The following bulleted points identify some of the conditions you might use to determine how segments should be segregated among tablespaces.
They are not prioritized here because the priority depends on your particular environment.
Using Automatic Storage Management (ASM) eliminates many of the contention issues listed with no additional effort by the DBA.
Big segments and small segments should be in separate tablespaces.
Table segments and their corresponding index segments should be in separate tablespaces.
A separate tablespace should be used for each application.
Segments with low usage and segments with high usage should be in different tablespaces.
Static segments should be separated from high DML segments.
Read-only tables should be in their own tablespace.
Staging tables for a data warehouse should be in their own tablespace.
Tablespaces should be created with the appropriate block size, depending on whether segments are accessed row by row or in full table scans.
Materialized views should be in a separate tablespace from the base table.
For partitioned tables and indexes, each partition should be in its own tablespace.
来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/7336830/viewspace-692778/，如需转载，请注明出处，否则将追究法律责任。