit 830 is using the slave database with highest priority as new primary database.
For example, in the example of Fig. 4 A, it is candidate from database be from database 430 and from database 440, and from number
It is 1 according to the priority that primary database is served as in library 430, the priority for serving as primary database from database 440 is 2, from database 430
It is identified with maximum transaction, then management server can the corresponding management server proxy module 410-4 of Xiang Yucong database 440
It sends about the information from database 430, so that management server proxy module 410-4 control is from database 440 from from data
430 Backup Data of library.After from database 440 from from 430 Backup Data of database, there can be maximum from database 440
Transaction Identifier, then selecting unit 830 can be by the slave database 440 after Backup Data as new primary database.
In addition, as also described above, primary database is responsible for the write-in of data, it is responsible for the reading of data from database,
And it can be from primary database replicate data from database.In order to forbid that data are written to from database, management server can be with
Attribute (such as super read-only (super read only) option) is replicated to from data lab setting, is held with limitation to from database
Row write operation.For example, management server 410 can give from database 430 to 450 and be respectively provided with super in the example of Fig. 4 A
Read only option.
In the disclosure, when selecting unit 830 using some from database as when new primary database, can first remove
From the duplication attribute of database.For example, in the example of Fig. 4 A, when selecting unit 830 by from database 440 as new master
When database, the super read only option from database 440 can be first removed.
In addition, in the disclosure, transmission unit 840 is to respectively with described at least two from other in database from data
The corresponding management server proxy module in library sends the information about new primary database, by with it is described other from database pair
The management server proxy module control answered it is described other from database from new primary database Backup Data.
For example, in the example of Fig. 4 A, management server by from database 440 as new primary database after, hair
Send unit 840 can the corresponding management server proxy module 410-3 of Xiang Yucong database 430, corresponding with from database 450
Management server proxy module 410-5 is sent about the information from database 440.Management server can pass through management server
Proxy module 410-3 control is from database 430 from from 440 Backup Data of database and management server proxy module 410-5
Control is from database 450 from from 440 Backup Data of database.
" information about new primary database " described above can be the identification information of new primary database, example
Such as, the static ip address (or fixed ip address) of new primary database.The quiet of new primary database can be passed through from database
State IP address, from new primary database Backup Data.For example, in the example of Fig. 4 A, management server will be from database 440
After new primary database, from database 430 and from database 450 can by the static ip address from database 440,
From from 440 Backup Data of database.
In addition, transmission unit 840 can be sent to management server proxy module corresponding with the new primary database
Configuration information, wherein the configuration information indicate the new primary database and it is described other from the duplicator between database
System, so that management server proxy module corresponding with the new primary database controls described in the new primary database use
Replicanism.
For example, in the example of Fig. 4 A, management server by from database 440 as new primary database after, hair
Send unit 840 can the corresponding management server proxy module 410-4 of Xiang Yucong database 440 send configuration information, wherein should
Configuration information indicate as the slave database 440 of new primary database and from database 430, from the duplication between database 450
Mechanism, the slave database 440 so as to management server proxy module 410-4 control as new primary database use the duplicator
System.
" replicanism " described above may include complete synchronous replicanism, asynchronous replication mechanism or partly synchronize multiple
Any one of making mechanism etc..In situation of the primary database using complete synchronous replicanism, primary database can according to
The write operation of family instruction execution data, and after having executed the write operation of data, it is not to return to confirm to user at once
Message, but wait it is all from database after primary database Backup Data again to user return confirmation message.In master data
Using in the situation of asynchronous replication mechanism, primary database can execute the write operation of data according to user instructions in library, and
After the write operation for having executed data, confirmation message is returned to user at once, it is standby from database from primary database without waiting
Confirmation message is returned to user again after part data.In situation of the primary database using half synchronous replicanism, primary database can
It is not at once to user to execute the write operation of data according to user instructions, and after having executed the write operation of data
Confirmation message is returned, but waits part from database (can be arranged by parameter ack count from the quantity of database) from master
Confirmation message is returned to user again after database back-up data.
In the disclosure, in new primary database and the feelings from the replicanism between database for half synchronous replicanism
In shape, new primary database can execute the write operation of data according to user instructions, and in the write-in behaviour for having executed data
After work, confirmation message is not returned to user not instead of at once, one or more is waited to back up from database from new primary database
Confirmation message is returned to user again after data, to guarantee the data consistency of new data-base cluster.It is synchronous by using half
Replicanism, though new data-base cluster is hit again, such as the primary database delay machine in new data-base cluster, but
Since the data of the slave database in new data-base cluster and primary database are consistent, in new data-base cluster
Slave database still ensure that the High Availabitity of database.Therefore, by using half synchronous replicanism, data base set can be assigned
The ability of the anti-two-hit of group.
In addition, in the disclosure, management server can be to each data base set group configuration dynamic IP (for example, floating
IP), externally to provide the unified IP of service using the dynamic IP as data-base cluster.User can be accessed by the dynamic IP
Data-base cluster, such as data are written to data-base cluster by the dynamic IP.In this case, management server can incite somebody to action
The dynamic IP of data-base cluster is tied to new primary database.In this way, user can be by dynamic IP to database
Data are written in cluster, without knowing which database in data-base cluster is primary database, to improve the body of user
Degree of testing.
For example, in the example of Fig. 4 A, management server by from database 440 as new primary database before, pipe
The Floating IP address of data-base cluster can be tied to primary database 410 by reason server.It will make from database 440 in management server
After new primary database, in step S506, the Floating IP address of data-base cluster can be tied to from number by management server
According to library 440.
By the above-mentioned method executed by management server, after the primary database delay machine in data-base cluster, management clothes
Some can be promoted to new primary database from database by business device, so that data-base cluster, which reverts to service, can use shape
State provides service to continue as user.However, the present disclosure is not limited thereto.Management server can also repair the master data of delay machine
Library.According to the disclosure example, transmission unit 840 can act on behalf of mould to management server corresponding with the primary database
Block sends restoration information, so that management server proxy module corresponding with the primary database repairs the primary database.This
In " restoration information " may include the instruction of restarting of database and the information about new primary database.
For example, in the example of Fig. 4 A, after 420 delay machine of primary database, management server can to primary database
420 corresponding management server proxy module 410-2 send restoration information, so that management server proxy module 410-2 is controlled
Primary database 420 processed closes and restarts and control the primary database after restarting successfully 420 from new primary database Backup Data,
So that management server proxy module 410-2 repairs primary database 420.
In this example, data-base cluster can be added in the primary database after reparation by management server.For example, management clothes
Be engaged in device can by the primary database 420 after reparation as the slave database from database 440 (being promoted to primary database) and
Data-base cluster is added.Alternatively, management server can delete the primary database after reparation from data-base cluster.
In addition, in this example, management server can carry out the verification of data consistency to the primary database after reparation,
Quadratic search is carried out, to guarantee the data consistency of data-base cluster.In addition, management server can also be by database instance
It is set as available mode, in order to provide database service.
In addition, in the disclosure, management server can also include notification unit (not shown), with to management service
The operator of device shows notification information, so that the operator of management server knows the operating status of data-base cluster.For example,
In the example of Fig. 4 A, when 420 delay machine of primary database, notification unit can show alarm signal to the operator of management server
Breath, so that the operator of management server knows 420 delay machine of primary database.In another example when the primary database in data-base cluster
Delay machine and when management server can not be new primary database be promoted to from database, notification unit can be to management server
Operator's display alarm information, so that master-slave swap can not be carried out to the operator of management server notice, thus by database
Cluster is set as maintenance state.In another example in the example of Fig. 4 A, after 420 delay machine of primary database and after being repaired, notification sheet
Member can be to operator's display alarm information of management server, so that the operator of management server knows primary database 420
It has been repaired.
It, can be from multiple from database when primary database breaks down by the management server of the embodiment of the present disclosure
It is middle to select one as new primary database, to ensure the High Availabitity of database service.In addition, the side for passing through the embodiment of the present disclosure
Method, management server operate database by management server proxy module, avoid relying on SSH and carry out to database
Operation meets the compliance requirement of financial scenario.In addition, by the method for the embodiment of the present disclosure, realize to from database
Extension, thus provide preferably read extended capability.
In addition, can also be by means of the framework shown in Fig. 9 for calculating equipment according to the management server of the embodiment of the present disclosure
To realize.Fig. 9 shows the framework of the calculating equipment.As shown in figure 9, calculate equipment 900 may include bus 910, one or
Multiple CPU 920, read-only memory (ROM) 930, random access memory (RAM) 940, the communication port for being connected to network
950, input output assembly 960, hard disk 970 etc..The storage equipment in equipment 900 is calculated, such as ROM 930 or hard disk 970 can
To store computer disposal and/or communicate program instruction performed by the various data used or file and CPU.Calculate equipment
900 can also include user interface 980.Certainly, framework shown in Fig. 9 is only exemplary, when realizing different equipment, root
According to actual needs, it is convenient to omit one or more components in calculating equipment shown in Fig. 9.
Embodiment of the disclosure also may be implemented as computer readable storage medium.According to the calculating of the embodiment of the present disclosure
Computer-readable instruction is stored on machine readable storage medium storing program for executing.It, can be with when the computer-readable instruction is run by processor
Execute the method according to the embodiment of the present disclosure referring to the figures above description.The computer readable storage medium includes but unlimited
In such as volatile memory and/or nonvolatile memory.The volatile memory for example may include that arbitrary access is deposited
Reservoir (RAM) and/or cache memory (cache) etc..The nonvolatile memory for example may include read-only storage
Device (ROM), hard disk, flash memory etc..
It will be appreciated by those skilled in the art that a variety of variations and modifications can occur in content disclosed by the disclosure.For example,
Various equipment described above or component can also pass through one in software, firmware or three by hardware realization
A little or whole combinations is realized.
In addition, as shown in the disclosure and claims, unless context clearly prompts exceptional situation, " one ", " one
It is a ", the words such as "an" and/or "the" not refer in particular to odd number, may also comprise plural number." first ", " second " used in the disclosure
And similar word is not offered as any sequence, quantity or importance, and be used only to distinguish different component parts.Together
The similar word such as sample, " comprising " or "comprising" means to occur after element or object before the word cover and appear in the word
The element that face is enumerated perhaps object and its equivalent and be not excluded for other elements or object." connection " or " connected " etc. are similar
Word be not limited to physics or mechanical connection, but may include electrical connection, either it is direct still
Indirectly.
In addition, flow chart has been used to be used to illustrate behaviour performed by system according to an embodiment of the present disclosure in the disclosure
Make.It should be understood that front or following operate not necessarily accurately carry out in sequence.On the contrary, can according to inverted order or
Various steps are handled simultaneously.It is also possible to during other operations are added to these, or from these processes remove a certain step
Or number step operation.
Unless otherwise defined, all terms (including technical and scientific term) used herein have leads with belonging to the present invention
The identical meanings that the those of ordinary skill in domain is commonly understood by.It is also understood that those of definition term such as in usual dictionary
The meaning consistent with their meanings in the context of the relevant technologies should be interpreted as having, without application idealization or
The meaning of extremely formalization explains, unless being clearly defined herein.
The disclosure is described in detail above, but it will be apparent to a person skilled in the art that the disclosure not limits
The fixed embodiment illustrated in this manual.The disclosure is not departing from the disclosure determined by the record of claims
Under the premise of objective and range, modifications and changes mode can be used as to implement.Therefore, the record of this specification is said with example
For the purpose of bright, for purposes of this disclosure not with the meaning of any restrictions.
Claims (15)
1. a kind of method for managing database, which comprises
Receive the running state information of primary database;
The operating status of the primary database is determined according to the running state information of the primary database;
When the operating status for determining the primary database is that the primary database breaks down, from least two from database
Select one from database as new primary database;And
To being sent respectively from other in database from the corresponding management server proxy module of database with described at least two
About the information of new primary database, by with it is described other from described in the corresponding management server proxy module control of database
Other are from database from new primary database Backup Data.
2. the method as described in claim 1, wherein the running state information for receiving primary database includes:
The running state information of the primary database is received from management server proxy module corresponding with the primary database.
3. the method as described in claim 1, wherein the running state information for receiving primary database includes:
It is received from described at least two from least one of database from the corresponding management server proxy module of database
The running state information of the primary database.
4. method as described in any one of claims 1 to 3, wherein it is described selected from database from least two one from number
Include: as new primary database according to library
Determine described at least two from database its Transaction Identifier be managed server acquisition slave database, wherein the pipe
Reason server is for managing the primary database and described at least two from database;
The identified predetermined ratio for whether being greater than described at least two quantity from database from the quantity of database of judgement;
When it is identified from the quantity of database be greater than described at least two from the predetermined ratio of the quantity of database when, by really
Fixed slave database is determined as candidate slave database;And
Selected from the slave database of the candidate one from database as new primary database.
5. method as claimed in claim 4, wherein
Each there is its priority for serving as primary database from database;
One is selected from the slave database of the candidate includes: as new primary database from database
Each of slave database according at least to the candidate is from the priority of database, from the slave database of the candidate
Select one from database as new primary database.
6. method as claimed in claim 5, wherein it is described according at least to each of slave database of the candidate from data
The priority in library, one is selected from the slave database of the candidate includes: as new primary database from database
Each of slave database according to the candidate is from the priority and Transaction Identifier of database, from the slave number of the candidate
According to selected in library one from database as new primary database.
7. method as claimed in claim 6, wherein according to each of slave database of the candidate from the preferential of database
Grade and Transaction Identifier, one is selected from the slave database of the candidate includes: as new primary database from database
Judge whether the slave database in the slave database of the candidate with highest priority has maximum transaction mark;And
When there is the slave database with highest priority maximum transaction to identify, the slave database with highest priority is made
For new primary database.
8. the method for claim 7, wherein according to each of slave database of the candidate from the preferential of database
Grade and Transaction Identifier, one is selected from the slave database of the candidate includes: as new primary database from database
When the slave database with highest priority do not have maximum transaction identify when, by with the slave number with highest priority
According to the corresponding management server proxy module in library, the slave database with highest priority is controlled from identifying with maximum transaction
From database back-up data, so that the slave database with highest priority is identified with maximum transaction;And
Using the slave database with highest priority as new primary database.
9. method as described in any one of claims 1 to 3, further includes:
Configuration information is sent to management server proxy module corresponding with the new primary database, wherein the configuration information
Indicate the new primary database and it is described other from the replicanism between database, so as to the new primary database pair
The management server proxy module answered controls the new primary database using the replicanism.
10. method as described in any one of claims 1 to 3, further includes:
To management server proxy module corresponding with the primary database send restoration information, so as to the primary database pair
The management server proxy module answered repairs the primary database.
11. a kind of for managing the device of database, comprising:
Receiving unit is configured as receiving the running state information of primary database;
Determination unit is configured as determining the operation shape of the primary database according to the running state information of the primary database
State;
Selecting unit is configured as determining the operating status of the primary database for primary database hair when the determination unit
When raw failure, from least two select from database one from database as new primary database;And
Transmission unit is configured as to taking respectively from other in database from the corresponding management of database with described at least two
Device proxy module of being engaged in sends information about new primary database, by with it is described other from the corresponding management server of database
Proxy module control it is described other from database from new primary database Backup Data.
12. device as claimed in claim 11, wherein the receiving unit is configured as from corresponding with the primary database
Management server proxy module receives the running state information of the primary database.
13. device as claimed in claim 11, wherein the selecting unit is configured to determine that described at least two from data
Its Transaction Identifier is managed the slave database of server acquisition in library, wherein the management server is for managing the master data
Library and described at least two is from database;Whether it is greater than described at least two from data from the quantity of database determined by judgement
The predetermined ratio of the quantity in library;It is greater than described at least two from the pre- of the quantity of database from the quantity of database when identified
When certainty ratio, it is determined as candidate slave database from database for identified;And it is selected from the slave database of the candidate
One is selected from database as new primary database.
14. a kind of for managing the device of database, comprising:
Processor;And
Memory, wherein computer-readable code is stored in the memory, the computer-readable code is when by described
When managing device operation, perform claim requires method described in any one of 1-11.
15. a kind of computer readable storage medium is stored thereon with instruction, described instruction by processor when being run, so that institute
It states processor and executes such as method of any of claims 1-11.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910353959.6A CN110069365B (en) | 2019-04-26 | 2019-04-26 | Method for managing database and corresponding device, computer readable storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910353959.6A CN110069365B (en) | 2019-04-26 | 2019-04-26 | Method for managing database and corresponding device, computer readable storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110069365A true CN110069365A (en) | 2019-07-30 |
CN110069365B CN110069365B (en) | 2023-07-04 |
Family
ID=67369589
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910353959.6A Active CN110069365B (en) | 2019-04-26 | 2019-04-26 | Method for managing database and corresponding device, computer readable storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110069365B (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111460039A (en) * | 2020-04-07 | 2020-07-28 | 中国建设银行股份有限公司 | Relational database processing system, client, server and method |
CN112199356A (en) * | 2020-12-09 | 2021-01-08 | 北京顺达同行科技有限公司 | Fault processing method, device, server and storage medium |
CN114676118A (en) * | 2022-05-30 | 2022-06-28 | 深圳市科力锐科技有限公司 | Database switching method, device, equipment and storage medium |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020194015A1 (en) * | 2001-05-29 | 2002-12-19 | Incepto Ltd. | Distributed database clustering using asynchronous transactional replication |
US20060222162A1 (en) * | 2005-03-31 | 2006-10-05 | Marcel Bank | Computer network system for building, synchronising and/or operating second database from/with a first database, and procedures for it |
EP2171619A1 (en) * | 2007-07-23 | 2010-04-07 | Telefonaktiebolaget LM Ericsson (PUBL) | Database system with multiple layer distribution |
JP2014026587A (en) * | 2012-07-30 | 2014-02-06 | Nippon Telegr & Teleph Corp <Ntt> | Distributed database system, first database device of distributed database system, second database device of distributed database system and computer program |
CN103914354A (en) * | 2012-12-31 | 2014-07-09 | 北京新媒传信科技有限公司 | Method and system for database fault recovery |
US20150278030A1 (en) * | 2012-12-28 | 2015-10-01 | Huawei Technologies Co., Ltd. | Distributed Database Synchronization Method and System |
US20170103000A1 (en) * | 2015-10-07 | 2017-04-13 | International Business Machines Corporation | Backup capability for object store used as primary storage |
CN108932295A (en) * | 2018-05-31 | 2018-12-04 | 康键信息技术(深圳)有限公司 | Primary database method for handover control, device, computer equipment and storage medium |
CN108984569A (en) * | 2017-06-05 | 2018-12-11 | 中兴通讯股份有限公司 | Database switching method, system and computer readable storage medium |
-
2019
- 2019-04-26 CN CN201910353959.6A patent/CN110069365B/en active Active
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020194015A1 (en) * | 2001-05-29 | 2002-12-19 | Incepto Ltd. | Distributed database clustering using asynchronous transactional replication |
US20060222162A1 (en) * | 2005-03-31 | 2006-10-05 | Marcel Bank | Computer network system for building, synchronising and/or operating second database from/with a first database, and procedures for it |
EP2171619A1 (en) * | 2007-07-23 | 2010-04-07 | Telefonaktiebolaget LM Ericsson (PUBL) | Database system with multiple layer distribution |
JP2014026587A (en) * | 2012-07-30 | 2014-02-06 | Nippon Telegr & Teleph Corp <Ntt> | Distributed database system, first database device of distributed database system, second database device of distributed database system and computer program |
US20150278030A1 (en) * | 2012-12-28 | 2015-10-01 | Huawei Technologies Co., Ltd. | Distributed Database Synchronization Method and System |
CN103914354A (en) * | 2012-12-31 | 2014-07-09 | 北京新媒传信科技有限公司 | Method and system for database fault recovery |
US20170103000A1 (en) * | 2015-10-07 | 2017-04-13 | International Business Machines Corporation | Backup capability for object store used as primary storage |
CN108984569A (en) * | 2017-06-05 | 2018-12-11 | 中兴通讯股份有限公司 | Database switching method, system and computer readable storage medium |
CN108932295A (en) * | 2018-05-31 | 2018-12-04 | 康键信息技术(深圳)有限公司 | Primary database method for handover control, device, computer equipment and storage medium |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111460039A (en) * | 2020-04-07 | 2020-07-28 | 中国建设银行股份有限公司 | Relational database processing system, client, server and method |
CN112199356A (en) * | 2020-12-09 | 2021-01-08 | 北京顺达同行科技有限公司 | Fault processing method, device, server and storage medium |
CN114676118A (en) * | 2022-05-30 | 2022-06-28 | 深圳市科力锐科技有限公司 | Database switching method, device, equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN110069365B (en) | 2023-07-04 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110311831B (en) | Container cloud-based system resource monitoring method and related equipment | |
CN107391294B (en) | Method and device for establishing IPSAN disaster recovery system | |
CN104679611B (en) | Data resource clone method and device | |
CN110069365A (en) | Manage the method and corresponding device, computer readable storage medium of database | |
CN109729129A (en) | Configuration modification method, storage cluster and the computer system of storage cluster | |
CN109788055B (en) | Service management system and method based on micro-service architecture | |
CN104036043B (en) | High availability method of MYSQL and managing node | |
CN108897658B (en) | Method and device for monitoring master database, computer equipment and storage medium | |
CN107666493B (en) | Database configuration method and equipment thereof | |
CN112583648B (en) | Intelligent service fault processing method based on DNS | |
CN108306760A (en) | For making the self-healing method and apparatus of managerial ability in a distributed system | |
CN116107828A (en) | Main node selection method, distributed database and storage medium | |
US20230004465A1 (en) | Distributed database system and data disaster backup drilling method | |
CN105323271B (en) | Cloud computing system and processing method and device thereof | |
CN116185697B (en) | Container cluster management method, device and system, electronic equipment and storage medium | |
CN115314361B (en) | Server cluster management method and related components thereof | |
JP3870174B2 (en) | Method for managing remotely accessible resources | |
CN115794769B (en) | Method for managing high-availability database, electronic equipment and storage medium | |
CN114297182A (en) | Industrial model data management method, device, equipment and readable storage medium | |
CN115878361A (en) | Node management method and device for database cluster and electronic equipment | |
CN108984356A (en) | A kind of IT product test method and device | |
CN112486718A (en) | Database fault automatic switching method and device and computer storage medium | |
CN114020279A (en) | Application software distributed deployment method, system, terminal and storage medium | |
CN114722029B (en) | Method, system, equipment and storage medium for repairing monitor database | |
CN115905270B (en) | Method and device for determining active data nodes in database and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |