Oracle Tips  by Burleson Consulting

The Data Warehouse Development Life Cycle

Parallelism And Oracle Data Warehousing Parallel Processing
Before we get too deep into our discussion of parallel processing, a distinction needs to be made between multitasking and multiprocessing. Multitasking refers to the ability of a software package to manage multiple concurrent processes, thereby allowing parallel processing. OS/2 and Windows-NT are examples of this technology, but multitasking is found within all midrange and mainframe databases. Multiprocessing, on the other hand, refers to the use of multiple CPUs within a distributed environment where a master program directs parallel operations against numerous machines. There are two areas of multiprocessing: hardware and software. At the hardware level, arrays of CPUs are offered, and at the software level, a single CPU is partitioned into separate "logical" processors. The Prism software in the IBM mainframe environment is an example of multiprocessing technology.

In any case, programming for multiprocessors is quite different from programming for linear systems. Multiprocessing programming falls into two arenas: data parallel programming and control parallel programming. Indata parallel programming the data is partitioned into discrete pieces and the same program is run in parallel against each piece (see Figure 7.1). Data parallel can be used in any type of Oracle environment, including uniprocessor, symmetric multiprocessor (SMP), or massively parallel processor (MPP) environments. With control parallel programming, independent functions are identified and independent CPUs are used to simultaneously solve each independent function. Control parallel is generally associated with MPP environments.

Figure 7.1 An example of parallel processing.

One of the greatest problems faced by developers when implementing parallel processing systems is the identification of parallelism. Parallelism refers to the ability of a computer system to perform processing on two data sources at the same time. Whereas many of the traditional database applications were linear in nature, today's systems have ample opportunities for parallel processing.

---

This is an excerpt from "High Performance Data Warehousing". To learn more about Oracle, try "Oracle Tuning: The Definitive Reference", by Donald K. Burleson.  You can buy it direct from the publisher at 30% off here:
http://www.rampant-books.com/book_1002_oracle_tuning_definitive_reference_2nd_ed.htm