CN105549909B - A kind of optimization method of cluster sequence class consensus information persistence - Google Patents
A kind of optimization method of cluster sequence class consensus information persistence Download PDFInfo
- Publication number
- CN105549909B CN105549909B CN201510926027.8A CN201510926027A CN105549909B CN 105549909 B CN105549909 B CN 105549909B CN 201510926027 A CN201510926027 A CN 201510926027A CN 105549909 B CN105549909 B CN 105549909B
- Authority
- CN
- China
- Prior art keywords
- information
- persistence
- value
- sequence class
- cluster sequence
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0602—Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
- G06F3/0614—Improving the reliability of storage systems
- G06F3/0619—Improving the reliability of storage systems in relation to data integrity, e.g. data losses, bit errors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0668—Interfaces specially adapted for storage systems adopting a particular infrastructure
- G06F3/0671—In-line storage system
- G06F3/0673—Single storage device
- G06F3/0674—Disk device
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Security & Cryptography (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of optimization methods of cluster sequence class consensus information persistence, when carrying out persistence to cluster sequence class consensus information, according to certain persistence interval, the partial information value chosen in cluster sequence class consensus information carries out persistence, that is the current persistent value of information of cluster sequence class consensus information is set as N, persistence interval is set as extent, then only carries out persistence to the value that the value of information is N+n*extent, n is nonnegative integer.The beneficial effects of the present invention are: interval persistence is carried out to cluster sequence class consensus information, reduces persistence frequency, so that cluster global failure can be coped with and improve the efficiency of persistence, reduces persistence cost.
Description
Technical field
The invention mainly relates to distributed computing field, specifically a kind of cluster sequence class consensus information persistence it is excellent
Change method.
Background technique
In the realization of cluster Consistency service, to cope with cluster global failure, such as cluster power-off, all node collapses
Extreme case, it is necessary to the consensus information persistence that will be stored.
There is a type of cluster consensus information, can be described as cluster sequence class consensus information, it is certain to refer to that those are deferred to
Calculation method is developed, successor value can be predicted and type of cluster consensus information with uniqueness.Such as it is
It unites revisions number SCN (i.e. system change number), it changes once, just increases 1 certainly.
Cluster sequence class consensus information, generally has following characteristics: 1, being allocated by cluster Consistency service, it is necessary to complete
Office is unique；2, change frequency may be very high；3, the IO amount very little of persistence.Therefore, to cluster sequence class consensus information into
When row persistence, although the IO total amount very little that sequence itself generates, due to change frequency height, IO times number is very high, according to disk
The characteristics of, the cost of persistence is still very big.
Summary of the invention
In order to solve deficiency in the prior art, the present invention provides a kind of the excellent of cluster sequence class consensus information persistence
Change method carries out interval persistence to cluster sequence class consensus information, reduces persistence frequency, so that cluster can be coped with
Global failure, and the efficiency of persistence can be improved, reduce persistence cost.
The present invention to achieve the above object, is achieved through the following technical solutions:
A kind of optimization method of cluster sequence class consensus information persistence is carried out to cluster sequence class consensus information
When persistence, according to certain persistence interval, the partial information value chosen in cluster sequence class consensus information is carried out persistently
Change, i.e. the current persistent value of information of cluster sequence class consensus information is set as N, and persistence interval is set as extent, then only right
The value that the value of information is N+n*extent carries out persistence, and n is nonnegative integer；The reality of the persistence of cluster sequence class consensus information
It is existing the following steps are included:
Step 1, starting cluster Consistency service；
Step 2 judges that the persistence file of cluster sequence class consensus information whether there is, if persistence file is deposited
3 are being thened follow the steps, if persistence file is not present, is thening follow the steps 4；
Step 3, read cluster sequence class consensus information persistence file in cluster sequence class consensus information with
And the value of information of cluster sequence class consensus information enters memory, gos to step 5 after the completion of reading；
The value of information of cluster sequence class consensus information is set as initial value by step 4, also as current persistence information
Value, sequence execute step 5；
Step 5, according to client request by cluster sequence class consensus information and cluster sequence class consensus information
The value of information synchronizes update between each node, and sequence executes step 6 after the completion of updating；
Step 6 judges whether the updated value of information meets the condition of the persistence of cluster sequence class consensus information, i.e.,
The updated value of information is set as M, and divided by extent after subtracting N with M, remainder zero, then updated value of information M meets cluster sequence
The persistence condition of column class consensus information, otherwise updated value of information M does not meet holding for cluster sequence class consensus information
Longization condition；If the updated value of information meets the condition of the persistence of cluster sequence class consensus information, then follow the steps
7, if the updated value of information does not meet the condition of the persistence of cluster sequence class consensus information, jumps and execute step 5；
Step 7 carries out persistence to the value of information of cluster sequence class consensus information, i.e., when cluster sequence class consistency is believed
When the value of information of breath is updated to M, and M meets the condition of persistence, and extent is divided between persistence, then to the information of M+extent
It is worth and carries out persistence, after the completion of persistence, jumps to step 5；
When step 8, cluster Consistency service normally stop, it is lasting that current cluster sequence class consensus information value is carried out
Change.
When carrying out persistence to cluster sequence class consensus information, only to the information in cluster sequence class consensus information
Value is that the value of the integral multiple at persistence interval carries out persistence.
The prior art is compared, the beneficial effects of the invention are that: it is lasting that the present invention carries out cluster sequence class consensus information
When change, according to certain persistence interval, the partial information value chosen in cluster sequence class consensus information carries out persistence, subtracts
Lack persistence frequency, so that cluster global failure can be coped with and improve the efficiency of persistence, reduces persistence cost.
Detailed description of the invention
Attached drawing 1 is broad flow diagram of the invention.
Specific embodiment
With reference to the drawings and specific embodiments, the invention will be further described.It should be understood that these embodiments are merely to illustrate
The present invention rather than limit the scope of the invention.In addition, it should also be understood that, after reading the content taught by the present invention, this field
Technical staff can make various changes or modifications the present invention, and such equivalent forms equally fall within the application the appended claims
Limited range.
A kind of optimization method of cluster sequence class consensus information persistence is carried out to cluster sequence class consensus information
When persistence, according to certain persistence interval, the partial information value chosen in cluster sequence class consensus information is carried out persistently
Change, i.e. the current persistent value of information of cluster sequence class consensus information is set as N, and persistence interval is set as extent, then only right
The value that the value of information is N+n*extent carries out persistence, and n is nonnegative integer；It is not every for cluster sequence class consensus information
Secondary variation all carries out persistence, but behind some persistence interval, then persistence is carried out, persistence frequency can be reduced in this way
Number, and cluster global failure can be coped with and improve the persistence efficiency of cluster sequence class consensus information, reduce persistence
Cost.As shown in Figure 1, the realization of the persistence of cluster sequence class consensus information the following steps are included:
Step 1, starting cluster Consistency service.
Step 2 judges that the persistence file of cluster sequence class consensus information whether there is, if persistence file is deposited
3 are being thened follow the steps, if persistence file is not present, is thening follow the steps 4.
Step 3, read cluster sequence class consensus information persistence file in cluster sequence class consensus information with
And the value of information of cluster sequence class consensus information enters memory, gos to step 5 after having read；For No. SCN, Ru Dangji
When the whole collapse of group restores again, No. SCN of persistence is read from disk, is updated since No. SCN of newest persistence,
Since No. SCN No. SCN for being necessarily smaller than newest persistence that cluster has been updated before, the SCN newly updated
Unique characteristic in complete or collected works group number is still maintained.
The value of information of cluster sequence class consensus information is set as initial value by step 4, also as current persistence information
Value, sequence execute step 5.
Step 5, according to client request by cluster sequence class consensus information and cluster sequence class consensus information
The value of information synchronizes update between each node, and sequence executes step 6 after the completion of updating；Cluster sequence class consistency is believed
Breath, is subject to the cluster sequence class consensus information value and its value of information (in memory) of update, for example, the value of information ratio of node A
The value of information of node B is new, then just with the value of information of the value of information overlay node B of node A.
Step 6 judges whether the updated value of information meets the condition of the persistence of cluster sequence class consensus information, i.e.,
The updated value of information is set as M, and divided by extent after subtracting N with M, remainder zero, then updated value of information M meets cluster sequence
The persistence condition of column class consensus information, otherwise updated value of information M does not meet holding for cluster sequence class consensus information
Longization condition；If the updated value of information meets the condition of the persistence of cluster sequence class consensus information, then follow the steps
7, if the updated value of information does not meet the condition of the persistence of cluster sequence class consensus information, jumps and execute step 5.
Step 7 carries out persistence to the value of information of cluster sequence class consensus information, i.e., when cluster sequence class consistency is believed
When the value of information of breath is updated to M, and M meets the condition of persistence, and extent is divided between persistence, then to the information of M+extent
It is worth and carries out persistence, after the completion of persistence, jumps to step 5；For example, for SCN, persistence gap size is
Extent, the current persistent value of information are that N, then by M+extent persistence, is updated to M when No. SCN when being updated to M for No. SCN
When+extent, then persistence M+2*extent, and so on；No. SCN of this section [M+1, M+extent-1] be not at this time
Persistence is carried out again.
When step 8, cluster Consistency service normally stop, it is lasting that current cluster sequence class consensus information value is carried out
Change.
When carrying out persistence to cluster sequence class consensus information, only to the information in cluster sequence class consensus information
Value is the value progress persistence of the integral multiple at persistence interval, for example, persistence gap size is extent for SCN,
The current persistent value of information is N, then N is the integral multiple of extent.
Obviously, the above embodiment of the present invention be only to clearly illustrate example of the present invention, and not be pair
The restriction of embodiments of the present invention.For those of ordinary skill in the art, may be used also on the basis of the above description
To make other variations or changes in different ways.There is no necessity and possibility to exhaust all the enbodiments.And these
Belong to the obvious changes or variations that connotation of the invention is extended out and still falls within protection scope of the present invention.
Claims (2)
1. a kind of optimization method of cluster sequence class consensus information persistence, which is characterized in that consistent to cluster sequence class
Property information when carrying out persistence, according to certain persistence interval, choose the partial information in cluster sequence class consensus information
Value carries out persistence, i.e. the current persistent value of information of cluster sequence class consensus information is set as N, and persistence interval is set as
Extent then only carries out persistence to the value that the value of information is N+n*extent, and n is nonnegative integer；Cluster sequence class consistency letter
The realization of the persistence of breath the following steps are included:
Step 1, starting cluster Consistency service；
Step 2 judges that the persistence file of cluster sequence class consensus information whether there is, if persistence file exists,
It executes step 3 and thens follow the steps 4 if persistence file is not present；
Step 3, read cluster sequence class consensus information persistence file in cluster sequence class consensus information and collection
The value of information of group's sequence class consensus information enters memory, gos to step 5 after the completion of reading；
The value of information of cluster sequence class consensus information is set as initial value by step 4, also as the current persistence value of information,
Sequence executes step 5；
Step 5, according to client request by the information of cluster sequence class consensus information and cluster sequence class consensus information
Value synchronizes update between each node, and sequence executes step 6 after the completion of updating；
Step 6 judges whether the updated value of information meets the condition of the persistence of cluster sequence class consensus information, that is, updates
The value of information afterwards is set as M, and divided by extent after subtracting N with M, remainder zero, then updated value of information M meets cluster sequence class
The persistence condition of consensus information, otherwise updated value of information M does not meet the persistence of cluster sequence class consensus information
Condition；If the updated value of information meets the condition of the persistence of cluster sequence class consensus information, 7 are thened follow the steps, such as
The updated value of information of fruit does not meet the condition of the persistence of cluster sequence class consensus information, then jumps and execute step 5；
Step 7 carries out persistence to the value of information of cluster sequence class consensus information, i.e., when cluster sequence class consensus information
When the value of information is updated to M, and M meets the condition of persistence, and extent is divided between persistence, then to the value of information of M+extent into
Row persistence after the completion of persistence, jumps to step 5；
When step 8, cluster Consistency service normally stop, current cluster sequence class consensus information value is subjected to persistence.
2. a kind of optimization method of cluster sequence class consensus information persistence according to claim 1, which is characterized in that
It is only lasting to the value of information in cluster sequence class consensus information when carrying out persistence to cluster sequence class consensus information
The value for changing the integral multiple at interval carries out persistence.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510926027.8A CN105549909B (en) | 2015-12-12 | 2015-12-12 | A kind of optimization method of cluster sequence class consensus information persistence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510926027.8A CN105549909B (en) | 2015-12-12 | 2015-12-12 | A kind of optimization method of cluster sequence class consensus information persistence |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105549909A CN105549909A (en) | 2016-05-04 |
| CN105549909B true CN105549909B (en) | 2019-03-12 |
Family
ID=55829112
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510926027.8A Active CN105549909B (en) | 2015-12-12 | 2015-12-12 | A kind of optimization method of cluster sequence class consensus information persistence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105549909B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102024032A (en) * | 2010-11-29 | 2011-04-20 | 广州明朝网络科技有限公司 | Distributed data caching and persisting method and system based on Erlang |
| CN104484469A (en) * | 2014-12-31 | 2015-04-01 | 天津南大通用数据技术股份有限公司 | Expansion method for supporting non-halt of multi-hash map database cluster system |
| CN104484472A (en) * | 2014-12-31 | 2015-04-01 | 天津南大通用数据技术股份有限公司 | Database cluster for mixing various heterogeneous data sources and implementation method |
| CN104615657A (en) * | 2014-12-31 | 2015-05-13 | 天津南大通用数据技术股份有限公司 | Expanding and shrinking method for distributed cluster with nodes supporting multiple data fragments |
| CN104866380A (en) * | 2015-06-18 | 2015-08-26 | 北京搜狐新媒体信息技术有限公司 | Method and device for processing state transition of cluster management system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8601222B2 (en) * | 2010-05-13 | 2013-12-03 | Fusion-Io, Inc. | Apparatus, system, and method for conditional and atomic storage operations |
-
2015
- 2015-12-12 CN CN201510926027.8A patent/CN105549909B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102024032A (en) * | 2010-11-29 | 2011-04-20 | 广州明朝网络科技有限公司 | Distributed data caching and persisting method and system based on Erlang |
| CN104484469A (en) * | 2014-12-31 | 2015-04-01 | 天津南大通用数据技术股份有限公司 | Expansion method for supporting non-halt of multi-hash map database cluster system |
| CN104484472A (en) * | 2014-12-31 | 2015-04-01 | 天津南大通用数据技术股份有限公司 | Database cluster for mixing various heterogeneous data sources and implementation method |
| CN104615657A (en) * | 2014-12-31 | 2015-05-13 | 天津南大通用数据技术股份有限公司 | Expanding and shrinking method for distributed cluster with nodes supporting multiple data fragments |
| CN104866380A (en) * | 2015-06-18 | 2015-08-26 | 北京搜狐新媒体信息技术有限公司 | Method and device for processing state transition of cluster management system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105549909A (en) | 2016-05-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11755675B2 (en) | Method and apparatus for managing region tag | |
| KR20160145482A (en) | Method and apparatus of implementing spiking neural network | |
| US9733835B2 (en) | Data storage method and storage server | |
| CN103136114A (en) | Storage method and storage device | |
| CN106445677A (en) | Load balancing method and device | |
| CN104301425A (en) | Object registration center service cluster implementation method based on object type and Hash ring | |
| CN104702691B (en) | Distributed load equalizing method and device | |
| CN109542684A (en) | A kind of storage remote copy method, device, equipment and storage medium | |
| Smékalová | Evaluating the cohesion policy: targeting of disadvantaged municipalities | |
| KR20170016168A (en) | Tile-based map data updating system and method thereof | |
| CN114385932A (en) | Meteorological data query method and device | |
| Dantas et al. | How past vicariant events can explain the Atlantic Forest biodiversity | |
| CN105549909B (en) | A kind of optimization method of cluster sequence class consensus information persistence | |
| Ye et al. | A two-layer geo-cloud based dynamic replica creation strategy | |
| CN108073472A (en) | A kind of memory correcting and eleting codes location mode perceived based on temperature | |
| CN105005586A (en) | Degree feature replacement policy based stream type graph sampling method | |
| CN104731716B (en) | A kind of date storage method | |
| JP6221701B2 (en) | Control program, control device, and control method | |
| JP5712851B2 (en) | Data division apparatus, data division method, and data division program | |
| CN112184900B (en) | Method, device and storage medium for determining elevation data | |
| CN104820633B (en) | Calculate the monitoring method and device of equipment | |
| CN104102557A (en) | Cloud computing platform data backup method based on clustering | |
| CN110909062A (en) | Data processing method and device, electronic equipment and readable storage medium | |
| De Keersmaeker et al. | A spatially explicit empirical model on actual and potential ancient forest plant diversity in a fragmented landscape | |
| US10402422B2 (en) | Method and system for synchronizing data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |