SYS.PLAN

The PLAN table contains the execution plan that the TimesTen query optimizer prepares after an application calls ttOptSetFlag. See "Generating a query plan" and "Modifying plan generation" in Oracle TimesTen In-Memory Database Operations Guide.

The execution plan includes the operation performed at each step and the table or index that it references.

Columns

Column name	Type	Description
STEP	TT_INTEGER NOT NULL	Ordinal number of the operation, starting at 1.
LEVEL	TT_INTEGER NOT NULL	Level of this operation in the plan tree.
OPERATION	TT_CHAR (31) NOT NULL	Type of operation, one of: `TblLkSerialScan` -- full table scan `RowLkSerialScan` -- full table scan `TblLkTtreeScan` -- ttree scan `RowLkTtreeScan` -- ttree scan `TblLkHashScan` -- hash lookup `RowLkHashScan` -- hash lookup `TblLkRowidScan` -- rowid lookup `RowLkRowidScan` -- rowid lookup `TblLkUpdate` -- updates one or more rows `RowLkUpdate` -- updates one or more rows `TblLkDelete` -- deletes one or more rows `RowLkDelete` -- deletes one or more rows `TblLkInsert` -- inserts one or more rows `RowLkInsert` -- inserts one or more rows `TmpTtreeScanTmpHashScan` -- create a temporary index `NestedLoop [OuterJoin \| SemiJoin]` -- nested loop join (with optional outer join or semi-join) `MergeJoin` -- merge join `OrderBy` -- sorts rows (requires extra temp space) `SortedDistinct` -- identifies distinct rows from a sorted list (requires minimal extra space) `Distinct` -- identifies distinct rows from an unsorted list (requires extra temporary space) `SortedGroupBy` -- identifies distinct groups from a sorted list (requires minimal extra space) `GroupBy` -- identifies distinct groups from an unsorted list (requires extra temp space) `TmpTable` -- materializes intermediate results (requires extra temporary space) `TblLkUpdView` -- updates a view based on changes to detail table(s) `RowLkUpdView` -- updates a view based on changes to detail table(s) `OracleInsert` -- flushes changes to Oracle `ZeroTblScan` -- evaluates a predicate on a single set of values (no scan required) `ViewUniqueMatchScan` -- uniquely identifies those view rows that need to be updated (requires extra temp space
TBLNAME	TT_CHAR (31)	Name of table scanned at this step. Column is NULL if no table is scanned.
IXNAME	TT_CHAR (31)	Name of index used at this step. T-tree index names may have a "(D)" after the name, which indicates a descending scan. Column is NULL if no index is scanned.
PRED	TT_VARCHAR (1024)	Predicate applied during table or index scan or join. Column is NULL if no predicate applies.
OTHERPRED	TT_VARCHAR (1024)	Predicate applied after table or index scan or join. Column is NULL if no predicate applies.