如何通过jdbc将数据插入Hive表中？

Question

I've tried to use jdbc to operate data in hive. When I used select or create operation, everything worked fine. But when I tried to use insert to add some rows into a table like:

String sql = "insert into agg_test_20170508 values (20170508, 100)";
Class.forName(JDBC_Driver);
connection = DriverManager.getConnection(JDBC_String, ID, pwd);
Statement stmt = connection.createStatement();
System.out.println(D_FORMAT.format(new Date()) + " Running:\r\n" + sql);
ResultSet res = stmt.executeQuery(sql);
System.out.println(D_FORMAT.format(new Date()) + " QUERY COMPELET.");

But after a long time, nothing returned. Are there any restriction when using jdbc to connect to hive?

The hadoop version in my company is 0.20.2.

Answer 1

You have to use executeUpdate with INSERT UPDATE DELETE :

int res = stmt.executeUpdate(sql);

int executeUpdate(String sql): executes an INSERT, UPDATE or DELETE statement and returns an update account indicating number of rows affected (eg 1 row inserted, or 2 rows updated, or 0 rows affected).

ResultSet executeQuery(String sql): executes a SELECT statement and returns a ResultSet object which contains results returned by the query.

Answer 2

请使用executeUpdate方法而不是executeQuery方法

Answer 3

executeQuery() --- This is used generally for reading the content of the database. The output will be in the form of ResultSet. Generally SELECT statement is used.

executeUpdate() --- This is generally used for altering the databases. Generally DROP TABLE or DATABASE, INSERT into TABLE, UPDATE TABLE, DELETE from TABLE statements will be used in this. The output will be in the form of int. This int value denotes the number of rows affected by the query.

execute() --- If you dont know which method to be used for executing SQL statements, this method can be used. This will return a boolean. TRUE indicates the result is a ResultSet and FALSE indicates it has the int value which denotes number of rows affected by the query.