MERGE
Performs update and insert operations on a target table based on the results of a join with another data set, such as a table or view. The join can match a source row with only one target row; otherwise, Vertica returns an error.
If a merge would violate a table or schema disk quota, the operation fails. For more information, see Disk quotas.
The target table cannot have columns of complex data types. The source table can, so long as those columns are not included in the merge operation.
Syntax
MERGE [ /*+LABEL (label-string)*/ ]
INTO [[database.]schema.]target-table [ [AS] alias ]
USING source-dataset
ON join-condition matching-clause[ matching-clause ]
Returns
Number of target table rows updated or inserted
Arguments
-
LABEL
Assigns a label to a statement to identify it for profiling and debugging.
[
database
.]
schema
Database and schema. The default schema is
public
. If you specify a database, it must be the current database.target-table
- The table on which to perform update and insert operations. MERGE takes an X (exclusive) lock on the target table during the operation. The table must not contain columns of complex types.
Important
The total number of target table columns cannot exceed 831. source-dataset
- The data to join to
target-table
, one of the following:-
[[
database
.]
schema.
]
table
[ [AS]
alias
]
-
[[
database
.]
schema.
]
view
[ [AS]
alias
]
-
(
subquery
)
sq-alias
The specified data set typically supplies the data used to update the target table and populate new rows. You can specify an external table.
-
ON
join-condition
- The conditions on which to join the target table and source data set.
Tip
The Vertica query optimizer can create an optimized query plan for a MERGE statement only if the target table join column has a unique or primary key constraint. For details, see MERGE optimization. matching-clause
- One of the following clauses:
MERGE supports one instance of each clause, and must include at least one.
-
WHEN MATCHED THEN UPDATE
- For each
target-table
row that is joined (matched) tosource-dataset
, specifies to update one or more columns:WHEN MATCHED [ AND update-filter ] THEN UPDATE SET { column = expression }[,...]
update-filter
optionally filters the set of matching rows. The update filter can specify any number of conditions. Vertica evaluates each matching row against this filter, and updates only the rows that evaluate to true. For details, see Update and insert filters.Note
Vertica also supports Oracle syntax for specifying update filters:
WHEN MATCHED THEN UPDATE SET { column = expression }[,...] [ WHERE update-filter ]
The following requirements apply:
-
A MERGE statement can contain only one WHEN MATCHED clause.
-
target-column
can only specify a column name in the target table. It cannot be qualified with a table name.
For details, see Merging table data.
-
-
WHEN NOT MATCHED THEN INSERT
- For each
source-dataset
row that is not joined (not matched) totarget-table
, specifies to:-
Insert a new row into
target-table
. -
Populate each new row with the values specified in
values-list
.
WHEN NOT MATCHED [ AND insert-filter ] THEN INSERT [ ( column-list ) ] VALUES ( values-list )
column-list
is a comma-delimited list of one or more target columns in the target table, listed in any order. MERGE mapscolumn-list
columns tovalues-list
values in the same order, and each column-value pair must be compatible. If you omitcolumn-list
, Vertica mapsvalues-list
values to columns according to column order in the table definition.insert-filter
optionally filters the set of non-matching rows. The insert filter can specify any number of conditions. Vertica evaluates each non-matching source row against this filter. For each row that evaluates to true, Vertica inserts a new row in the target table. For details, see Update and insert filters.Note
Vertica also supports Oracle syntax for specifying insert filters:
WHEN NOT MATCHED THEN INSERT [ ( column-list ) ] VALUES ( values-list ) [ WHERE insert-filter ]
The following requirements apply:
-
A MERGE statement can contain only one WHEN NOT MATCHED clause.
-
*
column-list*
can only specify column names in the target table. It cannot be qualified with a table name. -
Insert filter conditions can only reference the source data. If any condition references the target table, Vertica returns an error.
For details, see Merging table data.
-
Privileges
MERGE requires the following privileges:
-
SELECT permissions on the source data and INSERT, UPDATE, and DELETE permissions on the target table.
-
Automatic constraint enforcement requires SELECT permissions on the table containing the constraint.
-
SELECT permissions on the target table if the condition in the syntax reads data from the target table.
For example, the following GRANT statement grants user1
access to the t2
table. This allows user1
to run the MERGE statement that follows:
=> GRANT SELECT, INSERT, UPDATE, DELETE ON TABLE t2 to user1;
GRANT PRIVILEGE
=>\c - user1
You are now connected as user "user1".
=> MERGE INTO t2 USING t1 ON t1.a = t2.a
WHEN MATCHED THEN UPDATE SET b = t1.b
WHEN NOT MATCHED THEN INSERT (a, b) VALUES (t1.a, t1.b);
Improving MERGE performance
You can improve MERGE performance in several ways:
-
Use a source data set that is smaller than the target table.
For details, see MERGE optimization.
Constraint enforcement
If primary key, unique key, or check constraints are enabled for automatic enforcement in the target table, Vertica enforces those constraints when you load new data. If a violation occurs, Vertica rolls back the operation and returns an error.
Caution
If you run MERGE multiple times using the same target and source table, each iteration is liable to introduce duplicate values into the target columns and return with an error.Columns prohibited from merge
The following columns cannot be specified in a merge operation; attempts to do so return with an error:
-
IDENTITY columns, or columns whose default value is set to a named sequence.
-
Vmap columns such as
__raw__
in flex tables. -
Columns of complex types ARRAY, SET, or ROW.
Examples
See: