I have a data flow two lookups components (call them lookup1 and lookup2). They both query the same relational table but with different values. Each has a single row result set containing one column and the each of the two columns is mapped to a corresponding package-level variable. The original data flow sequence had lookup1 executing after lookup 2. Each component redirects errors to a separate text file.
Lookup1 succeeds but lookup2 fails on every row which populates its error text file; however I can construct a sql query from the lookup2 values that returns the expected result.
If I reverse the sequence of components (lookup2 followed by lookup1) lookup2 still fails on every row. Whenever both lookups are present in the dataflow, lookup2 fails for every row and all its rows are redirected to the error text file
Now this is where it gets interesting. If I omit either lookup1 or lookup 2 from the data flow, it works. If the data flow contains lookup1 only the destination is populated. If the dataflow contains only lookup2 then no errors are written to the error text file, all lookups succeed, and the destination is populated.
I'm stumped. Is it possible that both lookups selecting from the same table could cause a problem Each works independently, but when both are together in the data flow, lookup2 fails for every row. I've been over the configuration and code a dozen times and am positive there are no errors; besides, lookup2 runs fine if lookup1 is excluded from the data flow.