For each table that is referenced by your query, the Query Processor has an internal data structure; an internal representation of the table that is being searched.
Example:
If your query is searching for point values and historic point values, there will be two internal data structures:
-
A data structure representing the CDBPoint table (as this is the database table that contains the point values)
and
- A data structure that represents the CDBHistoric table (the database table that contains historic point data).
We also refer to an internal data structure as ‘the referenced table’.
Each internal data structure is associated with:
- An access path that determines how the rows in the table are to be searched
- Constraints that reduce the number of rows in the table that are to be searched for a Scan or Non-Unique Lookup access path. There can be 0 constraints, a single constraint, or multiple constraints
- A filter condition (which may be empty) that determines whether a row should be returned in the result set
- A Cost that is calculated based on the table type, access path, and constraints.
The attributes listed above are used by the Query Processor when determining an efficient way to execute your query.