CN106446182A

CN106446182A - Database engine

Info

Publication number: CN106446182A
Application number: CN201610860097.2A
Authority: CN
Inventors: 黄仕忠
Original assignee: Guizhou Media Education Technology Co Ltd
Current assignee: Guizhou Media Education Technology Co Ltd
Priority date: 2016-09-29
Filing date: 2016-09-29
Publication date: 2017-02-22

Abstract

The invention discloses a database engine. The database engine comprises a load balancing data fetching device and a database entity bean, and is characterized in that multiple data source pools which can be connected with databases of different types are established through data source configuration files, data source names are used to obtain data connection from the data source pools, the data connection is put in a thread data connection pool, and the data source pools and the data connection pool are successively connected with a basic sql execution layer, a persistent object sql conversion layer and an external operation interface layer. By adopting the database engine, the higher SQL execution efficiency can be reached, and unnecessary repeated database connection and related operations of jdbc processing can be reduced to gain time and efficiency for developers.

Description

A kind of database engine

Technical field

The present invention relates to a kind of database engine, belongs to database technical field.

Background technology

In Enterprise Project exploitation, in order to development efficiency is improved, reduce development cost, can cut in terms of database processing, Current conventional database processing scheme has two kinds：One is that this scheme relies on bag too much using the ORM that increases income, such as Hibernate, And left-hand seat difficulty is big, efficiency is also less high；Two are operated using pure JDBC, although this scheme execution efficiency highest, but When processing complex data table, developer will remember field name or sequence number, it is easy to malfunction, and be difficult investigation after mistake, And, current data processing scheme, the connection of substantial amounts of repeated data storehouse and jdbc operation is there is mostly, for this purpose, A set of database manipulation engine should be developed in conjunction with actual demand, reduce unnecessary repeated data storehouse connection and jdbc processes phase Operation is closed, is that developer gains time and efficiency.

Content of the invention

This practical technical problem to be solved is to provide a kind of database engine, can reach higher SQL execution efficiency, Unnecessary repeated data storehouse connection can be reduced and jdbc processes associative operation, be that developer gains time and efficiency, overcome The deficiencies in the prior art.

The technical scheme is that：A kind of database engine, which includes that load balancing is fetched data device, database entity Bean, by data source profile, sets up multiple data source ponds that can be connected with different types of data base, with data source Title obtains data cube computation from data source pond, and data cube computation is put into thread-data connection pool, described data source pond and number It is connected with basic sql execution level, persistence object sql conversion layer, peripheral operation interface layer according to connection pool successively.

Same connection is only existed in the same thread of above-mentioned thread-data connection pool.

In above-mentioned entity bean, data base's table name and literary name section are parsed with reflex mechanism.

The data base created by upper data source pond is linked back data source pond by the load balancing device that fetches data.

Prior art compares, the database engine of the present invention, by data source profile, sets up multiple energy and inhomogeneity The data source pond that the data base of type is connected, obtains data cube computation with DSN from data source pond, data cube computation is put Enter thread-data connection pool, a light-weighted underlying database operation framework can be so set up, is reached higher SQL and execute effect Rate, reduces unnecessary repeated data storehouse connection and jdbc processes associative operation, it is often more important that, engine is carried in same application Operate for multi-data source, be data base's solution of the High Availabitity such as data-base cluster, read and write abruption, database loads equilibrium, high capacity Certainly scheme provides the support of program application level, is that developer gains time and efficiency.

Described data source pond and Data Connection Pool successively with basic sql execution level, persistence object sql conversion layer, outer Portion's operate interface layer is connected, so during database table is automatically created：Obtained from data source pond by DSN Data cube computation, and thread-data connection pool is put into, same connection is only existed in same thread, is ensured with this same The affairs of thread request are submitted to, and the sql sentence of hoc queries data table information calls the interface polls of peripheral operation interface layer to go out All table structural informations in data base, loading of databases entity bean, the data of entity bean are parsed by reflex mechanism Storehouse table name and literary name section, judge the data base's table name with the presence or absence of the entity beanr for parsing in data base, do not exist, execute Operation is created, according to the data base's table name for parsing and field, assembling creates the sql sentence of database table, calls peripheral operation The interface of interface layer, executes sql sentence by the connection in line taking journey connection pool, completes database table establishment；Preserving data During：The content for needing to preserve is put in entity bean, the table structure of entity bean is parsed by reflex mechanism and is filled out Content is filled, group will become the insertion sql sentence of database table, the data cube computation in thread pool is obtained by DSN, is passed through The data cube computation of acquirement executes the insertion sql sentence of database table, completes data and preserves warehouse-in；In inquiry data procedures：Group Dress inquiry sql sentence, obtains the data cube computation in thread pool by DSN, executes data by the data cube computation for obtaining The inquiry sql sentence of storehouse table, Query Result is assembled into corresponding list of entities.

Same connection is only existed in the same thread of above-mentioned thread-data connection pool, so can be in same thread mistake When to obtain connection, it is ensured that got is same connection, ensure that with this affairs of same thread request are submitted to.

In above-mentioned entity bean, data base's table name and literary name section are parsed with reflex mechanism, can so be easy to repercussion The relevant information of database table and field is generated, most importantly field is explained and also generated in the lump.

The data base created by upper data source pond connects back to data source pond by the load balancing device that fetches data, and so may not be used Disconnected reservation and expanding data storehouse, meanwhile, load balancing fetch data device using a plurality of network link sharing data stream can be realized Amount.

Description of the drawings

Fig. 1 is the attachment structure schematic diagram of the present invention；

Specific embodiment

Embodiment 1. is as shown in figure 1, a kind of database engine, and which includes that load balancing is fetched data device, database entity Bean, it is characterised in that;By data source profile, multiple data sources that can be connected are set up with different types of data base Pond, obtains data cube computation with DSN from data source pond, and data cube computation is put into thread-data connection pool, described number It is connected with basic sql execution level, persistence object sql conversion layer, peripheral operation interface layer with Data Connection Pool successively according to source pond, Same connection is only existed in the same thread of thread-data connection pool, in entity bean, number is parsed with reflex mechanism According to storehouse table name and literary name section, the data base created by data source pond connects back to data source pond by the load balancing device that fetches data.

Its operation process

1. database table process is automatically created:

1. database entity bean is set up, and in entity, adds self-defining table name and field to explain.

2. pass through data source profile, multi-data source pond is set up, different types of number can be used in same application According to storehouse.

3. data cube computation is obtained from data source pond by DSN, and thread-data connection pool is put into, same Same connection is only existed in individual thread, ensures that with this affairs of same thread request are submitted to.

4. the sql sentence of hoc queries data table information, calls the interface polls of peripheral operation interface layer to go out in data base All table structural informations.

5. loading of databases entity bean, parses data base's table name and the literary name section of entity bean by reflex mechanism.

6. judge the data base's table name with the presence or absence of the entity beanr for parsing in data base.Do not exist and then execute establishment Operation.

7. according to the data base's table name for parsing and field, assembling creates the sql sentence of database table.

8. the interface of peripheral operation interface layer is called, sql sentence is executed by the connection in line taking journey connection pool, complete number Create according to storehouse table.

2. data procedures are preserved：

1. entity bean is set up.

2. the content for needing to preserve is put in entity bean.

3. table structure and the filling content of entity bean are parsed by reflex mechanism.

4. group will become the insertion sql sentence of database table.

5. the data cube computation in thread pool is obtained by DSN.

6. the data cube computation by obtaining executes the insertion sql sentence of database table.

7. data are completed and preserves warehouse-in.

3. data procedures are inquired about：

1. sql sentence is inquired about in assembling.

2. the data cube computation in thread pool is obtained by DSN.

3. the data cube computation by obtaining executes the inquiry sql sentence of database table.

4. Query Result is assembled into corresponding list of entities.

Claims

1. a kind of database engine, which includes that load balancing is fetched data device, database entity bean, it is characterised in that;By number According to source profile, multiple data source ponds that can be connected are set up with different types of data base, with DSN from data Obtain data cube computation in the pond of source, data cube computation be put into thread-data connection pool, described data source pond and Data Connection Pool according to Secondary with basic sql（sql）Execution level, persistence object sql conversion layer, peripheral operation interface layer are connected.

2. the database engine according to right 1 is required, it is characterised in that：The same thread of the thread-data connection pool In only exist same connection.

3. the database engine according to right 1 is required, it is characterised in that：In described entity bean, with reflex mechanism Parse data base's table name and literary name section.

4. the database engine according to right 3 is required, it is characterised in that：Data base (the data base created by data source pond Connection) being fetched data by load equalizer connects back to data source pond.