This topic describes the method and limits of executing INSERT INTO statements in Realtime Compute for Apache Flink.
Operation limits
| Table type | Limit |
|---|---|
| Source table | Can be referenced in only FROM clauses and does not support INSERT statements. |
| Dimension table | Can be referenced in only JOIN statements and does not support INSERT statements. |
| Result table | Supports only INSERT statements. |
| View | Can be referenced in only FROM clauses. |
Syntax
INSERT INTO tableName
[ (columnName[ , columnName]*) ]
queryStatement;Examples
INSERT INTO LargeOrders
SELECT * FROM Orders WHERE units > 1000;
INSERT INTO Orders(z,v)
SELECT c,d FROM OO;Note
- In Realtime Compute for Apache Flink, a single SQL job can contain multiple data manipulation language (DML) operations, data sources, data destinations, and dimension tables. For example, a job file can contain two snippets of SQL statements for independent services. You can execute the SQL statements to write data to different data destinations.
- Realtime Compute for Apache Flink does not allow you to execute a separate SELECT statement to query data. To execute a SELECT statement, you must include the SELECT statement in a CREATE VIEW or an INSERT INTO statement.
- You can execute an INSERT INTO statement to update an existing record. For example, you can insert a key value into an ApsaraDB for RDS result table that contains a primary key. If the key value already exists, the existing record is updated. If the key value does not exist, a new key value is inserted.