Implementing Siebel eBusiness Applications on DB2 UDB for z/OS and OS/390 > Preparing for Implementation on the DB2 Host > Setting Up the Subsystem >

Choosing the Database CCSID

Siebel Systems supports both ASCII- and EBCDIC-based coded character set IDs (CCSIDs) on DB2 UDB for z/OS and OS/390. The sort order used in list applets, among others, depends on the database CCSID.

Review the Release Notes on Siebel SupportWeb for information about known restrictions; for example, the following features are not supported on databases with EBCDIC code pages:

• Web Client migration
• Siebel Data Warehouse
• Siebel Dun and Bradstreet server components

You should run development databases with ASCII code pages, because databases with EBCDIC code pages do not support the following procedures in a development environment upgrade:

• Merging prior configuration changes into a new custom configuration repository (this applies to upgrades only)
• Compiling a new Siebel repository file (.srf file) from the new repository

You can partition table spaces on a database with either an ASCII or an EBCDIC code page, so that limit keys reflect the sort sequence difference between ASCII and EBCDIC. For more information, see Configuring the Database Layout.

You can deploy either an ASCII- or an EBCDIC-based CCSID to create your Siebel Schema, but it is generally recommended that you use ASCII.

• Use an ASCII-based encoding scheme to reduce the overhead required for character conversion. Character conversion is performed by DRDA, between the database and the midtier ASCII servers.
• Use an EBCDIC-based encoding scheme if you want to join Siebel tables with other EBCDIC DB2 tables.

CAUTION:  DB2 version 7 only supports a single code page for tables referenced within a single SQL statement.

Setting the CCSID

The CCSIDs must be configured in the fields ASCII CODED CHAR SET or EBCDIC CODED CHAR SET on the DB2 installation panel DSNTIPF. The field DEF ENCODING SCHEME on the same panel determines whether the default encoding scheme is ASCII or EBCDIC.

NOTE:  The field MIXED DATA on DSNTIPF should equal NO for all supported languages except Japanese.