Fujitsu 5.0L10 Manual Do Utilizador

Página de 225
Chapter 4 Storage Structure
 
From the application program viewpoint, database data is represented in table format. The application program 
manipulates data as if it were manipulating rows and columns of a table by using structured query language (SQL) 
statements. 
The structure for storing data represented in table format on physical pages is called the storage structure. The 
storage structure cannot be directly seen from the application program. Regardless of the storage structure used, from 
the application program viewpoint, the table rows and columns appear as if they are being manipulated according to 
SQL statements. 
However, since the physical data is stored according to the storage structure, the storage structure is an important 
factor in determining processing efficiency. 
Ignoring the interrelationships between transactions, such as exclusion, and focusing on storage structure, one can 
see the following elements affecting the data processing efficiency of an application program: 
 
·  Addition of an index 
Add an index for a table. 
 
·  Allocation of database space 
Carefully consider the amount of data to be processed and area usage patterns. Allocate database space for 
each component of the storage structure. 
 
·  Association with the shared buffer pool 
Select a page size and shared buffer pool appropriate to the data processing. 
This chapter explains the features of storage structures. The explanation covers the following topics: 
4.1 Features of Table Storage Structures 
4.2 Features of the Index Storage Structure 
4.3 Allocating Space 
4.4 Estimating the Required Amount of Database Space 
4.1 Features of Table Storage Structures
 
Specify the storage structure according to a data structure organization (DSO) definition. The three types of storage 
structures are SEQUENTIAL, RANDOM, OBJECT, and BTREE. The SEQUENTIAL, RANDOM, and OBJECT 
structures are used as storage structures for tables. The BTREE structure is used as the storage structure for indexes. 
This section explains the features of the table storage structures and the data processing appropriate to those 
structures. 
4.1.1 SEQUENTIAL structure
 
Data is stored in a SEQUENTIAL structure in the order that the data is inserted. 
Figure: Overview of SEQUENTIAL structure shows an overview of the SEQUENTIAL structure, using the STOCK 
table as an example. 
139