Restrictions - ODBC Driver for Teradata

ODBC Driver for Teradata® User Guide

Product
ODBC Driver for Teradata
Release Number
16.20
Published
August 2020
Language
English (United States)
Last Update
2020-08-25
dita:mapPath
fxv1527114222338.ditamap
dita:ditavalPath
Audience_PDF_product_legacy_odbc_include.ditaval
dita:id
B035-2526
lifecycle
previous
Product Category
Teradata Tools and Utilities
  • ODBC Driver for Teradata does not support EBCDIC session character sets for the Unicode ODBC driver. Note that KANJIEBCDIC5035_0I is supported on Windows. See Session Character Sets and Translation DLLs for more information.
  • Retrieving fixed character fields (for example, C01 CHAR(10)) utilizing the UTF8 session character set can result in padded strings due to the database export factor utilized when translating characters to the session character set. Use UTF16 session character set which contains an export factor of 2 for all characters, or utilize varchar() variables instead of char().
  • Kanji User ID and Password are not supported on an English-enabled client machine. Kanji User ID and passwords can be used on English language PCs if the database release is 12 or higher, the database is Kanji-enabled, and the session character set is UTF16 or UTF8. This is a Unicode Data Dictionary enhancement.
  • Characters not found during translations will be substituted by an error character. For example, given a session on a Japanese Windows machine, the string “abecedný” would convert the “ý” to “0xfcfc”, indicating a conversion error from the ODBC code converter. The character, “ý”, is not a valid SJIS character. See the “Character Conversion in Teradata® Database International Character Set Support (B035-1125).
  • ANSI applications requesting column names or other meta data containing non-ASCII characters through the SQLColAttribute method, using a Unicode session character set, will retrieve the non-ASCII characters in UTF8 encoding (not encoded by the application code page). ANSI applications using non-Unicode sessions will not encounter this problem. It could be a problem if a customer is in the process of migrating applications to Unicode having some ANSI and some Unicode applications, all running over Unicode sessions. A workaround is to use the old SQLColAttributes convention ('s' at the end).
  • Calling SQLColumns against a table with a 2 GB CLOB column can cause an error because the value in the BUFFER_LENGTH column overflows when bound to unsigned integer types. The maximum CLOB size recommended for use with UTF8 (export factor 2) and UTF16 session character sets is 1 GB.