direct reads
My application team is running ad-hoc queries against a large (500+ mil rows) table. These queries are resoved via a full table scan. When I monitor the database (10.2.0.3) via OEM I see the number of IO per second jumps from about 1500 to 100000 or 200000. Because these SQLs are ad-hoc, I don't think I will be able to create a new index to satisfy them. Also an additional index will slow down our load process. When I drill down in OEM I see that a lot of waits are direct reads. The table has DOP of 4. I am wondering whether increasing the
0