I have an application that reads in a flat file and fetches/creates and
updates the DB based on the information it finds. I am seeing strange
updates statements during commit that are basically overwriting information
with nulls.
Basically, the scenario that I am seeing is that one of the objects can have
2 identifiers in addition to its key. This application only cares about
ID1. ID2 is set by another process so if it is there, great; if not then
don't worry about it.
However, what I am seeing is that during commit, an update statement is
being generated for most of the objects that sets ID1 to its value and ID2
to null, effectively overwriting the data that is present. I am a little
confused on how this could happen since I am not setting ID2 anywhere in the
application.
Any thoughts?
This archive was generated by hypermail 2.0.0 : Thu Jan 22 2004 - 14:41:03 EST