Data access on DB2 UDB for iSeries: data access paths and methods

 

Data access methods are used to process queries and access data.

In general, the query engine has two kinds of raw material with which to satisfy a query request:

• The database objects that contain the data to be queries

• The executable instructions or operations to retrieve and transform the data into usable information

There are actually only two types of permanent database objects that can be used as source material for a query — tables and indexes (binary radix and encoded vector indexes). In addition, the query engine may need to create temporary objects or data structures to hold interim results or references during the execution of an access plan. The DB2^® UDB Symmetric Multiprocessing feature provides the optimizer with additional methods for retrieving data that include parallel processing. Finally, the optimizer uses certain methods to manipulate these objects.

• Permanent objects and access methods
  The database objects and access methods used by the query engine can be broken down into three basic types of operations that are used to manipulate the permanent and temporary objects -- Create, Scan, and Probe.

• Temporary objects and access methods
  Temporary objects are created by the optimizer in order to process a query. In general, these temporary objects are internal objects and cannot be accessed by a user.

• Objects processed in parallel
  The DB2 UDB Symmetric Multiprocessing feature provides the optimizer with additional methods for retrieving data that include parallel processing. Symmetrical multiprocessing (SMP) is a form of parallelism achieved on a single server where multiple (CPU and I/O) processors that share memory and disk resource work simultaneously toward achieving a single end result.

• Spreading data automatically
  DB2 Universal Database™ for iSeries™ automatically spreads the data across the disk devices available in the auxiliary storage pool (ASP) where the data is allocated. This ensures that the data is spread without user intervention.

 

Parent topic:

Performance and query optimization