Physical design of database design (2)

What does physical design do

One . Choose the right database management system  

Oracle、SqlServer、MySql And PgSQL

• Copyright 、 Cost considerations ：

• Functional considerations ：

Oracle： A good one in the industry DBMS, A high performance , It is suitable for large transaction operations .

• Operating system considerations ：

SqlServerL Only support Windows Under the system

Oracle、MySql And PgSQL： Can support linux System 、windows System

• Development language considerations ：

If you use .net development language , Use SQLServer better

• Application scenario considerations ：

mysql And PgSQL Applicable to Internet projects

Oracle、SqlServer Applicable to enterprise level projects （Oracle, E.g. carrier level 、 financial , Because it has good scalability , High safety ,SqlServer： Small and medium-sized enterprises erp）

Introduce Mysql Common storage engines

 

Two 、 Define database 、 Naming conventions for tables and fields

All object naming principles ：

1. Readability principle

  Be careful ： Case naming （ hump ）, yes , we have DBMS Sensitive to case , such as Mysql

2. The principle of expressiveness

The name of the object should be able to describe the object it identifies .

for example ： Table name , It can reflect the stored data content

          For stored procedures , The stored procedure name should reflect the function of the stored procedure

3. Long name principle

Use as few or no abbreviations as possible .

For databases DB Any object other than the name

Such a name , When only looking at the name of the watch 、 Field name , I have no idea what the watch is for . During development and post operation and maintenance , Poor readability , Need to rely entirely on manuals .

3、 ... and . According to the selected DBMS The system selects the appropriate field type

  Selection of data type of column ：1. It will affect the cost of data storage space ,2. Affect the performance of data query

The principle of field type selection ：

  When a column can select multiple data types , Should be Number type is preferred , Next is the date or binary type , Finally, the character type .

For the same level of data type , Should be Give priority to data types that take up less space .

 

 Mysql stay MySQL For internal use 4 Bytes INT Type to store timestamp data , Timestamp type , It can only be stored in 2038 year , Because now 2020 Years. , Use it with caution . Empathy ,int Type to store time should also be used with caution ！

  The above principles are mainly considered from the following two perspectives ：

1. When comparing data （ Query criteria 、join Conditions and order ） When waiting for operation ： The same data , Character processing is often slower than number processing

 2. In the database , Data processing is in pages , The smaller the length of the column , Good for performance .

How to select field type in database ：

•  char And varchar How to choose ？

1. If the length of data to be stored in the column is almost the same , Consider using char; Otherwise, we should consider using varchar . Such as ： ID number , cell-phone number

2. If the maximum data length in the column is less than 50Byte, Generally, we also consider char（ If the column is rarely used , Based on saving space and reducing I/O The consideration of , Or choose to use varchar）

3. In general, it is not appropriate to define more than 50Byte Of char Type column .（ almost 15 Characters ）

• decimal And float How to choose

 1.decimal For storing accurate data , Two float Can only be used to store imprecise data .

2. because float The cost of storage space is generally higher than decimal Small , Therefore, imprecise data is preferred float type

• How time types are stored

1. Use int To store the advantages and disadvantages of time fields

advantage ： Field length ratio detetime Small

shortcoming ： Inconvenient to use , To convert a function

Limit ： It can only be stored in 2038-01-19 11:14:07, because 2^32 by 2147483648

2. Time granularity to store  

year 、 month 、 Japan 、 Hours 、 branch 、 second 、 Zhou

Four 、 Other considerations for database design ：

1. How to choose a primary key

2. Avoid using foreign key constraints

•   Reduce the efficiency of data import , The impact of high concurrency environment is very serious
• Increased maintenance costs
• However, foreign key constraints are not recommended , But the associated column must be indexed

3. Avoid using triggers

1. Reduce the efficiency of data import

2. Unexpected data exceptions may occur , Especially when business requirements change

3. Complicate business logic  

4. About reserved fields

 1. Unable to know exactly the type of reserved field .

2. Unable to accurately know the contents stored in the reserved field

3. The cost of maintaining the reserved fields later , The cost of adding a field is the same

4. strictly prohibit Use reserved fields

 

5、 ... and . Anti normal design

  For the sake of performance and reading efficiency , Appropriately violate the requirements of the third paradigm , and A small amount of data redundancy is allowed . Space for time .

 

Why anti paradigm

1. Reduce the number of table associations

2. Increase data reading efficiency

3、 The anti paradigm design must be moderate