This page describes the different clients supported by HiveServer2.

Beeline – New Command Line Shell

HiveServer2 supports a new command shell Beeline that works with HiveServer2. It's a JDBC client that is based on the SQLLine CLI (http://sqlline.sourceforge.net/). There’s detailed documentation of SQLLine which is applicable to Beeline as well.

The Beeline shell works in both embedded mode as well as remote mode. In the embedded mode, it runs an embedded Hive (similar to Hive CLI) whereas remote mode is for connecting to a separate HiveServer2 process over Thrift. Starting in Hive 0.14, when Beeline is used with HiveServer2, it also prints the log messages from HiveServer2 for queries it executes to STDERR.

Beeline Example

% bin/beeline
Hive version 0.11.0-SNAPSHOT by Apache
beeline> !connect jdbc:hive2://localhost:10000 scott tiger org.apache.hive.jdbc.HiveDriver
!connect jdbc:hive2://localhost:10000 scott tiger org.apache.hive.jdbc.HiveDriver
Connecting to jdbc:hive2://localhost:10000
Connected to: Hive (version 0.10.0)
Driver: Hive (version 0.10.0-SNAPSHOT)
Transaction isolation: TRANSACTION_REPEATABLE_READ
0: jdbc:hive2://localhost:10000> show tables;
show tables;
+-------------------+
|     tab_name      |
+-------------------+
| primitives        |
| src               |
| src1              |
| src_json          |
| src_sequencefile  |
| src_thrift        |
| srcbucket         |
| srcbucket2        |
| srcpart           |
+-------------------+
9 rows selected (1.079 seconds)

% bin/beeline
beeline> !connect jdbc:hive2://<host>:<port>/<db>;auth=noSasl hiveuser pass org.apache.hive.jdbc.HiveDriver

Beeline Commands

Beeline Hive Commands

Hive specific commands (same as Hive CLI commands) can be run from Beeline, when the Hive JDBC driver is used.

Use ";" (semicolon) to terminate commands. Comments in scripts can be specified using the "--" prefix.

Beeline Command Options

The Beeline CLI supports these command line options:

Separated-Value Output Formats

Starting with Hive 0.14, there are improved SV output formats available, namely DSV, CSV2 and TSV2.  These conform better to standard CSV convention, which adds quotes around a cell value only if it contains special characters (such as the delimiter character or a quote character) or spans multiple lines.  These three formats differ only with the delimiter between cells, which is comma for CSV2, tab for TSV2, and configurable for DSV (delimiterForDSV property).

CSV and TSV output formats are maintained for backward compatibility, but beware as they add additional single-quote characters around all cell values contrary to this convention.

HiveServer2 Logging

Starting with Hive 0.14.0, HiveServer2 operation logs are available for Beeline clients. These parameters configure logging:

• hive.server2.logging.operation.enabled
• hive.server2.logging.operation.log.location
• hive.server2.logging.operation.verbose (Hive 0.14 to 1.1)
• hive.server2.logging.operation.level (Hive 1.2 onward)

JDBC

HiveServer2 has a new JDBC driver. It supports both embedded and remote access to HiveServer2.

Connection URLs

Connection URL Format

The HiveServer2 URL is a string with the following syntax:

jdbc:hive2://<host1>:<port1>,<host2>:<port2>/dbName;sess_var_list?hive_conf_list#hive_var_list

where

• <host1>:<port1>,<host2>:<port2> is a server instance or a comma separated list of server instances to connect to (if dynamic service discovery is enabled). If empty, the embedded server will be used.
• dbName is the name of the initial database.
• sess_var_list is a semicolon separated list of key=value pairs of session variables (e.g., user=foo;password=bar).
• hive_conf_list is a semicolon separated list of key=value pairs of Hive configuration variables for this session (e.g., hive.server2.transport.mode=http;hive.server2.thrift.http.path=hs2).
• hive_var_list is a semicolon separated list of key=value pairs of Hive variables for this session.

Connection URL for Remote or Embedded Mode

The JDBC connection URL format has the prefix jdbc:hive2:// and the Driver class is org.apache.hive.jdbc.HiveDriver. Note that this is different from the old HiveServer.

• For a remote server, the URL format is jdbc:hive2://<host>:<port>/<db> (default port for HiveServer2 is 10000).
• For an embedded server, the URL format is jdbc:hive2:// (no host or port).

Connection URL When HiveServer2 Is Running in HTTP Mode

JDBC connection URL:  jdbc:hive2://<host>:<port>/<db>?hive.server2.transport.mode=http;hive.server2.thrift.http.path=<http_endpoint>, where:

• <http_endpoint> is the corresponding HTTP endpoint configured in hive-site.xml. Default value is cliservice.
• Default port for HTTP transport mode is 10001

Connection URL When SSL Is Enabled in HiveServer2

JDBC connection URL:  jdbc:hive2://<host>:<port>/<db>;ssl=true;sslTrustStore=<trust_store_path>;trustStorePassword=<trust_store_password>, where:

• <trust_store_path> is the path where client's truststore file lives.
• <trust_store_password> is the password to access the truststore.

In HTTP mode:  jdbc:hive2://<host>:<port>/<db>;ssl=true;sslTrustStore=<trust_store_path>;trustStorePassword=<trust_store_password>?hive.server2.transport.mode=http;hive.server2.thrift.http.path=<http_endpoint>.

Using JDBC

You can use JDBC to access data stored in a relational database or other tabular format.

1. Load the HiveServer2 JDBC driver. As of 1.2.0 applications no longer need to explicitly load JDBC drivers using Class.forName().
   
   For example:

   Class.forName("org.apache.hive.jdbc.HiveDriver");
   

2. Connect to the database by creating a Connection object with the JDBC driver.
   
   For example:

   Connection cnct = DriverManager.getConnection("jdbc:hive2://<host>:<port>", "<user>", "<password>");
   

   The default <port> is 10000. In non-secure configurations, specify a <user> for the query to run as. The <password> field value is ignored in non-secure mode.

   Connection cnct = DriverManager.getConnection("jdbc:hive2://<host>:<port>", "<user>", "");
   

   In Kerberos secure mode, the user information is based on the Kerberos credentials.

3. Submit SQL to the database by creating a Statement object and using its executeQuery() method.
   
   For example:

   Statement stmt = cnct.createStatement();
   ResultSet rset = stmt.executeQuery("SELECT foo FROM bar");
   

4. Process the result set, if necessary.

These steps are illustrated in the sample code below.

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 HiveJdbcClient {
  private static String driverName = "org.apache.hive.jdbc.HiveDriver";

  /**
   * @param args
   * @throws SQLException
   */
  public static void main(String[] args) throws SQLException {
      try {
      Class.forName(driverName);
    } catch (ClassNotFoundException e) {
      // TODO Auto-generated catch block
      e.printStackTrace();
      System.exit(1);
    }
    //replace "hive" here with the name of the user the queries should run as
    Connection con = DriverManager.getConnection("jdbc:hive2://localhost:10000/default", "hive", "");
    Statement stmt = con.createStatement();
    String tableName = "testHiveDriverTable";
    stmt.execute("drop table if exists " + tableName);
    stmt.execute("create table " + tableName + " (key int, value string)");
    // show tables
    String sql = "show tables '" + tableName + "'";
    System.out.println("Running: " + sql);
    ResultSet res = stmt.executeQuery(sql);
    if (res.next()) {
      System.out.println(res.getString(1));
    }
       // describe table
    sql = "describe " + tableName;
    System.out.println("Running: " + sql);
    res = stmt.executeQuery(sql);
    while (res.next()) {
      System.out.println(res.getString(1) + "\t" + res.getString(2));
    }

    // load data into table
    // NOTE: filepath has to be local to the hive server
    // NOTE: /tmp/a.txt is a ctrl-A separated file with two fields per line
    String filepath = "/tmp/a.txt";
    sql = "load data local inpath '" + filepath + "' into table " + tableName;
    System.out.println("Running: " + sql);
    stmt.execute(sql);

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

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

Running the JDBC Sample Code

# Then on the command-line
$ javac HiveJdbcClient.java

# To run the program using remote hiveserver in non-kerberos mode, we need the following jars in the classpath
# from hive/build/dist/lib
#     hive-jdbc*.jar
#     hive-service*.jar
#     libfb303-0.9.0.jar
#  	  libthrift-0.9.0.jar
# 	  log4j-1.2.16.jar
# 	  slf4j-api-1.6.1.jar
# 	  slf4j-log4j12-1.6.1.jar
# 	  commons-logging-1.0.4.jar
#
#
# To run the program using kerberos secure mode, we need the following jars in the classpath 
#     hive-exec*.jar
#     commons-configuration-1.6.jar
#  and from hadoop
#     hadoop-core*.jar
#
# To run the program in embedded mode, we need the following additional jars in the classpath
# from hive/build/dist/lib
#     hive-exec*.jar
#     hive-metastore*.jar
#     antlr-runtime-3.0.1.jar
#     derby.jar
#     jdo2-api-2.1.jar
#     jpox-core-1.2.2.jar
#     jpox-rdbms-1.2.2.jar
# and from hadoop/build
#     hadoop-core*.jar
# as well as hive/build/dist/conf, any HIVE_AUX_JARS_PATH set, 
# and hadoop jars necessary to run MR jobs (eg lzo codec)

$ java -cp $CLASSPATH HiveJdbcClient

Alternatively, you can run the following bash script, which will seed the data file and build your classpath before invoking the client. The script adds all the additional jars needed for using HiveServer2 in embedded mode as well.

#!/bin/bash
HADOOP_HOME=/your/path/to/hadoop
HIVE_HOME=/your/path/to/hive

echo -e '1\x01foo' > /tmp/a.txt
echo -e '2\x01bar' >> /tmp/a.txt

HADOOP_CORE=$(ls $HADOOP_HOME/hadoop-core*.jar)
CLASSPATH=.:$HIVE_HOME/conf:$(hadoop classpath)

for i in ${HIVE_HOME}/lib/*.jar ; do
    CLASSPATH=$CLASSPATH:$i
done

java -cp $CLASSPATH HiveJdbcClient

JDBC Data Types

The following table lists the data types implemented for HiveServer2 JDBC.

JDBC Client Setup for a Secure Cluster

When connecting to HiveServer2 with Kerberos authentication, the URL format is:

jdbc:hive2://<host>:<port>/<db>;principal=<Server_Principal_of_HiveServer2>

The client needs to have a valid Kerberos ticket in the ticket cache before connecting.

NOTE: If you don't have a "/" after the port number, the jdbc driver does not parse the hostname and ends up running HS2 in embedded mode . So if you are specifying a hostname, make sure you have a "/" or "/<dbname>" after the port number.

In the case of LDAP, CUSTOM or PAM authentication, the client needs to pass a valid user name and password to the JDBC connection API.

To use sasl.qop, add the following to the sessionconf part of your Hive JDBC hive connection string, e.g.

jdbc:hive://hostname/dbname;sasl.qop=auth-int

For more information, see Setting Up HiveServer2.

Multi-User Scenarios and Programmatic Login to Kerberos KDC

In the current approach of using Kerberos you need to have a valid Kerberos ticket in the ticket cache before connecting. This entails a static login (using kinit, key tab or ticketcache) and the restriction of one Kerberos user per client. These restrictions limit the usage in middleware systems and other multi-user scenarios, and in scenarios where the client wants to login programmatically to Kerberos KDC.

One way to mitigate the problem of multi-user scenarios is with secure proxy users (see HIVE-5155). Starting in Hive 0.13.0, support for secure proxy users has two components:

• Delegation token based connection for Oozie (OOZIE-1457). This is the common mechanism for Hadoop ecosystem components.
• Direct proxy access for privileged Hadoop users (HIVE-5155). This enables a privileged user to directly specify an alternate session user during the connection. If the connecting user has Hadoop level privilege to impersonate the requested userid, then HiveServer2 will run the session as that requested user.

The other way is to use a pre-authenticated Kerberos Subject (see HIVE-6486). In this method, starting with Hive 0.13.0 the Hive JDBC client can use a pre-authenticated subject to authenticate to HiveServer2. This enables a middleware system to run queries as the user running the client.

Using Kerberos with a Pre-Authenticated Subject

To use a pre-authenticated subject you will need the following changes.

1. Add hive-exec*.jar to the classpath in addition to the regular Hive JDBC jars (commons-configuration-1.6.jar and hadoop-core*.jar are not required).
2. Add auth=kerberos and kerberosAuthType=fromSubject JDBC URL properties in addition to having the “principal" url property.
3. Open the connection in Subject.doAs().

The following code snippet illustrates the usage (refer to HIVE-6486 for a complete test case):

static Connection getConnection( Subject signedOnUserSubject ) throws Exception{
       Connection conn = (Connection) Subject.doAs(signedOnUserSubject, new PrivilegedExceptionAction<Object>()
           {
               public Object run()
               {
                       Connection con = null;
                       String JDBC_DB_URL = "jdbc:hive2://HiveHost:10000/default;" ||
                                              "principal=hive/localhost.localdomain@EXAMPLE.COM;" || 
                                              "kerberosAuthType=fromSubject";
                       try {
                               Class.forName(JDBC_DRIVER);
                               con =  DriverManager.getConnection(JDBC_DB_URL);
                       } catch (SQLException e) {
                               e.printStackTrace();
                       } catch (ClassNotFoundException e) {
                               e.printStackTrace();
                       }
                       return con;
               }
           });
       return conn;
}

Python Client

A Python client driver is available on github. For installation instructions, see Setting Up HiveServer2: Python Client Driver.

Ruby Client

A Ruby client driver is available on github at https://github.com/forward3d/rbhive.

Integration with SQuirrel SQL Client 

1. Download, install and start the SQuirrel SQL Client from the SQuirrel SQL website.
2. Select 'Drivers -> New Driver...' to register Hive's JDBC driver that works with HiveServer2.

   1. Enter the driver name and example URL:

         Name: Hive
         Example URL: jdbc:hive2://localhost:10000/default
      

3. Select 'Extra Class Path -> Add' to add the following jars from your local Hive and Hadoop distribution.

      HIVE_HOME/build/dist/lib/*.jar
      HADOOP_HOME/hadoop-*-core.jar 

4. Select 'List Drivers'. This will cause SQuirrel to parse your jars for JDBC drivers and might take a few seconds. From the 'Class Name' input box select the Hive driver for working with HiveServer2:

      org.apache.hive.jdbc.HiveDriver
      

5. Click 'OK' to complete the driver registration. 

6. Select 'Aliases -> Add Alias...' to create a connection alias to your HiveServer2 instance.
   1. Give the connection alias a name in the 'Name' input box.
   2. Select the Hive driver from the 'Driver' drop-down.
   3. Modify the example URL as needed to point to your HiveServer2 instance.
   4. Enter 'User Name' and 'Password' and click 'OK' to save the connection alias.
   5. To connect to HiveServer2, double-click the Hive alias and click 'Connect'.

When the connection is established you will see errors in the log console and might get a warning that the driver is not JDBC 3.0 compatible. These alerts are due to yet-to-be-implemented parts of the JDBC metadata API and can safely be ignored. To test the connection enter SHOW TABLES in the console and click the run icon.

Also note that when a query is running, support for the 'Cancel' button is not yet available.

Advanced Features for Integration with Other Tools

Supporting Cookie Replay in HTTP Mode 

HIVE-9709 introduced support for the JDBC driver to enable cookie replay. This is turned on by default so that incoming cookies can be sent back to the server for authentication purpose.  

The JDBC connection URL when enabled should look like: jdbc:hive2://<host>:<port>/<db>?transportMode=http;httpPath=<http_endpoint>;cookieAuth=true;cookieName=<cookie_name>

• cookieAuth is set to default as true.
• cookieName: If any of the incoming cookies' key matches the value of cookieName, the JDBC driver will not send any login credentials/Kerberos ticket to the server. The client will just send the cookie alone back to the server for authentication purpose. The default value of cookieName is hive.server2.auth (this is the HiveServer2 cookie name). 
• To turn off cookie replay, cookieAuth=false must be used in the JDBC URL.
• Important Note: As part of HIVE-9709, we upgraded Apache http-client and http-core components of Hive to 4.4. To avoid any collision between this upgraded version of HttpComponents and other any versions that might be present in your system (such as the one provided by Apache Hadoop 2.6 which uses http-client and http-core components version of 4.2.5), the client is expected to set CLASSPATH in such a way that Beeline related jars appear before HADOOP lib jars. This is achieved via setting HADOOP_USER_CLASSPATH_FIRST=true before using hive-jdbc. In fact, in bin/beeline.sh we do this!

Using 2-way SSL in HTTP Mode 

HIVE-10447 enabled the JDBC driver to support 2-way SSL in HTTP mode. Please note that HiveServer2 currently does not support 2-way SSL. So this feature is handy when there is an intermediate server such as Knox which requires client to support 2-way SSL.

JDBC connection URL: jdbc:hive2://<host>:<port>/<db>;ssl=true;twoWay=true;sslTrustStore=<trust_store_path>;trustStorePassword=<trust_store_password>;sslKeyStore=<key_store_path>;keyStorePassword=<key_store_password>?hive.server2.transport.mode=http;hive.server2.thrift.http.path=<http_endpoint>

• <trust_store_path> is the path where client's truststore file lives. This is a mandatory non-empty field.
• <trust_store_password> is the password to access the truststore.
• <key_store_path> is the path where client's keystore file lives. This is a mandatory non-empty field.
• <key_store_password> is the password to access the keystore.

Passing HTTP Header Key/Value Pairs via JDBC Driver

HIVE-10339 introduced an option for clients to provide custom HTTP headers that can be sent to the underlying server (Hive 1.2.0 and later).

JDBC connection URL: jdbc:hive2://<host>:<port>/<db>?hive.server2.transport.mode=http;hive.server2.thrift.http.path=<http_endpoint>;http.header.<name1>=<value1>;http.header.<name2>=<value2>

When the above URL is specified, Beeline will call underlying requests to add an HTTP header set to <name1> and <value1> and another HTTP header set to <name2> and <value2>. This is helpful when the end user needs to send identity in an HTTP header down to intermediate servers such as Knox via Beeline for authentication, for example http.header.USERNAME=<value1>;http.header.PASSWORD=<value2>.