CN103399943A - Communication method and communication device for parallel query of clustered databases - Google Patents
Communication method and communication device for parallel query of clustered databases Download PDFInfo
- Publication number
- CN103399943A CN103399943A CN2013103539762A CN201310353976A CN103399943A CN 103399943 A CN103399943 A CN 103399943A CN 2013103539762 A CN2013103539762 A CN 2013103539762A CN 201310353976 A CN201310353976 A CN 201310353976A CN 103399943 A CN103399943 A CN 103399943A
- Authority
- CN
- China
- Prior art keywords
- query
- node
- data file
- data
- destination node
- 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.)
- Pending
Links
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a communication method and a communication device for parallel query of clustered databases. The communication method comprises the following steps: dividing query results coming from query nodes to obtain division blocks; transmitting the division blocks to a target node in an independent data file form; under the condition of transmitting all the data files to the target node, performing data merging on the data files on the target node. Through the transmission of the query results in an independent data file form, the query results are still saved in a magnetic disk when the data nodes or networks fail, so that the problem that communication data between the database nodes is lost and not easy to recover when the data nodes or the networks fail is avoided.
Description
Technical field
The present invention relates to the data base querying field, specifically, relate to a kind of means of communication and communication device of Cluster Database parallel query.
Background technology
In data-base cluster, tables of data is distributed on a plurality of database nodes according to certain Data Partition Strategy usually, so query processing is usually directed to executed in parallel and the internodal Data Transmission Controlling of a plurality of database nodes.In large scale system, single parts (such as network, node or the disk etc. of certain server) are inevitable while breaking down.How to guarantee when node failure, the communication state that can preserve between each node is a huge challenge that faces in data-base cluster.
At high-performance computing sector, usually adopt the message transmission to carry out communication, and, by preserving the mode of checkpoint, regularly the state of all concurrent processes is preserved, the reliability of system is provided.And in data base cluster system, the data volume that relates in query processing process is huge, thisly regularly the mode of preserving all process executing states, can introduce a large amount of disk read-writes, therefore can cause expense huge, will seriously reduce the performance of query processing.For this reason, people have utilized internal storage data structure and this dual mode of database temporary table Query Result is preserved and transmitted, still, when practical application, although this dual mode has solved the Communication between each node in the data base cluster system, also has many deficiencies, for example:
1) adopt the internal storage data structure to carry out data communication, the Memory Allocation management is relatively complicated, when it is carried out query processing, needs the Query Design personnel to realize various search algorithms, can't take full advantage of the query executor of database itself.In addition, when certain concurrent process breaks down, be difficult to carry out data and recover, reliability is relatively low.
2) adopt the mode of database temporary table to carry out data communication, can effectively utilize the query processing function of database, need not to realize separately search algorithm.This mode need to be written to the Query Result in each stage in the tables of data of target database.But the data of database temporary table write the constrained that is subject to database ACID characteristic (atomicity, consistance, integrality and permanent), the warehouse-in Performance Ratio of data is lower, this has also limited the performance of whole query processing, particularly relate in query processing data volume larger in.
When adopting internal memory and database temporary table to carry out data communication in correlation technique, the problem of reliability and warehouse-in poor-performing, not yet propose effective solution at present.
Summary of the invention
When adopting internal memory and database temporary table to carry out data communication in correlation technique, the problem of reliability and warehouse-in poor-performing, the present invention proposes a kind of means of communication and communication device of Cluster Database parallel query, can be good at solving the reliability of Query Result when preserving and transmit and the problem of warehouse-in poor-performing.
Technical scheme of the present invention is achieved in that
A kind of means of communication of Cluster Database parallel query are provided according to an aspect of the present invention.
These means of communication comprise:
Query Result from query node is divided, obtained divided block;
Each divided block is transferred to destination node with the form of independent data files;
, in the situation that all data files all transfer to destination node, the data file on destination node is carried out data merge.
Wherein, it is the next stage query node of query node at destination node, and destination node need to further be inquired about and while sorting the Query Result from query node, can be when the Query Result to from query node to be divided, and the mode that the employing scope is divided is divided.
In addition; it is the next stage query node of query node at destination node; and when destination node need to carry out correlation inquiry or classified statistics to the Query Result from query node, can be when the Query Result to from query node to be divided, the mode that adopts Hash to divide is divided.
Wherein, the quantity of divided block conforms to the quantity of the next stage query node of query node.
Wherein, after the Query Result to from query node is divided, can identify each divided block according to pre-configured identification information, and the divided block after sign is preserved with the form of data file, wherein, identification information comprise following one of at least: the numbering of query task numbering, the query node corresponding with the query task numbering, with query task, number the numbering of corresponding destination node.
Wherein, when each divided block is transferred to destination node with the form of independent data files, can, according to the identification information of each data file, determine query task, query node and the destination node corresponding with this data file; And according to query task, query node and the destination node determined, this data file copies to destination node, and in the situation that copy successfully, is deleted this data file from query node,, in the situation that copy unsuccessfully, record Operation Log.
In addition, when the data file on destination node is carried out the data merging, can analyze to the data file data message that the store path of specified data file and data file comprise; And the data message that comprises according to store path and the data file of established data file, the generating database external table.
A kind of communication device of Cluster Database parallel query is provided according to a further aspect in the invention.
This communication device comprises:
Divide module, be used for the Query Result of the node from inquiry is divided, obtain divided block;
Transport module, be used for each divided block is transferred to destination node with the form of independent data files;
Merge module, be used in the situation that all data files all transfer to destination node, the data file on destination node is merged.
Wherein, it is the next stage query node of query node at destination node, and destination node need to further be inquired about and while sorting, divide module when the Query Result to from query node is divided the Query Result from query node, and the mode of can the employing scope dividing is divided.
Wherein, next stage query node at the destination node query node; and when destination node need to carry out correlation inquiry or classified statistics to the Query Result from query node, divide module when the Query Result to from query node is divided, the mode that can adopt Hash to divide is divided.
Wherein, the quantity of divided block conforms to the quantity of the next stage query node of query node.
In addition, this communication device also comprises identification module, be used for after the Query Result to from query node is divided, according to pre-configured identification information, each divided block is identified, and the divided block after sign is preserved with the form of data file, wherein, identification information comprise following one of at least: the numbering of query task numbering, the query node corresponding with the query task numbering, with query task, number the numbering of corresponding destination node.
Wherein, transport module further comprises determination module and copy module, and wherein, determination module, be used for the identification information according to each data file, determines query task, query node and the destination node corresponding with this data file; The copy module, be used for according to query task, query node and the destination node determined, with this data file copies to destination node, and in the situation that copy successfully, this data file is deleted from query node,, in the situation that copy unsuccessfully, record Operation Log.
Wherein, merge module and further comprise analysis module and generation module, wherein, analysis module, be used for the data file is analyzed the data message that the store path of specified data file and data file comprise; Generation module, be used for according to the store path of established data file and the data message of data file, generating database external table.
The present invention carries out the transmission of Query Result by the form that adopts independent data files, make when database node or network transmission fault, the Query Result of query execution can still be kept in disk, thereby avoided when database node or network break down, the communication data between each database node is lost and is difficult for the problem of recovering.
In addition, the present invention carries out the merging of data by the mode that adopts the database external table, make when carrying out the data merging, data write does not need to carry out log information recording and safeguard the ACID characteristic, therefore, has solved in existing correlation technique and to have carried out Query Result while preserving, the problem of warehouse-in poor-performing, and the data integrity of also having avoided database to write bringing and the expense of consistency maintenance, improved the data transmission performance between each database node.
Description of drawings
In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, below will the accompanying drawing of required use in embodiment be briefly described, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skills, under the prerequisite of not paying creative work, can also obtain according to these accompanying drawings other accompanying drawing.
Fig. 1 is the schematic flow sheet according to the means of communication of the Cluster Database parallel query of the embodiment of the present invention;
Fig. 2 carries out according to the employing independent data files of the embodiment of the present invention and database external table the schematic flow sheet that parallel query is processed;
Fig. 3 is the structural representation according to the communication device of the Cluster Database parallel query of the embodiment of the present invention.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, obviously, described embodiment is only the present invention's part embodiment, rather than whole embodiment.Based on the embodiment in the present invention, the every other embodiment that those of ordinary skills obtain, belong to the scope of protection of the invention.
, according to embodiments of the invention, a kind of means of communication of Cluster Database parallel query are provided.
As shown in Figure 1, comprise according to the means of communication of the Cluster Database parallel query of the embodiment of the present invention:
Step S101, divide the Query Result from query node, obtains divided block;
Step S103, transfer to destination node to each divided block with the form of independent data files;
Step S105,, in the situation that all data files all transfer to destination node, carry out data to the data file on destination node and merge.
Due in large scale system, it is inevitable that single parts (for example CPU, internal memory, network interface card) break down, if the mode of employing internal storage data structure and message transmission is carried out the data communication between data-base cluster, so when breaking down, communication data between data-base cluster will be lost, but also be difficult to recover, therefore,, for fear of the generation of this situation, adopted the mode with independent data files that the Query Result data are transferred to destination node from query node in above-mentioned steps.The benefit that the mode of employing independent data files is transmitted is no matter be that database node breaks down or network breaks down, the Query Result of query execution can be kept in disk all the time, thereby factor data bank node or the network situation that causes loss of data that breaks down that effectively reduces occurs.
In the process of carrying out inquiry, if destination node is the next stage query node of query node, and destination node need to further be inquired about and while sorting the Query Result from query node, so when the Query Result to from query node is divided, the mode of can the employing scope dividing is divided.
Here the scope of saying is divided the data of the subrange that refers to a Query Result of each divided block preservation, and corresponding and a database node, for example, to { x1, x2, x3 ... xn, max_value } carry out the scope division, first divided block is exactly the data acquisition of value＜x1 so, and second divided block is exactly the data acquisition of x1≤value＜x2, by that analogy, last divided block is the data acquisition of xn≤value＜max_value.
In addition; in the process of carrying out inquiry; if destination node is the next stage query node of query node; and when destination node need to carry out correlation inquiry or classified statistics to the Query Result from query node; when the Query Result to from query node was divided, the mode that can adopt Hash to divide was divided so.
Here the Hash of saying is divided and is referred to adopt the mode of hash to divide Query Result, for example, and Hash(value)=bucket_id; wherein, bucket_id is the divided block numbering, and also database node corresponding to sign numbered simultaneously; for example, in 0～node_num-1.
In technique scheme; no matter be that the mode that the employing scope is divided is divided Query Result; still the mode that adopts Hash to divide is divided Query Result; the quantity of the divided block after it is divided all conforms to the quantity of the next stage query node of query node; here conforming to of saying can be identical, can be also the multiple relation.
In addition, in order to distinguish the divided block after division, the present invention can also be after the Query Result to from query node be divided, according to pre-configured identification information, each divided block is identified, and the divided block after sign is preserved with the form of data file, wherein, identification information comprise following one of at least: the numbering of query task numbering, the query node corresponding with the query task numbering, with query task, number the numbering of corresponding destination node.
When actual disposition, can adopt and self think that being easy to sign format is configured according to operating personnel's custom, for example, file name to be named, naming rule can be as follows:
${TASKID}_${SRCNODEID}_${DSTNODEID}.dat
Wherein, first refers to the numbering of query task, and second portion refers to the numbering (namely producing the node serial number of this result) of query task XM, and third part is the numbering of destination node, identifies this destination file and need to copy that node to.Adopt underscore to separate between three parts, the suffix name of file can be " dat ", from this file name, be easy to just to understand this document and belong to which query task, result from which query node (XM), need to be sent to which database node (destination node).
After the data division, to transmit with the form of independent data files, therefore, when each divided block is transferred to destination node with the form of independent data files, can, according to the identification information of each data file, determine query task, query node and the destination node corresponding with this data file; Then according to query task, query node and the destination node determined, with this data file copies to destination node, and, in the situation that copy successfully, this data file is deleted from query node,, in the situation that copy unsuccessfully, record Operation Log.
When actual transmissions, can adopt the mode of telecopy to copy, when the mode of employing telecopy copies, the first-selected data storage directory FPATH that need to read out each database node, then the data file FILEN under each this catalogue is carried out following steps: (this data file is when configuration identifier information a) to resolve the title of this data file, adopt above-mentioned naming rule to be configured), the numbering of acquisition mission number TASKID, source node (query node) and destination node; B) ID by database node obtains source node name and destination node name (perhaps address) IPADDR corresponding with source node numbering and destination node numbering; C) copy by the telecopy file, for example, scp $ FILEN} $ IPADDR}:$ FPATH}/; D) determining step c) whether successful,, if source file FILEN is deleted in success,, if unsuccessful, record corresponding daily record, and after copy is completed, export corresponding log recording.
In addition, in order to improve the data transmission performance between each database node in data-base cluster, the present invention can also when the data file on destination node is carried out the data merging, adopt the mode of creation database external table a plurality of data files in destination node to be combined into the integral body of a logic.And, because the Query Result data are recorded in independently in data file, be also to adopt a kind of regular mode, therefore, when the creation database external table, can adopt the Data Analysis mode of the database of standard to carry out Data Analysis.
The reason that the present invention adopts the database external table to carry out the data merging is that the database external table is when creating, need not carry out log information recording and safeguard the ACID characteristic, therefore, the data integrity that can avoid database to write bringing and the expense of consistency maintenance, both solve the problem of the warehouse-in poor-performing in existing correlation technique, improved again the data transmission performance between each database node.
The step that the present invention adopts the mode of database external table to realize that data merge is: the data file is analyzed, the data message that the store path of specified data file and data file comprise, according to the data message that store path and the data file of established data file comprises, generating database external table.
When actual creation database external table, can adopt oracle database to create, in the process that creates, the structure of table depends on the structure of last Query Result, the decollator sign that the partitioning scheme of data file adopts while depending on writing in files.Data file is the data file of coming from each query node copy previous stage, and concrete steps are as follows:
Certainly, when actual creation database external table, also can utilize other database to carry out the creation database external table, for example, postgresql.
Below by instantiation, technique scheme of the present invention is described, understand technique scheme of the present invention to facilitate.
Fig. 2 adopts independent data files and database external table to carry out the schematic flow sheet that parallel query is processed, as can be seen from Figure 2, each database node is after executing query task, need to divide the result of inquiry, and after division, the data of dividing are kept at respectively in different data files, then different data files is copied on corresponding database node by network remote, and corresponding database node is after receiving the data file that telecopy comes, need to adopt the mode of database external table, the data file of these dispersions is aggregated into the integral body of a logic, to facilitate the calculating of next query task.
, according to embodiments of the invention, also provide a kind of communication device of Cluster Database parallel query.
As shown in Figure 3, comprise according to the communication device of the Cluster Database parallel query of the embodiment of the present invention:
In the process of carrying out inquiry by said structure, if destination node is the next stage query node of query node, and destination node need to further be inquired about and while sorting the Query Result from query node, divide module 31 when the Query Result to from query node is divided, the mode of can the employing scope dividing is divided.
In addition; in the process of carrying out inquiry by said structure; if destination node is the next stage query node of query node; and when destination node need to carry out correlation inquiry or classified statistics to the Query Result from query node; divide module 31 when the Query Result to from query node is divided, the mode that can adopt Hash to divide is divided.
In addition; in the process of carrying out inquiry by said structure; divide module 31 no matter be that the mode that the employing scope is divided is divided the Query Result from query node; still the mode that adopts Hash to divide is divided the Query Result from query node; the quantity of the divided block after division all conforms to the quantity of the next stage query node of query node; here said conforming to can be identical, can be also the relation of multiple.
in addition, in order to distinguish the divided block after division, also comprise the identification module (not shown) according to the communication device of the Cluster Database parallel query of the embodiment of the present invention, be used for after the Query Result to from query node is divided, according to pre-configured identification information, each divided block is identified, and the divided block after sign is preserved with the form of data file, wherein, identification information comprise following one of at least: the query task numbering, the numbering of the query node corresponding with the query task numbering, the numbering of the destination node corresponding with the query task numbering.
In addition, transport module 32 is when transferring to destination node with each divided block with the form of independent data files, realize by determination module (not shown) and copy module (not shown), wherein, determination module, be used for the identification information according to each data file, determine query task, query node and the destination node corresponding with this data file; The copy module, be used for according to query task, query node and the destination node determined, with this data file copies to destination node, and in the situation that copy successfully, data file is deleted from query node,, in the situation that copy unsuccessfully, record Operation Log.
Same, merge module 33 when the data file with on destination node merges, carry out merge data file by analysis module (not shown) and generation module (not shown) generating database external table, wherein, analysis module, be used for the data file is analyzed the data message that the store path of specified data file and data file comprise; Generation module, be used for according to the store path of established data file and the data message of data file, generating database external table.
In sum, by means of technique scheme of the present invention, carry out the transmission of Query Result by the form that adopts independent data files, make when database node or network transmission fault, the Query Result of query execution can still be kept in disk, thereby avoided when database node or network break down, the communication data between each database node is lost and is difficult for the problem of recovering.
In addition, by means of technique scheme of the present invention, carry out the merging of data by the mode that adopts the database external table, make when carrying out the data merging, data write not to need to carry out log information recording and safeguards the ACID characteristic, therefore, solved in existing correlation technique and to have carried out Query Result while preserving, the problem of warehouse-in poor-performing, and, the data integrity of also having avoided database to write bringing and the expense of consistency maintenance, improved the data transmission performance between each database node.
The foregoing is only preferred embodiment of the present invention, in order to limit the present invention, within the spirit and principles in the present invention not all, any modification of doing, be equal to replacement, improvement etc., within all should being included in protection scope of the present invention.
Claims (10)
1. the means of communication of a data-base cluster parallel query, is characterized in that, comprising:
Query Result from query node is divided, obtained divided block;
Each divided block is transferred to destination node with the form of independent data files;
, in the situation that all data files all transfer to described destination node, the data file on described destination node is carried out data merge.
2. the means of communication according to claim 1, it is characterized in that, if described destination node is the next stage query node of described query node, and described destination node need to further be inquired about and sort the Query Result from described query node, when the Query Result to from described query node was divided, the mode that the employing scope is divided was divided.
3. the means of communication according to claim 1; it is characterized in that; if described destination node is the next stage query node of described query node; and described destination node need to carry out correlation inquiry or classified statistics to the Query Result from described query node; when the Query Result to from described query node was divided, the mode that adopts Hash to divide was divided.
4. the means of communication according to claim 1, is characterized in that, the quantity of described divided block conforms to the quantity of the next stage query node of described query node.
5. the means of communication according to claim 1, is characterized in that, after the Query Result from described query node is divided, further comprises:
According to pre-configured identification information, each divided block is identified, and the divided block after sign is preserved with the form of data file, wherein, described identification information comprise following one of at least: the numbering of query task numbering, the described query node corresponding with described query task numbering, with described query task, number the numbering of corresponding described destination node.
6. the means of communication according to claim 5, is characterized in that, each divided block transferred to destination node with the form of independent data files comprise:
, according to the identification information of each data file, determine query task, query node and the destination node corresponding with this data file;
According to described query task, query node and the destination node determined, with this data file copies to destination node; And in the situation that copy successfully, this data file is deleted from query node,, in the situation that copy unsuccessfully, record Operation Log.
7. the means of communication according to claim 1, is characterized in that, the data file on described destination node carried out the data merging comprise:
Described data file is analyzed, determined the store path of described data file and the data message that described data file comprises;
According to the store path of the described data file of determining and the data message that described data file comprises, generating database external table.
8. the communication device of a Cluster Database parallel query, is characterized in that, comprising:
Divide module, be used for the Query Result from query node is divided, obtain divided block;
Transport module, be used for each divided block is transferred to destination node with the form of independent data files;
Merge module, be used in the situation that all data files all transfer to described destination node, the data file on described destination node is merged.
9. communication device according to claim 8, is characterized in that, further comprises:
Identification module, be used for after the Query Result to from query node is divided, according to pre-configured identification information, each divided block is identified, and the divided block after sign is preserved with the form of data file, wherein, described identification information comprise following one of at least: the numbering of query task numbering, the described query node corresponding with described query task numbering, with described query task, number the numbering of corresponding described destination node.
10. communication device according to claim 8, is characterized in that, described merging module further comprises analysis module and generation module, wherein,
Analysis module, be used for described data file is analyzed, and determines the store path of described data file and the data message that described data file comprises;
Generation module, be used for according to the store path of the described data file of determining and the data message of described data file, generating database external table.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013103539762A CN103399943A (en) | 2013-08-14 | 2013-08-14 | Communication method and communication device for parallel query of clustered databases |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013103539762A CN103399943A (en) | 2013-08-14 | 2013-08-14 | Communication method and communication device for parallel query of clustered databases |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103399943A true CN103399943A (en) | 2013-11-20 |
Family
ID=49563571
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2013103539762A Pending CN103399943A (en) | 2013-08-14 | 2013-08-14 | Communication method and communication device for parallel query of clustered databases |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103399943A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104156400A (en) * | 2014-07-22 | 2014-11-19 | 中国科学院信息工程研究所 | Storage method and device of mass network flow data |
| CN104239417A (en) * | 2014-08-19 | 2014-12-24 | 天津南大通用数据技术股份有限公司 | Dynamic adjustment method and dynamic adjustment device after data fragmentation in distributed database |
| CN106156202A (en) * | 2015-04-22 | 2016-11-23 | 中国移动通信集团广东有限公司 | A kind of processing method and processing device of database failure |
| CN108108473A (en) * | 2018-01-02 | 2018-06-01 | 联想(北京)有限公司 | Data query method and server |
| CN109977077A (en) * | 2019-03-25 | 2019-07-05 | 腾讯科技（深圳）有限公司 | Model file storage method, device, readable storage medium storing program for executing and computer equipment |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102158531A (en) * | 2010-02-01 | 2011-08-17 | 北京大学 | Distributed transmission method for query data stream |
| CN102402586A (en) * | 2011-10-24 | 2012-04-04 | 深圳华强电子交易网络有限公司 | Distributed data storage method |
| CN102467570A (en) * | 2010-11-17 | 2012-05-23 | 日电（中国）有限公司 | Connection query system and method for distributed data warehouse |
| CN102521307A (en) * | 2011-12-01 | 2012-06-27 | 北京人大金仓信息技术股份有限公司 | Parallel query processing method for share-nothing database cluster in cloud computing environment |
| CN102831120A (en) * | 2011-06-15 | 2012-12-19 | 腾讯科技（深圳）有限公司 | Data processing method and system |
| CN103123652A (en) * | 2013-03-14 | 2013-05-29 | 曙光信息产业(北京)有限公司 | Data query method and cluster database system |
| CN103136363A (en) * | 2013-03-14 | 2013-06-05 | 曙光信息产业(北京)有限公司 | Inquiry processing method and cluster data base system |
-
2013
- 2013-08-14 CN CN2013103539762A patent/CN103399943A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102158531A (en) * | 2010-02-01 | 2011-08-17 | 北京大学 | Distributed transmission method for query data stream |
| CN102467570A (en) * | 2010-11-17 | 2012-05-23 | 日电（中国）有限公司 | Connection query system and method for distributed data warehouse |
| CN102831120A (en) * | 2011-06-15 | 2012-12-19 | 腾讯科技（深圳）有限公司 | Data processing method and system |
| CN102402586A (en) * | 2011-10-24 | 2012-04-04 | 深圳华强电子交易网络有限公司 | Distributed data storage method |
| CN102521307A (en) * | 2011-12-01 | 2012-06-27 | 北京人大金仓信息技术股份有限公司 | Parallel query processing method for share-nothing database cluster in cloud computing environment |
| CN103123652A (en) * | 2013-03-14 | 2013-05-29 | 曙光信息产业(北京)有限公司 | Data query method and cluster database system |
| CN103136363A (en) * | 2013-03-14 | 2013-06-05 | 曙光信息产业(北京)有限公司 | Inquiry processing method and cluster data base system |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104156400A (en) * | 2014-07-22 | 2014-11-19 | 中国科学院信息工程研究所 | Storage method and device of mass network flow data |
| CN104156400B (en) * | 2014-07-22 | 2017-07-11 | 中国科学院信息工程研究所 | The storage method and device of a kind of mass network flow data |
| CN104239417A (en) * | 2014-08-19 | 2014-12-24 | 天津南大通用数据技术股份有限公司 | Dynamic adjustment method and dynamic adjustment device after data fragmentation in distributed database |
| CN104239417B (en) * | 2014-08-19 | 2017-06-09 | 天津南大通用数据技术股份有限公司 | Dynamic adjusting method and device after a kind of distributed data base data fragmentation |
| CN106156202A (en) * | 2015-04-22 | 2016-11-23 | 中国移动通信集团广东有限公司 | A kind of processing method and processing device of database failure |
| CN106156202B (en) * | 2015-04-22 | 2019-11-19 | 中国移动通信集团广东有限公司 | A kind of processing method and processing device of database failure |
| CN108108473A (en) * | 2018-01-02 | 2018-06-01 | 联想(北京)有限公司 | Data query method and server |
| CN109977077A (en) * | 2019-03-25 | 2019-07-05 | 腾讯科技（深圳）有限公司 | Model file storage method, device, readable storage medium storing program for executing and computer equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102411637B (en) | Metadata management method of distributed file system | |
| JP6044539B2 (en) | Distributed storage system and method | |
| CA2913036C (en) | Index update pipeline | |
| CN106406896B (en) | Block chain block building method based on parallel Pipeline technology | |
| US9864527B1 (en) | Distributed data storage management | |
| CN107391758B (en) | Database switching method, device and equipment | |
| CN102142032B (en) | Method and system for reading and writing data of distributed file system | |
| US20100023564A1 (en) | Synchronous replication for fault tolerance | |
| CN107193490A (en) | A kind of distributed data-storage system and method based on block chain | |
| CN102411639B (en) | Multi-copy storage management method and system of metadata | |
| CN105138615A (en) | Method and system for building big data distributed log | |
| EP4213038A1 (en) | Data processing method and apparatus based on distributed storage, device, and medium | |
| CN102214205A (en) | Logical replication in clustered database system with adaptive cloning | |
| CN106708653B (en) | Mixed tax big data security protection method based on erasure code and multiple copies | |
| CN103500120A (en) | Distributed cache high-availability processing method and system based on multithreading asynchronous double writing | |
| CN103399943A (en) | Communication method and communication device for parallel query of clustered databases | |
| CN106407224A (en) | Method and device for file compaction in KV (Key-Value)-Store system | |
| CN103399894A (en) | Distributed transaction processing method on basis of shared storage pool | |
| CN105045917A (en) | Example-based distributed data recovery method and device | |
| CN104965835B (en) | A kind of file read/write method and device of distributed file system | |
| CN106873902B (en) | File storage system, data scheduling method and data node | |
| CN115114370B (en) | Master-slave database synchronization method and device, electronic equipment and storage medium | |
| CN102541693A (en) | Multi-copy storage management method and system of data | |
| WO2024021470A1 (en) | Cross-region data scheduling method and apparatus, device, and storage medium | |
| CN109542860B (en) | Service data management method based on HDFS and terminal equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20131120 |