to*****@hotmail.com wrote:
Hi,
I am using udb version 8 fixpack 12 installed on Linux.
I need to move a schema from one db to another. All tables in the
target and source schemas are created using 'CCSID UNICODE'.
I used db2move export command from the db2 client on Windows XP.
However, when I import the data into the target schema, I noticed some
special characters were replaced by a right-arrow character.
What do I need to do to keep the data intact during the transfer?
Thanks in advance.
Tom
I believe that by default DB2 converts data to the client application
codepage. This includes the output of export. Try setting the DB2CODEPAGE
environment variable to 1208 before doing the export. It may also have to
be set prior to import or load.
When using UTF-8, even when using characters that convert correctly you can
get import/load errors due to truncation if the client is not running with
the correct codepage. E.g. Some characters may correctly change from
multi-byte to single byte and a char column will get padded with blanks.
When importing/loading the reverse happens but the blanks may not be
removed and an error is the result.
Colin