User Guide

10.0
- 11.0

Back Next

Defining Hadoop Targets

You can configure Fast Clone to unload source data to output files on a Cloudera, Hortonworks, or MapR distribution of Hadoop or to a Hive data warehouse target. With these Hadoop target types, Fast Clone does not generate a load script for a target load utility.

On the

Target DB

tab, select

CLOUDERA

HIVE

HORTONWORKS

, or

MAPR

in the

Database Type

list.

Enter the version, host name, port number, and user name for connecting to the target.

For MapR targets, you can specify a cluster name instead of a host name in the

Hostname

field. Use the following syntax:

/mapr/cluster_name

The cluster_name value must match the cluster name in the

$MAPR_HOME/conf/mapr-clusters.conf

configuration file.

When you specify a cluster name, the Fast Clone Console does not use the port number that you specify in the

Port

field.

To specify the directory in the distributed file system for the Cloudera, Hortonworks, or MapR output files, click

Select

and then either browse to the directory or enter the path to the directory in the

Path

field.

The user name that you specify for connecting to the Hadoop target must have write permissions on this directory.

Click

Connect

The Fast Clone Console connects to the target.

The Fast Clone Console requires Java Runtime Environment (JRE) 1.7 x64 or later to connect to these Hadoop targets.

For Hive targets, select a database in the

Database

list.

The Fast Clone Console uses the definition of the selected database to determine the location of flat files that store table data on the HDFS.

Before you unload source data to a Hive target, enter the field delimiter that the Hive target uses in data files in the

Column separator

field on the

Runtime Settings

tab >

Format Settings

view. By default, Hive data warehouses use the ASCII start of heading (SOH) character as the field delimiter.

Rename Saved Search

Table of Contents

User Guide

User Guide

Defining Hadoop Targets

Defining Hadoop Targets