If an information object retrieves data from only one data source, the query execution plan consists of a single node: a SQL or ODA node that displays the query sent to the data source. If an information object retrieves data from more than one data source, the query execution plan consists of several nodes, including the following:
One SQL or ODA node for each data source query
Nodes that represent the joins between data source queries
For example, the query execution plan in Figure 3‑55 displays two SQL nodes and a join node. The SQL nodes represent the native SQL queries that are sent to two different databases. The join node indicates that the join uses the Nested Loop join algorithm.
Figure 3‑55 An execution plan for a query using a join of two SQL queries
You can experiment with different join algorithms and determine which one performs best by doing the following:
Displaying the execution time for each node
Calculating the total execution time for the information object by adding the execution times for the individual nodes
One SQL or ODA node for each data source query