ODPS JDBC

Chinese Docs

MaxCompute JDBC介绍

Installation

Generally, there are two ways to use ODPS JDBC driver in your project.

1.The first one is to use the standalone library:

• Download the with-dependencies-jar from release page.
• Checkout the CHANGELOG.

2.The second is to rely on maven to resolve the dependencies for you:

<dependency>
  <groupId>com.aliyun.odps</groupId>
  <artifactId>odps-jdbc</artifactId>
  <version>VERSION</version>
</dependency>

Getting Started

Using ODPS JDBC driver is just as using other JDBC drivers. It contains the following few steps:

1. Explictly load the ODPS JDBC driver using Class.forName():

Class.forName("com.aliyun.odps.jdbc.OdpsDriver");

2. Connect to the ODPS by creating a Connection object with the JDBC driver:

Connection conn = DriverManager.getConnection(url, accessId, accessKey);

The ODPS server works with RESTful API, so the url looks like:

String url = "jdbc:odps:ENDPOINT?project=PROJECT_NAME&charset=UTF-8";

The connection properties can also be passed through Properties. For example:

Properties config = new Properties();
config.put("access_id", "...");
config.put("access_key", "...");
config.put("project_name", "...");
config.put("charset", "...");
Connection conn = DriverManager.getConnection("jdbc:odps:<endpoint>", config);

3. Submit SQL to ODPS by creating Statement object and using its executeQuery() method:

Statement stmt = conn.createStatement();
ResultSet rs = stmt.executeQuery("SELECT foo FROM bar");

4. Process the result set.

For example:

while (rs.next()) {
    ...
}

Connection String Parameters

It is recommended that the key and value in URL should be encoded by using java.net.URLEncoder#encode(java.lang.String).

Basic

URL key	Property Key	Required	Default value	Description
endpoint	end_point	True		The endpoint of your MaxCompute service
project	project_name	True		The name of your MaxCompute project
accessId	access_id	True		Your Alibaba Cloud access key ID
accessKey	access_key	True		Your Alibaba Cloud access key secret
interactiveMode	interactive_mode	False	false	For MCQA, enable MCQA
logview	logview_host	False	Provided by MC	The endpoint of MaxCompute Logview
tunnelEndpoint	tunnel_endpoint	False	Provided by MC	The endpoint of the MaxCompute Tunnel service
enableOdpsLogger	enable_odps_logger	False	false	Enable MaxCompute JDBC logger

Advanced

URL key	Property Key	Required	Default value	Description
stsToken	sts_token	False		The Alibaba Cloud STS token
logConfFile	log_conf_file	False		The configuration path for SLF4J
charset	charset	False	UTF-8	The charset of the inputs and outputs
executeProject	execute_project_name	False		For MCQA, the name of the MaxCompute project in which actually execute the queries
alwaysFallback	always_fallback	False	false	For MCQA, fall back to regular mode if any exception happened
instanceTunnelMaxRecord	instance_tunnel_max_record	False	-1 (unlimited)	For MCQA, max number of records within a result set, enableLimit option should set to false
instanceTunnelMaxSize	instance_tunnel_max_size	False	-1 (unlimited)	For MCQA, max size of a result set in byte
enableLimit	enable_limit	False	true(limited)	For MCQA, download permission won't be checked if enableLimit is set true, but your result record count will be limited to 10000
autoLimitFallback	auto_limit_fallback	False	False(no auto fallback)	For non-MCQA mode, result record count will be limited to 10000 when no download permission exception happened and autoLimitFallback is set to true

Example

JDBC Client Sample Code

import java.sql.SQLException;
import java.sql.Connection;
import java.sql.ResultSet;
import java.sql.Statement;
import java.sql.DriverManager;

public class OdpsJdbcClient {
  private static String driverName = "com.aliyun.odps.jdbc.OdpsDriver";

  /**
   * @param args
   * @throws SQLException
   */
  public static void main(String[] args) throws SQLException {
    try {
      Class.forName(driverName);
    } catch (ClassNotFoundException e) {
      e.printStackTrace();
      System.exit(1);
    }

    // fill in the information here
    String accessId = "your_access_id";
    String accessKey = "your_access_key";
    Connection conn = DriverManager.getConnection("jdbc:odps:https://service.odps.aliyun.com/api?project=<your_project_name>", accessId, accessKey);
    Statement stmt = conn.createStatement();
    String tableName = "testOdpsDriverTable";
    stmt.execute("drop table if exists " + tableName);
    stmt.execute("create table " + tableName + " (key int, value string)");

    String sql;
    ResultSet rs;

    // insert a record
    sql = String.format("insert into table %s select 24 key, 'hours' value from (select count(1) from %s) a", tableName, tableName);
    System.out.println("Running: " + sql);
    int count = stmt.executeUpdate(sql);
    System.out.println("updated records: " + count);

    // select * query
    sql = "select * from " + tableName;
    System.out.println("Running: " + sql);
    rs = stmt.executeQuery(sql);
    while (rs.next()) {
      System.out.println(String.valueOf(rs.getInt(1)) + "\t" + rs.getString(2));
    }

    // regular query
    sql = "select count(1) from " + tableName;
    System.out.println("Running: " + sql);
    rs = stmt.executeQuery(sql);
    while (rs.next()) {
      System.out.println(rs.getString(1));
    }

    // do not forget to close
    stmt.close();
    conn.close();
  }
}

Running the JDBC Sample Code

# compile the client code
mvn clean package -DskipTests

# run the program with specifying the class path
# using prepared shell script (linux)
./jdbc_test.sh 'jdbc:odps:http://service.odps.aliyun.com/api?project=odpsdemo&accessId=...&accessKey=...&charset=UTF-8&logconffile=logback/logback.xml' 'select * from dual'

# using java command
java -cp "target/odps-jdbc-2.2-jar-with-dependencies.jar:logback/logback-core-1.2.3.jar:logback/logback-classic-1.2.3.jar" com.aliyun.odps.jdbc.JdbcTest "jdbc:odps:http://service.odps.aliyun.com/api?project=odpsdemo&accessId=...&accessKey=...&charset=UTF-8&logconffile=logback/logback.xml" "select * from dual"

Setting SQL task properties

stmt.execute("set biz_id=xxxxxx");
stmt.execute("set odps.sql.mapper.split.size=512");

Third-party Integration

It is also recommended to use ODPS by using other third-party BI tools or DB visualizer that supports JDBC.

For example:

• SQLWorkbench/J
• Squrriel SQL
• Pentaho

Getting Involved

The project is under construction (and not fully JDBC-compliant). If you dicover any good features which have not been implemented, please fire me an Email or just pull a request.

Architecture

Build and run unitest

1.Build from source locally:

git clone ....
cd odps-jdbc
mvn package -DskipTests

2.Copy out a configuration file:

cp ./src/test/resources/conf.properties.example ./src/test/resources/conf.properties

3.Fill in your connection strings:

access_id=...
access_key=...
end_point=...
project_name=...
logview_host=...
charset=UTF-8

4.Run maven test command (or just test it in IntelliJ IDEA):

mvn test

Data Type Mapping

Currently, 13 ODPS data types are supported. Please see the following table for supported ODPS data types and corresponding JDBC interfaces.

ODPS Type	JDBC Interface	JDBC Type
TINYINT	java.sql.ResultSet.getByte	TINYINT
SMALLINT	java.sql.ResultSet.getShort	SMALLINT
INT	java.sql.ResultSet.getInt	INTEGER
BIGINT	java.sql.ResultSet.getLong	BIGINT
FLOAT	java.sql.ResultSet.getFloat	FLOAT
DOUBLE	java.sql.ResultSet.getDouble	DOUBLE
BOOLEAN	java.sql.ResultSet.getBoolean	BOOLEAN
DATETIME	java.sql.ResultSet.getTimestamp	TIMESTAMP
TIMESTAMP	java.sql.ResultSet.getTimestamp	TIMESTAMP
VARCHAR	java.sql.ResultSet.getString	VARCHAR
STRING	java.sql.ResultSet.getString	VARCHAR
DECIMAL	java.sql.ResultSet.getBigDecimal	DECIMAL
BINARY	java.sql.ResultSet.getBytes	BINARY

NOTE: Possible timezone issue

DATETIME in MaxCompute is actually defined as EPOCH in milliseconds, which is UTC, and so is TIMESTAMP in JDBC. This driver fill the DATETIME value directly into JDBC TIMESTAMP and do no parse or format action. When application that using JDBC display a DATETIME as a human-readable string format, it is the application itself did the format using application defined or OS defined timezone. It is suggested to keep your application/OS timezone setting same to MaxCompute to avoid inconsistent datetime parse/format.

Type Conversion

Implicit type conversion happens when accessing a ODPS data type with JDBC interfaces other than the recommended one. Please see the following table for supported implicit conversions.

JAVA\ODPS	TINYINT	SMALLINT	INT	BIGINT	FLOAT	DOUBLE	DECIMAL	VARCHAR	STRING	DATETIME	TIMESTAMP	BOOLEAN	BINARY
byte	Y	Y	Y	Y	Y	Y	Y
short	Y	Y	Y	Y	Y	Y	Y		Y
int	Y	Y	Y	Y	Y	Y	Y		Y
long	Y	Y	Y	Y	Y	Y	Y		Y
float	Y	Y	Y	Y	Y	Y	Y		Y
double	Y	Y	Y	Y	Y	Y	Y		Y
BigDecimal							Y
String	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y
byte[]	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y	Y
Date									Y	Y	Y
Time									Y	Y	Y
Timestamp									Y	Y	Y
boolean	Y	Y	Y	Y	Y	Y	Y		Y			Y

MaxCompute Service Compatibility and Recommended JDBC version

Since Sprint27, MaxCompute tunnel service supported a feature named instance tunnel that allowing client read query result set through tunnel endpoint, to release client from creating temporary table. And this JDBC driver began adopt using instance tunnel since version 2.0.

However, for users using MaxCompute deploy that is earlier than Sprint27 (especially Private Cloud cases), please stick to the latest version before 2.0.

MaxCompute	JDBC
Public Service	latest
Non PRC Public Service	latest
<= Sprint27	1.9.2

Authors && Contributors

• Cheng Yichao
• Li Ruibo
• Zhao Zhenyi
• Wen Shaojin
• Wang Zhong

License

licensed under the Apache License 2.0