We have a new vendor application that requires a specific character set (Western European), which we do not have installed in our existing Oracle instances (US7ASCII). Since the new application will interact with other internal applications, we want to install them on the same instance.
Is it possible within an Oracle instance to have multiple NLS character sets defined by tablespace or table? Is this a case for converting an instance to UNICODE? Also, are there any conversion issues if we try to join tables across these character sets, be it either in the same instance or through a call across instances?
Oracle can do some conversion between character sets. However it might be better to use UNICODE so that the expected character sets can be handled with little or no conversion. US7ASCII is a subset of the Western European set of characters so there may not be many problems. However, I would consider a rebuild to the Western European characterset and using a National set of US7ASCII.
Dig Deeper on Oracle database design and architecture
Related Q&A from Mike Ault
How to find the definition or structure of a dropped table? I know the table's name but I don't know the columns and datatypes. It no longer exists. Continue Reading
I am trying to remove carriage returns at the end of clob fields in SQL*Plus. This just nulls out the field in the table. What do I need to change to... Continue Reading
I want to find the length of a numeric datatype field in my table. How can I find it? Continue Reading
Have a question for an expert?
Please add a title for your question
Get answers from a TechTarget expert on whatever's puzzling you.