I am working on a map that reads from over 30 different DB2 database tables that have large volume of records. An existing map (made by someone else) uses separate DBLOOKUP calls to read from each table. These are done in a series of functional maps. The map has had performance issues and is large. Would it make more sense to have the database spit out a text file where the fields are pipe delimited and the rows are CR/LF delimited and have that file trigger a map? Or should the database have a view that contains all records for a claim? How have you solved this problem in your project?