To set up Oracle for use with single or multiple byte characters:
CHARACTER SET "UTF8"
open_cursors= <set to appropriate value>
For example, the minimum value for WebLogic equals number of threads (across all application servers) + (connection pool size X prepared statement pool size)
cursor_sharing=FORCE
compatible=<set to appropriate value, or remove to default to current release>
timed_statistics=true
db_block_size=8192
optimizer_mode=ALL_ROWS
If you are using multi-byte character set, set the following and restart Oracle:
nls_length_semantics=CHAR
Alternatively you can run the following prior to running any create table scripts:
alter session set nls_length_semantics = CHAR
Setting this attribute ensures that the field sizes are not impacted by the number of bytes a data type can store. For example, Varchar(40) would now be able to store 40 Japanese characters instead of 40/3 bytes in the UTF-8 character set.
For the Japanese locale, the AL32UTF-8 character set or the UTF-16 character set must be used.
When you change the multi-byte character set to CHAR by setting nls_length_semantics = CHAR, Oracle reserves space equivalent to 'n' chars, which is more than 'n' bytes. Therefore, when you run the dbverify.cmd command, the reduced entries in table columns are printed in the EFrame_Drops.lst file.
The Oracle JDBC driver can be found in the JDBC Driver Downloads section at:
http://www.oracle.com/technology/software/tech/java/sqlj_jdbc/index.html
For Oracle database supported version information, go to the Support Portal at http://www.ibm.com/support/entry/portal/Planning
To create the Oracle database to handle multiple byte characters, do not modify the DDL.