First page Back Continue Last page Graphics

Clustering and Index Access

Notes:

We can test the optimizers reaction to the clustering (or lack of) data. Clustering factor is a measure of the sorted relationship between an index and it’s data. i.e. Is the data in the same sorted sequence as the index? Clustering factor has a big impact on index range scans. A poor clustering factor raises the cost of the index range scan and decreases the probability that CBO will choose that access path.
In this demo, I have an index on the OBJECT_ID column of the BIG table, but Oracle is not using it, even though just 10% of the data is requested.