That is, as far as I remember, the 11th commandment on the tablets
handed to Moses on Mt. Sinai:
"Thou shalt always use index range scan"
In the beginning, there was a rule base optimizer, which used a ranked list of access methods. The optimizer has always selected the access method with the highest rank. Full table scan was at the bottom of the list, single row scan by rowid and single row scan by PK were at the very top of the list. Hash access methods were not on the list. And God saw that RBO was good simple. But then the devil has created CBO and made a set of complex rules which change from release to release. However, it is ordained that faithful should still use RBO principles and ignore CBO, statistics and all related devil's handiwork.
On 04/14/2018 11:19 AM, Frits Hoogland wrote:
My question would be: why an index range scan MUST be used? What is the reason for this person to deem it mandatory?