CN110427437A - A kind of relevant database mixing isomery interrogation model and method towards big data - Google Patents
A kind of relevant database mixing isomery interrogation model and method towards big data Download PDFInfo
- Publication number
- CN110427437A CN110427437A CN201910701575.9A CN201910701575A CN110427437A CN 110427437 A CN110427437 A CN 110427437A CN 201910701575 A CN201910701575 A CN 201910701575A CN 110427437 A CN110427437 A CN 110427437A
- Authority
- CN
- China
- Prior art keywords
- index
- data
- database
- hdfs
- global
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/10—File systems; File servers
- G06F16/18—File system types
- G06F16/182—Distributed file systems
-
- 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
- G06F16/22—Indexing; Data structures therefor; Storage structures
- G06F16/2228—Indexing structures
-
- 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
- G06F16/24—Querying
-
- 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
- G06F16/28—Databases characterised by their database models, e.g. relational or object models
- G06F16/284—Relational databases
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of relevant database mixing isomery interrogation model towards big data, structure are as follows: the bottom is center database, is used to store the data to be inquired；Middle layer is Hadoop distributed file system HDFS, store metadata and intermediate result, meanwhile addition data buffer storage layer and global index's layer are used to store global index's information, index list and the indexed cache of all data dictionary table and its result buffering and all tables of data；Top layer uses MapReduce programming model, provides to parallel data processing in HDFS and guarantees fault-tolerance；Four Design of Middleware DB connector, data loader, index maker and query engine functional modules.The model supports original data dynamic divides, and improves pretreatment efficiency, on the basis of keeping former data complete, improves the response time of inquiry request under large-scale consumer.
Description
Technical field
The present invention relates to a kind of relevant database mixing isomery interrogation model and method towards big data, belongs to information
Technical treatment field.
Background technique
In big data analysis field, with the sharp increase of data volume, high scalability and high-performance are that big data analysis is flat
The characteristic of platform indispensability, parallel database include advanced technological means and algorithm, such as index, data compression, Materialized View, result
Shared, data connection of optimization of buffering, I/O etc.；But it is lacking in scalability, only supports limited spread mostly, i.e., it is hundreds of
The scale MapReduce of node is proposed by Googel, is towards the unstructured data on extensive low-cost server cluster
It is disposable to handle and design.MapReduce has enhanced scalability, it can arbitrarily adds or remove in the cluster section
Point, and have little influence on the execution of existing task；But under same hardware condition, the property of MapReduce processing structure data
It can be far below parallel database, so that its performance when processing connects (key operation of data analysis field) is especially not to the utmost such as
People's will.The system for being based solely on parallel database or realization is not the ideal solution of big data analysis, therefore, great Liang Yan
Study carefully and set about combining both, desired design goes out the big data analysis platform for having both the two advantage.HaoopDB is as mixing collection
At the representative studies in system, system architecture is divided into two layers: upper layer uses the decomposition and scheduling of carry out task, and lower layer uses
The inquiry and processing of relational database progress data.Its innovation is: preferable by obtaining by Hadoop frame
Fault-tolerance and support to isomerous environment；Preferable performance is obtained by that will inquire to execute in pushed data library as far as possible.In
Data need to use after the overall situation divides and database loads in HaoopDB, and the dynamic of data is not supported still to divide,
It needs disposably to divide data.Cause the database data of large-scale data to pre-process cost prohibitive, while data being put down
It is assigned to each node, cannot keep the independent completion of original each database.
Summary of the invention
Goal of the invention: being directed to the problems of the prior art, uses for reference HadoopDB system principle, open one kind is towards big data
Relevant database mixing isomery interrogation model and method, model supports original data dynamic divide, improve pretreatment efficiency,
On the basis of keeping former data complete, the response time of inquiry request under large-scale consumer is improved.
Technical solution: a kind of relevant database mixing isomery interrogation model towards big data, including use for reference
HadoopDB carries out query task decomposition, scheduling and inquiry, the processing of relevant database completion data by Hadoop and mixes
A kind of relevant database mixing isomery interrogation model towards big data is divided into four parts at system architecture by intersection, including
Data Layer, Hadoop distributed file system HDFS, MapReduce programming model and the centre of the central database of the bottom
Part, the data Layer of the central database of the bottom are used to store the data to be inquired, by establishing global index's mechanism and classification
Inquiry mechanism improves tables of data search efficiency；Data Layer includes multiple database nodes, and middle layer is Hadoop distributed document
System HDFS stores metadata and intermediate result, meanwhile, it adds data buffer storage layer and global index's layer is used to store all numbers
According to dictionary table, global index's information, index list and the indexed cache of result buffering and all tables of data；Top layer uses
MapReduce programming model provides to parallel data processing in HDFS and ensures fault-tolerance；Middleware includes database connection
Four device, data loader, the index maker of middleware and query engine functional modules, global index's layer include that index generates
Device, lru algorithm (at least using algorithm, Least Recently Used in the recent period), index generate caching, index list.
Further, Hadoop file system is connected with database, is made by Design of Middleware DB connector
MapReduce can direct Accessing Oracle Database, using Hadoop provide DBInptuFormat class access database,
Database table data is read into HDFS；Recycle TextOutpuFormat class that obtained result is written in database table.
Further, Design of Middleware data loader, by data dictionary table and the local index of tables of data from database
The middle back end loaded into HDFS, to data dictionary table, due to its enquiry frequency height, but data volume very little and table it is substantially stationary
It is constant, therefore all data dictionary tables are all loaded into the data buffer storage layer into HDFS；To tables of data, only its local index is added
The global index's layer being loaded into.
Further, middleware index maker, using MapReduce programming model, by load into local index close
And be global index, it is stored in the back end of HDFS, after global index generates, the local index in HDFS will be discharged to subtract
Few space will be divided into suitable size if global index is excessive automatically, and index of reference catalogue records every piece of position
It sets, index list uses format, is stored in the namenode of HDFS, and index buffering at least uses algorithm, that is, LRU using recent
(Least Recently Used), most-often used global index's block is moved in namenode from the back end of HDFS.
Further, Design of Middleware query engine provides different inquiry requests different implementation strategies.To data
When dictionary table is inquired, HDFS will be directly accessed.When to not having to establish the Field Inquiry indexed in tables of data, bottom will be directly accessed
Database；When to the Field Inquiry for having built up index in tables of data, which number global index's layer in first access is obtained into
It should be accessed according to library, then concurrent access these databases, each database completes partial query, finally closes obtained result
And it returns.
Further, data Layer establishes local index and global index, establishes to physical location field in database table
Catalogue can quickly access the specific information in table using index, and database data has multiple copies, can use data duplicate energy
The memory access performance of raising task improves memory access speed and utilizes system on the basis of Analysis and Screening goes out user and commonly inquires field
Every part of data have the characteristics of a multiple copies in system, establish global index for tables of data, improve the search efficiency of tables of data, right
Each database node, DB connector allow the direct access number of MapReduce for connecting Hadoop and database
According to library, data loader is loaded the local index of tables of data into HDFS file system with MapReduce by DB connector
Back end in system；Index maker will load local index merging into HDFS according to index field, using MapReduce
For global index, it is stored in the back end of HDFS；If global index is excessive, index maker will be according to HDFS file format
It is required that being divided into suitable data block N block, N is the natural number greater than 1, and updates index slit record to record every piece of position
It sets；The local index HDFS file in back end is discharged, to save space.
Further, index tree is established by global index's layer in back end to access global index's file, to obtain
Global index's record of user query condition must be met, global index's mechanism accessing step is as follows:
Step 1, index tree leaf node is generated, one or more in global index's file is scanned, when in certain row record
It when the value of index field meets querying condition, marks the row and establishes pointer, pointer is distributed according to each tag field voluntarily
Onto corresponding leaf node, leaf node number is the pointer list that n, that is, each leaf node has non-empty；
Step 2, the intermediate node and root node for generating index tree, according to the topological diagram of system, by each leaf node pair
The pointer list of intermediate node and root node the insertion index tree answered, intermediate node and root node is temporarily empty；
Step 3, the pointer list on all nodes is updated, remembers that the number of nodes of index tree is nindex(nindex≤ n, n are leaf
Son node number) remember that the finger in all leaf nodes is counted as wpointerSo each node should averagely possess p=wpointer/
nindexA pointer, if having more than p pointer in the finger meter list of leaf node, p pointer is motionless before retaining, by remaining
Pointer moves into the pointer list of the corresponding intermediate node of the leaf node；
Step 4, the location information field table on all nodes is generated, to each node i, scan pointer list, according to section
Point attribute finds each pointer in global index's document to capable location information field, generates location information list.
A kind of relevant database mixing isomery querying method towards big data, including with following steps:
Step 1, by pattern query engine, SQL query statement is submitted；
Step 2, Data Position is searched according to the global index that middleware index maker generates, MapReduce is programmed
The Map function of model is dispatched to corresponding node；
Step 3, if global index there are the data to be inquired, data are loaded by data loader；If inquiry word
Section then directly accesses HDFS without establishing index；
Step 4, Map function will execute in the database of SQL query push-in bottom data node, what each database returned
Intermediate result carries out reduction by MapReduce programming model Reduce function again；
It step 5, will be in the result write-in HDFS after reduction.
The utility model has the advantages that
A kind of relevant database mixing isomery interrogation model towards big data of the invention supports former data dynamic to draw
Point, pretreatment efficiency is improved, on the basis of keeping former data complete, when improving the response of inquiry request under large-scale consumer
Between.
Detailed description of the invention
Fig. 1 is a kind of relevant database mixing isomery interrogation model system construction drawing towards big data.
Specific embodiment
Further explanation is done to the present invention with reference to the accompanying drawing.
Data Layer establishes local index and global index, to the catalogue that fields certain in database table are established, uses index
The specific information in table can quickly be accessed.General database data have multiple copies, and can use data duplicate can improve task
Memory access performance, improve obtain memory access speed-up ratio.On the basis of going out user's the most commonly used inquiry field by Analysis and Screening,
Every part of data have the characteristics of multiple copies in reutilization system, establish global index for tables of data.Pass through the index of middleware
Generator generates index leaf node, intermediate node and root node and establishes index tree to access global index's file, to be expired
The global index of sufficient user query condition records, and global index is put in cache layer, reading rate with higher.
A kind of relevant database mixing isomery interrogation model towards big data, middle layer are Hadoop distributed document
System HDFS stores metadata and intermediate result, meanwhile, it adds data buffer storage layer and global index's layer is used to store all numbers
According to dictionary table and its global index's information, index list and the indexed cache of result buffering and all tables of data；Top layer uses
MapReduce programming model provides to parallel data processing in HDFS and guarantees fault-tolerance；Middleware includes database connection
Four device, data loader, index maker and query engine functional modules.Design of Middleware DB connector, will
Hadoop is connected with database, and MapReduce is allow to directly access the database node data loader for access frequency
Higher data dictionary table and the local index of tables of data load the back end into HDFS from database.Index maker benefit
With MapReduce programming model, by load into local index merge into global index, be stored in the back end of HDFS.It looks into
It askes engine and different implementation strategies is provided different inquiry requests.When to the inquiry of data dictionary table, HDFS will be directly accessed.It is right
When not having to establish the Field Inquiry indexed in tables of data, the database of bottom will be directly accessed；To having built up rope in tables of data
When the Field Inquiry drawn, will first access in global index layer obtain which database should be accessed, then concurrent access these
Database, each database complete partial query, finally merge obtained result and return.
The present invention proposes a kind of relevant database mixing isomery towards big data and looks on the basis of HadoopDB
The innovation for asking model and method is: keep the independent completion of original relational database, on the basis of increase data buffer storage layer,
For storing all data dictionary table and its result buffering, the concurrent inquiry request under large-scale consumer can be timely responded to；For
A kind of global index of design of database system with multiple data duplicates supports when by index field inquiry tables of data
Former data dynamic divides, and improves pretreatment efficiency, on the basis of keeping former data complete, improves to inquire under large-scale consumer and ask
The response time asked.
The above is only a preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications are also answered
It is considered as protection scope of the present invention.
Claims (8)
1. a kind of relevant database mixing isomery interrogation model towards big data, which is characterized in that including in the bottom
Data Layer, Hadoop distributed file system HDFS, MapReduce programming model and the middleware of heart database, the bottom
The data Layer of central database is used to store the data to be inquired, is mentioned by establishing global index's mechanism and classified inquiry mechanism
High tables of data search efficiency；Data Layer includes multiple database nodes, and middle layer is Hadoop distributed file system HDFS, is deposited
Metadata and intermediate result are put, meanwhile, it adds data buffer storage layer and global index's layer is used to store all data dictionary tables, knot
Global index's information, index list and the indexed cache of fruit buffering and all tables of data；Top layer programs mould using MapReduce
Type provides to parallel data processing in HDFS and ensures fault-tolerance；Middleware include DB connector, data loader, in
Between part four functional modules of index maker and query engine, global index layer include index maker, lru algorithm, index
Generate caching, index list.
2. a kind of relevant database mixing isomery interrogation model towards big data as described in claim 1, feature exist
In: Hadoop file system is connected with database, visits MapReduce directly by Design of Middleware DB connector
It asks oracle database, accesses database using the DBInptuFormat class that Hadoop is provided, database table data is read into
HDFS；Recycle TextOutpuFormat class that obtained result is written in database table.
3. a kind of relevant database mixing isomery interrogation model towards big data as described in claim 1, feature exist
In: Design of Middleware data loader loads data dictionary table and the local index of tables of data from database into HDFS's
All data dictionary tables are all loaded the data buffer storage layer into HDFS to data dictionary table by back end；To tables of data, only
By its local index load into global index layer.
4. a kind of relevant database mixing isomery interrogation model towards big data as described in claim 1, feature exist
In middleware index maker, using MapReduce programming model, by load into local index merge into global index, deposit
It is placed on the back end of HDFS, after global index generates, the local index in HDFS will be discharged to reduce space, if global
It indexes excessive, will be divided into suitable size automatically, and index of reference catalogue records every piece of position, index list uses lattice
Formula is stored in the namenode of HDFS, and index buffering uses lru algorithm, by most-often used global index's block from HDFS's
Back end moves in namenode.
5. a kind of relevant database mixing isomery interrogation model towards big data as described in claim 1, feature exist
In: Design of Middleware query engine provides different inquiry requests different implementation strategies, when inquiring data dictionary table,
HDFS directly being accessed, when to not having to establish the Field Inquiry indexed in tables of data, will directly access the database of bottom；Logarithm
When according to the Field Inquiry for having built up index in table, global index's layer in first access is obtained which database should be interviewed
It asks, then concurrent access these databases, each database completes partial query, finally merges obtained result and returns.
6. a kind of relevant database mixing isomery interrogation model towards big data as described in claim 1, feature exist
In: data Layer establishes local index and global index, the catalogue established to physical location field in database table；To each number
According to library node, DB connector directly accesses the database MapReduce, number for connecting Hadoop and database
According to loader by DB connector, the local index of tables of data is loaded into HDFS file system with MapReduce
Back end；The local index for loading into HDFS is merged into the overall situation according to index field, using MapReduce by index maker
Index, is stored in the back end of HDFS；If global index is excessive, index maker will be incited somebody to action according to HDFS file format requirements
It is divided into suitable data block N block, and N is the natural number greater than 1, and updates index slit record to record every piece of position；Release
Local index HDFS file in back end, to save space.
7. a kind of relevant database mixing isomery interrogation model towards big data as described in claim 1, feature exist
In: index tree is established by global index's layer in back end and accesses global index's file, meets user query to obtain
The global index of condition records, and global index's mechanism accessing step is as follows:
Step 1, index tree leaf node is generated, one or more in global index's file is scanned, is indexed when in certain row record
It when the value of field meets querying condition, marks the row and establishes pointer, pointer is assigned to phase according to each tag field voluntarily
On the leaf node answered, leaf node number is the pointer list that n, that is, each leaf node has non-empty；
Step 2, the intermediate node and root node for generating index tree, it is according to the topological diagram of system, each leaf node is corresponding
The pointer list of intermediate node and root node insertion index tree, intermediate node and root node is temporarily empty；
Step 3, the pointer list on all nodes is updated, remembers that the number of nodes of index tree is nindex(nindex≤ n, n are leaf section
Points) remember that the finger in all leaf nodes is counted as wpointerSo each node should averagely possess p=wpointer/nindexIt is a
Pointer, if having more than p pointer in the finger meter list of leaf node, p pointer is motionless before retaining, remaining pointer is moved
In the pointer list for entering the corresponding intermediate node of the leaf node；
Step 4, the location information field table on all nodes is generated, to each node i, scan pointer list, according to node category
Property find each pointer in global index's document to capable location information field, generate location information list.
8. a kind of relevant database mixing isomery querying method towards big data as described in claim 1, feature exist
In, including with following steps:
Step 1, by pattern query engine, SQL query statement is submitted；
Step 2, Data Position is searched according to the global index that middleware index maker generates, by MapReduce programming model
Map function be dispatched to corresponding node；
Step 3, if global index there are the data to be inquired, data are loaded by data loader；If inquiry field does not have
There is foundation to index, then directly accesses HDFS；
Step 4, Map function will execute in the database of SQL query push-in bottom data node, the centre that each database returns
As a result reduction is carried out by MapReduce programming model Reduce function again；
It step 5, will be in the result write-in HDFS after reduction.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910701575.9A CN110427437A (en) | 2019-07-31 | 2019-07-31 | A kind of relevant database mixing isomery interrogation model and method towards big data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910701575.9A CN110427437A (en) | 2019-07-31 | 2019-07-31 | A kind of relevant database mixing isomery interrogation model and method towards big data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110427437A true CN110427437A (en) | 2019-11-08 |
Family
ID=68411748
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910701575.9A Pending CN110427437A (en) | 2019-07-31 | 2019-07-31 | A kind of relevant database mixing isomery interrogation model and method towards big data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110427437A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110413651A (en) * | 2019-08-13 | 2019-11-05 | 中科驭数（北京）科技有限公司 | The Connection inquiring method and device of Relational DBMS |
| CN111104426A (en) * | 2019-11-22 | 2020-05-05 | 深圳智链物联科技有限公司 | Data query method and system |
| CN111324603A (en) * | 2019-12-27 | 2020-06-23 | 湖南星汉数智科技有限公司 | Grouping-based database processing method and device, computer device and computer readable storage medium |
| CN111914066A (en) * | 2020-08-17 | 2020-11-10 | 山东合天智汇信息技术有限公司 | Multi-source database global search method and system |
| CN113868249A (en) * | 2021-09-23 | 2021-12-31 | 广东电网有限责任公司 | Data storage method and device, computer equipment and storage medium |
| CN115061952A (en) * | 2022-08-19 | 2022-09-16 | 飞狐信息技术（天津）有限公司 | Data caching method and device, electronic equipment and computer storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102063486A (en) * | 2010-12-28 | 2011-05-18 | 东北大学 | Multi-dimensional data management-oriented cloud computing query processing method |
| US20150363467A1 (en) * | 2013-01-31 | 2015-12-17 | Hewlett-Packard Development Company, L.P. | Performing an index operation in a mapreduce environment |
| CN109800222A (en) * | 2018-12-11 | 2019-05-24 | 中国科学院信息工程研究所 | A kind of HBase secondary index adaptive optimization method and system |
-
2019
- 2019-07-31 CN CN201910701575.9A patent/CN110427437A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102063486A (en) * | 2010-12-28 | 2011-05-18 | 东北大学 | Multi-dimensional data management-oriented cloud computing query processing method |
| US20150363467A1 (en) * | 2013-01-31 | 2015-12-17 | Hewlett-Packard Development Company, L.P. | Performing an index operation in a mapreduce environment |
| CN109800222A (en) * | 2018-12-11 | 2019-05-24 | 中国科学院信息工程研究所 | A kind of HBase secondary index adaptive optimization method and system |
Non-Patent Citations (1)
| Title |
|---|
| 骆涛: "面向大数据处理的并行计算模型及性能优化", 《中国博士学位沦为全文数据库 信息科技辑》 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110413651A (en) * | 2019-08-13 | 2019-11-05 | 中科驭数（北京）科技有限公司 | The Connection inquiring method and device of Relational DBMS |
| CN110413651B (en) * | 2019-08-13 | 2020-12-08 | 中科驭数（北京）科技有限公司 | Connection query method and device for relational database management system |
| CN111104426A (en) * | 2019-11-22 | 2020-05-05 | 深圳智链物联科技有限公司 | Data query method and system |
| CN111104426B (en) * | 2019-11-22 | 2024-04-05 | 深圳智链物联科技有限公司 | Data query method and system |
| CN111104426B8 (en) * | 2019-11-22 | 2024-04-23 | 北京傲速科技有限公司 | Data query method and system |
| CN111324603A (en) * | 2019-12-27 | 2020-06-23 | 湖南星汉数智科技有限公司 | Grouping-based database processing method and device, computer device and computer readable storage medium |
| CN111324603B (en) * | 2019-12-27 | 2023-04-14 | 湖南星汉数智科技有限公司 | Packet-based database processing method and device, computer device and computer readable storage medium |
| CN111914066A (en) * | 2020-08-17 | 2020-11-10 | 山东合天智汇信息技术有限公司 | Multi-source database global search method and system |
| CN111914066B (en) * | 2020-08-17 | 2024-02-02 | 山东合天智汇信息技术有限公司 | Global searching method and system for multi-source database |
| CN113868249A (en) * | 2021-09-23 | 2021-12-31 | 广东电网有限责任公司 | Data storage method and device, computer equipment and storage medium |
| CN115061952A (en) * | 2022-08-19 | 2022-09-16 | 飞狐信息技术（天津）有限公司 | Data caching method and device, electronic equipment and computer storage medium |
| CN115061952B (en) * | 2022-08-19 | 2022-12-27 | 飞狐信息技术（天津）有限公司 | Data caching method and device, electronic equipment and computer storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110427437A (en) | A kind of relevant database mixing isomery interrogation model and method towards big data | |
| Lawrence | Integration and virtualization of relational SQL and NoSQL systems including MySQL and MongoDB | |
| Cattell | Scalable SQL and NoSQL data stores | |
| CN105487818B (en) | For the efficient De-weight method of repeated and redundant data in cloud storage system | |
| Tatarowicz et al. | Lookup tables: Fine-grained partitioning for distributed databases | |
| Gajendran | A survey on nosql databases | |
| CN108664516A (en) | Enquiring and optimizing method and relevant apparatus | |
| CN108595664A (en) | A kind of agricultural data monitoring method under hadoop environment | |
| CN108536692A (en) | A kind of generation method of executive plan, device and database server | |
| CN109446358A (en) | A kind of chart database accelerator and method based on ID caching technology | |
| Xiong et al. | Data vitalization: a new paradigm for large-scale dataset analysis | |
| Borkar et al. | Have your data and query it too: From key-value caching to big data management | |
| CN104123356A (en) | Method for increasing webpage response speed under large data volume condition | |
| Kuzochkina et al. | Analyzing and Comparison of NoSQL DBMS | |
| Matei et al. | Column-oriented databases, an alternative for analytical environment | |
| Maroulis et al. | Resource-aware adaptive indexing for in situ visual exploration and analytics | |
| CN106462591A (en) | Partition filtering using smart index in memory | |
| CN113722274A (en) | Efficient R-tree index remote sensing data storage model | |
| CN107273443B (en) | Mixed indexing method based on metadata of big data model | |
| Li et al. | Aggregate nearest keyword search in spatial databases | |
| D’silva et al. | Secondary indexing techniques for key-value stores: Two rings to rule them all | |
| Nascimento et al. | Access structures for moving points | |
| Liu | Data warehousing technologies for large-scale and right-time data | |
| Arnold et al. | HRDBMS: Combining the best of modern and traditional relational databases | |
| Luo | Partial materialized views |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191108 |
|
| RJ01 | Rejection of invention patent application after publication |