This works 99% of the time, but under certain conditions when I check the
recordset after a read (select * from ...) it appears empty (i.e.BOF and
adoEOF are both true) and there is not any error generated ( I suspected a
deadlock since that can occur in SQLServer, but there was no error generated
and a read in Oracle does not grab any locks).
Since I never expect this record to be empty the program logs an application
error that includes the SQL statement that was executed in order to create
the recordset. When I execute this SQL manually the record is present.
According to my understanding of Oracle, reading a record and updating the
same record from separate threads should never interfere with each other.
So my question is what is going on or how can I go about figuring out what
is going on. The Oracle log files did not reveal anything. The only unique
thing about this record is that it contains a BLOB and the BLOB is one of
the fields that gets updated.
Thanks in advance for any help,
Addie