We live in a world where queries are across large amounts of data. We've
heard a variety of ways to work with the ORA-1555 error, but none that are
all that satisfactory. In our case, we tend to have a related "commit across
fetch" issue where we are fetching from a long-running cursor and then
updating based on the data. Because of the volume of data, we are also
forced to commit, and we tend to sometimes force our own "snapshot too old"
error. This is all done with COBOL embedded SQL. We wonder if there isn't
some standard approach that people have determined to use over the years.
For example, we have considered fetching the data into a temporary flat file
and then reading back the table to accomplish the updates. That way we avoid
having the long running fetch/update situation. Is there a better way?
-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
Received on Tue Nov 24 1998 - 19:07:24 CST
