Partitioning is transparent to applications, and no changes to SQL statements are required
to take advantage of partitioning. However, in situations where specifying a partition would be
advantageous, you can specify both the table name and the partition name in a SQL statement;
this improves both parse and select performance. Examples of syntax using explicit partition
names in a select statement are found later in this chapter, in the section ???Splitting, Adding,
and Dropping Partitions.???
Creating Partitioned Tables
Several methods of partitioning are available in the Oracle database, and some of these are new
to Oracle 10g, such as list-partitioned index-organized tables (IOTs); other methods are new to
Oracle 11g, such as composite list-hash, list-list, list-range, and range-range partitioning. In the
next few sections, we??™ll cover the basics of range partitioning, hash partitioning, list partitioning,
six types of composite partitioning, as well as interval partitioning, reference partitioning,
application-controlled partitioning, and virtual column partitioning. I??™ll also show you how
to selectively compress partitions within the table to save on I/O and disk space.
Using Range Partitioning Range partitioning is used to map rows to partitions based on ranges
of one or more columns in the table being partitioned. Also, the rows to be partitioned should be
fairly evenly distributed among each partition, such as by months of the year or quarter.
Pages:
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835