How does DB2 multilingual support work? Is it installation specific or a database specific setting? Is there a way to check if a current installation supports multilingual data?

TIA

does this not mean that data is save in different codepages ?
sysibm.syscolumns keeps codepage for each character field and if clients with different codepage come in and create their tables you will have data with different codepage origine. Or if having an utf database and using graphic fields, you could save data for different languages. will have a lookup on this.

according the infocenter, this means that db2 has been installed with multi languages (english and other..) changing the default system language will return the messages in that language if previously installed. It is an installation issue and copying the language repositories that have been specified instead of default english.

There are 2 main kinds of multi-lingual support:

1. Multiple languages with a single-byte character set (English, Italian, German, French, Brazilian Portuguese, Spanish, etc).

2. Supporting languages that require double-byte character sets such as Japanese, Korean, Simplified Chinese, Traditional Chinese (This version also supports single byte character sets).

DB2 Install Code
Some versions (like Linux and Windows) are language independent and have only one install (or fixpack) install code. AIX has a separate instal code (and separate fixpacks) depending on whether you want option 1 or 2 above. You can check the fixpack page for more details about your specific platform.
http://www-306.ibm.com/software/data...ownloadv8.html

Enabling Double Byte Character Support at the Database Level
Double byte character support is usually enabled at the database level by specifying UTF-8 Codeset when creating the database:
CODESET UTF-8 COLLATE USING IDENTITY_16BIT
The territory code only affects the default of built-in client on the server and does not affect the ability to support multiple languages in the database server.

Column Considerations for Double-Byte Support
When storing "double-byte" characters in a UTF-8 database, you should allow for column lengths that are about 3 times the normal length (this is only necessary if you actually store Japanese, Korean, Simplified Chinese, Traditional Chinese characters in a column). Unfortunately, the amount of space needed varies by character, and cannot be exactly determined in advance. For this reason, you should normally use VARCHAR for columns that can have both single and double-byte characters in them (such as English and Japanese).

Enabling Double-Byte Support at the Table Level
You can enable double-byte character sets at the table level (on a non UTF-8 database) by specifying UNICODE when creating the table, but there are some extreme limitations that usually make this impractical (see the UNICODE parameter in CREATE TABLE in the SQL Reference Vol. 2).