Hi All,

We seem to found a solution to this problem. And again it seems to be a
bug/defect with IDS 9.30.HC2.

We had an index on the timestamp field which we were updating in the
FOREACH loop. Now multiple instances of this program were
simultaneously running to process different set of records. Since each
instance of the program was updating the timestamp field for its set,
the index on timestamp field was also being updated. It seems that with
this update of the indexes through different threads, the records
somehow were getting locked and were being skipped when SELECT
statement was executed (although it is to be noted that the SELECT
statements were trying to select some other set of records and not the
ones being updated).

Can somebody throw some light on this behaviour ?

Regards,
Ravi.