If client system connections are to use the defined character sets, the Teradata Database system administrator specifies which character set is assigned to which client system (see HostsInfoV). Otherwise, the standard default is used. Also, the user may specify a defined character set after a session is started (see CharSetsV).
When specifying a character set for a session, the choice should be compatible with the internal code of the logon client system; that is, an EBCDIC-compatible character set for sessions initiated from an IBM mainframe, ASCII-compatible sets for all others. It is suggested, therefore, that a convention be used for naming character sets which differentiates between EBCDIC and ASCII compatibility (see the example below).
If the character set is user-defined, the ID should also exist in DBC.CharTranslations.CharSetID.
For constraints not involving comparison of character data or not using CHARSET_COLL for evaluation, the value is NULL.
The Teradata Database must be reset to install the rows containing a Y in the InstallFlag field. If the value of InstallFlag is Y in 12 rows or fewer, each Y row is loaded. If InstallFlag is Y in more than 12 rows, then the CharSetName values are sorted in ascending ASCII sequence, and rows are loaded in alphabetical order until 12 sets are installed or the names are exhausted.