Say I have the classic:
select * from employees where dob < to_date('10/10/1985', 'DD/MM/YYYY');
The table is huge, so I want to parallelise this query.
Going by what I can see here:
http://docs.oracle.com/cd/B10500_01/server.920/a96524/c20paral.htm#13255
Essentially what we're wanting to do is arbitarily chop the table into n parts, and run our select statement on each chunk on a different thread, then join them together at the end.
- Is parallelisation appropriate here?
- How would I write the query?