Introduction to Oracle Parallel Query Oracle Tips by Burleson Consulting

Oracle has implemented parallel query features that allow a query to effectively parallelize queries with both symmetric multiprocessing (SMP) and massively parallel processing (MPP) architectures. Using these parallel features on a massively parallel machine, it is possible to read a one-gigabyte table with subsecond response time. Let's begin with a review of these architectures.

Note: Oracle parallel query works only with databases that perform full-table scans. A well-tuned online transaction database will seldom perform full-table scans and will not benefit from Oracle parallel query.

As I stated, Oracle parallel query will only work with queries that perform a full-table scan, and it is very important that the Remote DBA understand that indexes are the enemy of parallel query. To invoke parallel query, you must also force the SQL optimizer to perform a full-table scan. Hence, it follows that Oracle parallel query will only improve queries that must read the majority of the data blocks in a table.

Oracle parallel query achieves improved speed because multiple processes can be directed to read a table. Parallel query works best on servers that have multiple CPUs because multiple CPUs allow for simultaneous queries. A later section in this chapter will show you how to see how many CPUs you have on your database server.

Starting with Oracle release 7.2, you can partition a SQL query into subqueries and dedicate separate processors to each one. Here's how it works: Instead of having a single query server to manage the I/O against the table, parallel query allows the Oracle query server to dedicate many processes to simultaneously access the whole table (see Figure 10-1).

Figure 1: An Oracle parallel query with full-table scan

Let’s take a look at the techniques for invoking parallel query and determining the optimal degree of parallelism.

Invoking Oracle Parallel Query

There are several init.ora parameters to be set when using Oracle parallel query. Many of these are default values and are set by Oracle when your database is created. Oracle parallel query can be turned on in several ways. You can turn it on permanently for a table, or you can isolate the parallel query to a single table.

Permanent parallelism (Not recommended for cost-based optimization)

Alter table customer parallel degree 35;

Single query parallelism

select /*+ FULL(emp) PARALLEL(emp, 35) */
         emp_name
      from
        emp;

Note the use of the double hints in the preceding query. Most Oracle Remote DBAs always use the full hint with the parallel hint because they are both required to use Oracle parallel query.

WARNING: It is very dangerous to enable Oracle parallel query for a table with the alter table command. Once a table is marked for parallel query, the CBO may change the execution plan for existing queries to use parallel full-table scans instead of index scans. This well-intentioned mistake has crippled many databases, since queries that used to run fast with indexes will now use a full-table scan.

Most Oracle Remote DBAs identify those tables that perform full-table scans and then add hints to specify the degree of parallelism for the query. This way, all full-table scans against the tables will invoke Oracle parallel query. It can be very dangerous to turn on parallelism with the alter table command because the CBO may change other queries' execution plans to perform full-table scans based on the new setting for parallelism. This could adversely affect hundreds of SQL statements.

This is an excerpt from "Oracle High-Performance SQL Tuning" by Donald K. Burleson, published by Oracle Press.