Oracle: coding problems may occur when importing data

Question

In Oracle 11, I dumped my data using exp / imp to transfer to another database.

I tried importing the dump file into my local database without any problems.

But then my colleague tried the same thing on his machine, and some tables could not be imported due to an error:

can bind a LONG value only for insert into a LONG column. 

I don't have a long type, but I read that this error can also be selected when the size exceeds the varchar2 type. So I checked the database character set, I have standard Windows encoding and it has utf8 charset. What do you think, maybe the same data length is represented with a large number of bytes, and this leads to such an error?

Do I need to change the encoding of the database and create another dump? I am looking for the best solution, because it also needs to be imported into the customer database, which, as I know, has a completely different encoding.

Answer 1

Any set of legacy windows, by definition, is not multibyte. When you create a multibyte (utf8) db, each individual character can be converted to 1-3 bytes during import. Thus, you should automatically increase each column of the row type up to x3 times before importing. If you accept the limit of 4096, use the Clob type instead.