We have fragemnted a table by expression.
The column used for fragmenting is also a part of the index on the table.
When an sql query is fired which uses the index the cost of query comes out to br 1.
When a sql query is fired which does not uses the index and uses the column by which the table is fragmented then the cost of query increases.
My questions is when a fragment from which the data is to be retrieved is identified, can then the index work on this particular fragment. Can we somehow build the index on the fragment level so that retrieval is quick once the fragment is identified.
Or there is no need to fragment when the query uses the index. we actually had more than 23 lacs rows in the table due to which the i/o operation was becoming very slow.
We are trying to fragment this table so that we can achieve better performance.