RaqForum 25 No.
1 Reply  •  548 View  • 
     ...

Performance Optimization Practice

 
Performance Optimization Practice(38)

Data preparation

Below are structures of data tables used in this handbook:

Table Field Description
Categories [CategoryID] [int] NOT NULL,
[CategoryName] [nvarchar](50) NULL,
[Description] [nvarchar](100) NULL 		Category ID
Category name
Category description
Customers [CustomerID] [nvarchar](10) NOT NULL,
[CustomerName] [nvarchar](50) NOT NULL,
[ContactName] [nvarchar](50) NULL,
[ContactTitle] [nvarchar](50) NULL,
[Address] [nvarchar](50) NULL,
[City] [nvarchar](20) NULL,
[Region] [nvarchar](20) NULL,
[PostalCode] [nvarchar](20) NULL,
[Country] [nvarchar](20) NULL,
[Phone] [nvarchar](30) NULL,
[Fax] [nvarchar](30) NULL 		Customer ID
Customer name
Contact
Contact title
Customer address
City
Region
Postal code
Country
Telephone
Fax
Employees [EmployeeID] [int] NOT NULL,
[LastName] [nvarchar](50) NULL,
[FirstName] [nvarchar](50) NULL,
[Title] [nvarchar](50) NULL,
[TitleOfCourtesy] [nvarchar](50) NULL,
[Birthday] [date] NULL,
[HireDate] [date] NULL,
[Address] [nvarchar](50) NULL,
[City] [nvarchar](50) NULL,
[Region] [nvarchar](50) NULL,
[PostalCode] [nvarchar](50) NULL,
[Country] [nvarchar](50) NULL,
[HomePhone] [nvarchar](50) NULL,
[Gender] [nvarchar](50) NULL 		Employee ID
Last name
First name
Title
Courtesy title
Birthday
Hire date
Address
City
Region
Postal code
Country
Home phone
Gender
Orders [CustomerID] [nvarchar](10) NULL,
[OrderDate] [datetime] NULL,
[ProductID] [nvarchar](50) NOT NULL,
[Quantity] [int] NULL,
[Unit] [nvarchar](50) NULL,
[Price] [decimal](18, 2) NULL,
[Amount] [decimal](18, 2) NULL,
[EmployeeID] [int] NULL,
[EmployeeName] [string] NULL,
[ShipVia] [nvarchar](20) NULL 		Customer ID
Order date
Product ID
Quantity
Unit
Unit price
Amount of money
Employee ID
Employee name
Means of transportation
Products [ProductID] [nvarchar](50) NOT NULL,
[ProductName] [nvarchar](50) NOT NULL,
[SupplierID] [int] NULL,
[CategoryID] [int] NULL,
[QuantityPerUnit] [nvarchar](20) NULL,
[UnitPrice] [decimal](18, 2) NULL 		Product ID
Product name
Supplier ID
Category ID
Quantity per unit
Unit price
Suppliers [SupplierID] [int] NOT NULL,
[CompanyName] [nvarchar](50) NULL,
[ContactName] [nvarchar](50) NULL,
[ContactTitle] [nvarchar](50) NULL,
[Address] [nvarchar](100) NULL,
[City] [nvarchar](50) NULL,
[Region] [nvarchar](50) NULL,
[PostalCode] [nvarchar](50) NULL,
[Country] [nvarchar](50) NULL,
[Phone] [nvarchar](50) NULL,
[Fax] [nvarchar](50) NULL,
[HomePage] [nvarchar](100) NULL 		Supplier ID
Supplier name
Contact name
Contact title
Address
City
Region
Postal code
Country
Telephone
Fax
Website
Areas [Country] [nvarchar](50) NULL,
[Region] [nvarchar](50) NULL,
[City] [nvarchar](50) NULL,
[CityName] [nvarchar](50) NULL,
[PostalCode] [nvarchar](50) NULL 		Country
Region
City
City name
Postal code

Relationships between tables:

imagepng

As databases lack generality, we use text files to store the test data. Small data tables are directly stored in files, and large data tables can be generated based on script below.

Download data.zip

Table of contents

1 Using file storage

1.1 Bin file
1.2 Composite table
1.3 Converting date to small integer
1.4 Conversion from enumerated string to small integer

2 Understanding aggregation

2.1 COUNT DISTINCT
2.2 DISTINCT
2.3 Getting record(s) containing the max/min value
2.4 Getting top/bottom N values
2.5 Getting records containing top/bottom N
2.6 When there is a redundant grouping field
2.7 Getting the first/last record from each group

3 Order-based storage

3.1 Storing data in time order
3.2 Order-based filtering
3.3 Storing data in account order
3.4 Order-based DISTINCT
3.5 Order-based COUNT DISTINCT
3.6 Order-based grouping & summarization
3.7 Getting the first/last record from each group
3.8 Complicated judgment on each group

4 Dimension table pointed by foreign key

4.1 Storing dimension table in memory
4.2 Computations on a dimension table
4.3 Wide table
4.4 Numberized dimension table
4.5 Filtering on dimension table
4.6 Alignment sequence
4.7 Temp dimension table and segmented dimension table
4.8 External memory dimension table

5 Primary-key-based Association

5.1 Order-based merge
5.2 Computing intersection and union
5.3 Filtering sub table according to primary table
5.4 Filtering primary table according to sub table

6 Multi-purpose traversal

6.1 Basic concept
6.2 Application scenarios

7 Search tasks

7.1 Equivalent value Search
7.2 Interval-based search
7.3 Batch search
7.4 Full-text index

Performance Optimization Practice(38)
 
 •  548 View  •