Relational database

In this article DigitalOcean In the article , We will try to understand some of the most commonly used 、 The most popular relational database management system （RDBMS） Kernel differences . We will explore the bottom differences —— Features and functions , How they work , In which way is better , To help programmers choose the right RDBMS.

Database management system
A database is a space for storing model data in an organized way , Store various types of information （ data ）. Every database , Except for modeless , There is a model , Provide a structural description of the data . Database management system is to manage database structure 、 Application of size and sorting （ Or library ）.

Relational database management system
The relational database system realizes the relational model , And use it to process data . A relational model associates information with fields in a table （ That is to say schemas）, To store data .

This kind of database management system needs structure （ Such as table ） Defined before storing data . With a watch , Each column （ Field ） Store a different type （ data type ） Information about . Every record in the database , All have their own unique key, As a row belonging to a table , Each information in the row corresponds to a column in the table —— All relationships together , Form a relational model .

Relationships and data types
Relationships can be seen as a mathematical set containing a series of attributes that collectively represent the maintained database and related information . This type of identification and collection method allows relational databases to operate in their own way .

When defining a table into which data can be inserted , Every element that forms a record （ for example ： attribute ） Must match the defined data type ( for example ： One integer, One date wait .). Different relational database management systems realize different data types – They don't always convert directly to each other .

Collaboration with constraints , As we have already introduced , It is very common in the use of relational databases . in fact , Restrictions form the core of the relationship .

Be careful : If you need it, it doesn't really matter , Random data （ For example, a document ）, You may use NoSQL Interested in ( Modeless database ).

Important and popular relational databases
In this paper , We will introduce three main and important open source relational database management systems , It is they who have influenced the world of application development .

SQLite:

A powerful embedded relational database management system .

MySQL:

One of the most popular RDBMS.

PostgreSQL:

Most advanced SQL Open source objective-RDBMS.

notes : Open source applications are always free to use . Most of the time , Replication Engineering （ Using code ） Creating new applications is also allowed . If you are right about DBMS Interested in , You can see some branch projects based on these projects , for example MariaDB.

SQLite
SQLite Is an extraordinary database , He can process in the application that uses it . As a self-contained 、 File based database ,SQLite Provides an excellent tool set , Can handle all types of data , There are no restrictions , And it is much easier to use than the process server running on the server .

An application uses SQLite when , Its functions are directly integrated into it , The application will directly access the file containing the data （ namely SQLite database ）, Not through some ports （port, socket） To interact . Thanks to this underlying technology , This makes SQLite Become very fast and efficient , And very powerful .

SQLite Supported data types
NULL:

NULL value .

INTEGER:

Signed integers , Use 1、2、3、4、6 or 8 Byte storage .

REAL:

Floating point numbers , Use 8 byte IEEE Floating point storage .

TEXT:

Text string , Use database encoding to store （UTF-8, UTF-16BE or UTF-16LE）.

BLOB:

Binary big object , Store as you type .

notes : Want to learn more about SQLite Data type information , You can view the Official documents .

SQLite The advantages of
Based on the file :

The entire database is contained in a file on disk , So it has good mobility .

Standardization :

Although it looks like a “ Simplified edition ” The database of ,SQLite I do support SQL. It omits some functions (RIGHT OUTER JOIN and FOR EACH STATEMENT), however , At the same time, some other functions are added .

Great for development and even testing :

In the development stage of most applications , Most people really need the flexibility of concurrency in the solution .SQLite Rich functional foundation , It can provide more than the development needs , And it's so simple that you only need one file and one C Link library .

SQLite The shortcomings of
No user management :

Advanced databases can support user systems , for example , It can manage the access rights of database connections to databases and tables . But because of SQLite The purpose and nature of the production （ There is no multi-user concurrent high-level design ）, It doesn't have this function .

Lack of additional flexibility to optimize performance ：

Still from the beginning of design ,SQLite It does not support the use of various techniques for additional performance optimization . This library is easy to configure , Easy to use . Since it is not complicated , Theoretically, it can't be faster than now , In fact, it has been very fast now .

When to use SQLite ？
Embedded applications :

All need mobility , Applications that do not need expansion , for example , Single user local application , Mobile apps and games .

Instead of disk access :

In many cases , Need frequent direct reading / Application of writing disk files , Are very suitable for use SQLite , Can benefit from SQLite Use SQL Bring functionality and simplicity .

test :

It can spike most of the application business logic （ That is, the main purpose of the application ： Can complete the function ） Test of .

When not to SQLite ？
Multi user applications :

If the application you are developing needs to be accessed by multiple users , And these users all use the same database , So compared SQLite It is better to choose a relational database with complete functions （ for example MySQL）.

Applications that need to write data in a large area :

SQLite One of the defects of is its write operation . This database allows only one write operation at a time , Therefore, the throughput is limited .

MySQL
MySQL The most popular of all large database servers . It has rich characteristics , The open source nature of the product makes it drive a large number of online websites and Applications . To start MySQL Relatively simple , Developers can access a lot of information about this database on the Internet .

Be careful : Because of the popularity of this product , A large number of third-party applications 、 Tools and integration libraries are important for operating this RDBCMS It's very helpful in all aspects .

Mysql No attempt to achieve SQL All of the Standards , Instead, it provides users with many useful functions . As an independent database server , The application is the same as Mysql Interaction of daemons , Tell it to access the database itself – It's not like SQLite.

MySQL Supported data types
TINYINT:

A very small integer .

SMALLINT:

A small integer .

MEDIUMINT:

An integer of intermediate size .

INT or INTEGER:

An integer of normal size .

BIGINT:

A large integer .

FLOAT:

A small ( Single precision ) Floating point numbers , It can't be unsigned .

DOUBLE, DOUBLE PRECISION, REAL:

A normal size ( Double precision ) Floating point number , You can't make the unsigned one .

DECIMAL, NUMERIC:

Floating point numbers that are not wrapped . You can't make the unsigned one .

DATE:

A date .

DATETIME:

A combination of date and time .

TIMESTAMP:

A timestamp .

TIME:

A time .

YEAR:

One with two or 4 The year in digit format ( The default is 4 position ).

CHAR:

A fixed length string , When storing, it is always right aligned in its fixed length space .

VARCHAR:

A variable length string .

TINYBLOB, TINYTEXT:

One BLOB perhaps TEXT Column , Maximum length 255 (2^8 - 1) Characters .

BLOB, TEXT:

One BLOB perhaps TEXT Column , Maximum length 65535 (2^16 - 1) Characters .

MEDIUMBLOB, MEDIUMTEXT:

One BLOB perhaps TEXT Column , Maximum length 16777215 (2^24 - 1) Characters .

LONGBLOB, LONGTEXT:

One BLOB perhaps TEXT Column , Maximum length 4294967295 (2^32 - 1) Characters .

ENUM:

An enumerated type .

SET:

A collection .

MySQL The advantages of
Easy to use :

install MySQL Very easy to . Third party Library , Including visualization （ There is a GUI） The library of makes it very easy to use the database .

Rich in functions :

MySQL Support most relational databases should have SQL function —— Some direct support , Some indirect support .

Security :

MYSQL There are many security features , Some of them are quite advanced .

Flexible and powerful :

MySQL Can handle a lot of data , In addition, if necessary , It can also “ To adapt to ” Data of all sizes .

Fast :

Abandon support for certain standards , Give Way MySQL More efficient and able to use shortcuts , Therefore, the speed is improved .

MySQL The shortcomings of
Known limitations :

From the beginning of design ,MySQL I didn't plan to be omniscient , So it has some functional limitations , Unable to meet the needs of some top-level applications .

Reliability issues :

MySQL For the implementation of some functions ( for example , quote , Business , Data review, etc ) This makes it a little less reliable than some other relational databases .

Development stalled :

Even though MySQL Theoretically, it is still an open source product , Some people complain that it has been updated slowly since its birth . However , It should be noted that some are based on MySQL And a completely integrated database （ Such as MariaDB）, In standard MySQL On this basis, it brings additional value .

When to use MySQL？
Distributed operations :

When SQLite When what you provide cannot meet your needs , You can put MySQL Include in your deployment stack , Just like any independent database server , It will bring a lot of freedom of operation and some advanced functions .

High security :

MySQL Safety function of , In a simple way for data access （ And use ） Provides reliable protection .

Web Website and Web application :

Most websites （ and Web Applications ） You can simply work in MySQL On . This flexible and extensible tool is easy to use and easy to manage —— This proved to be very helpful for long-term operation .

Custom solutions :

If you work on a highly customized solution ,MySQL Can easily follow and enforce your rules , Thanks to its rich configuration settings and operation modes .

When not to MySQL？
SQL Obedience :

because MySQL No, [ to want to ] Realization SQL All standards , So this tool is not completely consistent with SQL. If you need to integrate such a relational database management system , from MySQL Switching is not easy .

Concurrent :

Even if MySQL And some storage engines are really good at reading , But there are still problems with concurrent reading and writing .

Lack of characteristics :

Again , According to the selection criteria of database engine ,MySQL Will lack certain characteristics , Such as full-text search .

PostgreSQL
PostgreSQL It's an advanced , Open source [ object ]- Relational database management system , Its main goal is to achieve standards and scalability . PostgreSQL, Or rather, Postgres, Try to put right ANSI/ISO SQL The adoption and revision of standards are combined .

Compared with other RDBMS, PostgreSQL With it for objects - Relationship and \ Or relational database function , For example, for reliable transactions , For example, atomicity , Uniformity , Isolation and persistence （ACID） Full support for , The high demand and collective support of these things , To show its uniqueness .

Due to the powerful underlying technology , Postgres It is good at efficiently completing many processing tasks . Thanks to its multi version concurrency control (MVCC) The implementation of the , Concurrency can be achieved without reading locks , This also ensures ACID Implementation .

PostgreSQL It's highly programmable , Therefore, it can be called “ stored procedure ” The custom program of . These functions can be created to simplify a write repetition 、 The execution of tasks that are complex and often require database operations .

Although the features are powerful , But this DBMS did not MySQL So popular , But there are still many fascinating third-party tools and libraries designed to make PostgreSQL It's easy to operate . Nowadays, it is easy to get through the default package manager of many operating systems PostgreSQL Has become possible .

PostgreSQL Supported data types
bigint:

Signed octet integer

bigserial:

Self increasing octet integer

bit [(n)]:

Fixed length bit string

bit varying [(n)]:

Variable length bit strings

boolean:

Logical Boolean (true/false)

box:

A rectangular box on a plane

bytea:

binary data (“ Digit group ”)

character varying [(n)]:

Variable length string

character [(n)]:

Fixed length string

cidr:

IPv4 perhaps IPv6 network address

circle:

A circle on the plane

date:

Calendar date ( Specific date )

double precision:

Double precision floating point （8 position ）

inet:

IPv4 perhaps IPv6 The host address

integer:

Signed four digit integer

interval [fields] [§]:

time span

line:

An infinite straight line on the plane

lseg:

A line segment on the plane

macaddr:

MAC ( Media access control ) Address

money:

The amount of money

numeric [(p, s)]:

Precise numbers with optional precision

path:

Geometric path on a plane

point:

A geometric point on a plane

polygon:

A closed geometric path on a plane

real:

Single-precision floating-point (4 position )

smallint: Signed two digit integer

serial:

Self growth 4 An integer

text:

Variable length character creation

time [§] [without time zone]:

Time of day ( No time zone )

time [§] with time zone:

Time of day , Include time zone

timestamp [§] [without time zone]:

Date and time （ No time zone ）

timestamp [§] with time zone:

Date and time , Include time zone

tsquery:

Text search query

tsvector:

Text search documents

txid_snapshot:

User level transactions ID snapshot

uuid:

Universal unique identifier

xml:

XML data

PostgreSQL The advantages of
Standard support SQL Open source relational database :

PostgreSQL It's an open source , free (free) Of , At the same time, a very powerful relational data management system .

A strong community :

PostgreSQL There is a passionate and experienced community behind it , Support can be obtained through knowledge base and Q & a website , Free 24 / 7 .

Strong third-party support :

Even if its own function is very powerful ,PostgreSQL There are still many powerful open source third-party tools to assist the design of the system 、 Management and use .

Extensibility :

You can use pre stored processes to programmatically extend PostgreSQL , An advanced relational database should be like this .

object-oriented :

PostgreSQL Not just a relational database , It is also an object-oriented database —— Support nested , And some other functions .

PostgreSQL The shortcomings of ：
performance :

For simple and heavy read operations , PostgreSQL Will make a mountain out of a molehill and may appear better than peers （ Such as MySQL） Lower performance .

Universal :

According to the nature of the tool , In terms of popularity, it still lacks sufficient backstage support , Despite a large number of deployments —— This may affect the ease with which support can be obtained .

trusteeship :

Due to the influence of the above factors , Let the host or service provider propose to use PostgreSQL Examples are difficult .

When to use PostgreSQL？
Data integrity :

When reliability and data integrity are absolutely necessary without reason ,PostgreSQL It's a better choice .

Complex customization process :

If you need your database to perform a custom process , Extensible PostgreSQL It's a better choice .

Integrate :

In the future , If possible, migrate the entire database system to another appropriate solution （ for example Oracle） in ,PostgreSQL For this kind of switching, it will be the most compatible and easy to operate .

Complex design :

Compared with other open source and free RDBMS（ Relational database management system ） To achieve , For complex database design ,PostgreSQL Provides most of the functions and possibilities , At the same time, I didn't give up other valuable places .

When not to PostgreSQL？
Speed :

If all you need is a fast read operation , PostgreSQL Not a tool for this .

Simplify the system :

Unless you need absolute data integrity , Atomicity , Uniformity , Isolation, , Durability , Or complex design ,PostgreSQL It is a killer for simplifying the system .

Copy :

Unless you are willing to spend a lot of time , Energy and resources , Otherwise, for those who lack experience in database and system management , Realization and MySQL Of （ Master-slave ） Copying may not be easy .