We have an Oracle 12c (126.96.36.199) database server configured with the AL32UTF8 character set (UTF-8 Universal Unicode 5.0 character set) in accordance with Oracle recommendations.
However, the application that uses this database is accompanied by a recommendation from the provider to instead use the character set WE8MSWIN1252 (code code code MS Windows Page 1252 in Western Europe 8-bit) .
I know, at least in general, how to convert one character set to another, using the Data Migration Wizard for Unicode or through a full export / import.
But there are two areas on which I am not clear.
The first is that Oracle says that there is a potential for data truncation problems (and explains how / why). Is there a way to find out if there is a problem in my database, unless you convert the database into a test and try to do a line-by-line comparison between converted and unconverted data?
Most importantly, since the provider wants us to use a different character set than Oracle recommends, is there a way to measure the potential impact on conversion performance? From what I can tell, the server is about to convert the incoming data from the WE8MSWIN1252 character set to AL32UTF8 when inserting / updating and vice versa when selecting, no ? Are there any insights that would allow me to see how long the database spends processing these conversions? I really can not imagine that the performance is enough to justify the conversion, but I would like to know for sure.