We have several customers that are moving their data from SQL Server (a lot)
and Oracle (a few) to DB2 systems. Some of these databases are in the multi-millions
of rows in size. The issue: SQL Server and Oracle allow the databases to
be installed as case in-sensitive, DB2 does *NOT*. IBM's solution is to
either upper-case all the data (which the customer finds unacceptable, as
this makes reports look like garbage), or change all of our SQL code to use
UCASE (which is unacceptable since it doesn't use indexes, and searches take
hours), or use a special command that creates a copy of each of the index
columns in the database as upper case. The last solution is a potential,
but we take an English like query, and build our SQL on-the-fly with security
and application constraints built into the SQL query. Talking to IBM, I
get two opposite answers to the question of "do we need to change our code
to go look at the "special columns" or will the query optimizer do this automatically?

The question: Will the query optimizer use the "special columns" automatically,
or will we have to re-write code and have a special version for DB2 that
adds the UCASE around each field in the Where clause?

Second part, is there any other known solution to this rather large deficiency
in the DB2 database? IBM agrees that this is an issue and actually has a
requirement for this, but they have told us this is low on their priority
list...

Any help would be appreciated!