An execution process generated by the optimizer for an SQL statement and is generally expressed as a tree of operators.
When an execution plan only involves a single table or a single partition of a partitioned table, and the table or partition is in the current node, the plan is defined as a "local plan".
When an execution plan only involves a single table or a single partition of a partitioned table, and the table or partition is in a node other than the current one, the plan is defined as a "remote plan".
When an execution plan involves multiple tables or partitions, the plan is defined as a distributed plan.
The method by which a table is accessed, including primary table access and secondary index access. For more information, see Access path.
The order by which tables are joined. Currently, OceanBase Database supports only left-deep join trees. For more information, see Join order.
The algorithm for joining two tables, including Nested Loop Join, Merge Join, and Hash Join.
This is a process that changes your query to an equivalent alternative with a better execution plan.
Execution plan binding
This is a process that allows you to specify the execution plan of an SQL statement by using a given outline. For more information, see Plan binding.
SQL Plan Management (SPM)
A plan evolution mechanism. With the evolution mechanism, the optimizer verifies if any new plan will cause a performance regression. In the case of a performance regression, the plan is rejected.
Adaptive Cursor Sharing (ACS)
A mechanism that allows the optimizer to store multiple plans for each parameterized SQL and select the appropriate plan based on the selection probability space of predicates in an SQL statement.
Data Flow Object (DFO)
Taking data redistribution points as the boundary, a distributed plan is divided into multiple logical subplans that can be executed concurrently. Each subplan is encapsulated in a DFO.