|Oracle® Big Data Connectors User's Guide
Release 2 (2.0)
Part Number E36961-03
|PDF · Mobi · ePub|
This chapter describes the Oracle Big Data Connectors, provides installation instructions, and identifies the permissions needed for users to access the connectors.
This chapter contains the following sections:
Oracle Big Data Connectors facilitate data access between data stored in a Hadoop cluster and Oracle Database. They can be licensed for use on either Oracle Big Data Appliance or a Hadoop cluster running on commodity hardware.
These are the connectors:
Oracle SQL Connector for Hadoop Distributed File System (previously Oracle Direct Connector for HDFS): Enables Oracle Database to access data stored in Hadoop Distributed File System (HDFS) files or a Hive table. The data can remain in HDFS or the Hive table, or it can be loaded into an Oracle database. Oracle SQL Connector for HDFS is a command-line utility that accepts generic command line arguments supported by the org.apache.hadoop.util.Tool interface. It also provides a preprocessor for Oracle external tables.
Oracle Loader for Hadoop: Provides an efficient and high-performance loader for fast movement of data from a Hadoop cluster into a table in an Oracle database. Oracle Loader for Hadoop prepartitions the data if necessary and transforms it into a database-ready format. It optionally sorts records by primary key or user-defined columns before loading the data or creating output files. Oracle Loader for Hadoop is a MapReduce application that is invoked as a command-line utility. It accepts the generic command-line options that are supported by the org.apache.hadoop.util.Tool interface.
Oracle Data Integrator Application Adapter for Hadoop: Extracts, transforms, and loads data from a Hadoop cluster into tables in an Oracle database, as defined using a graphical user interface.
Oracle R Connector for Hadoop: Provides an interface between a local R environment, Oracle Database, and Hadoop, allowing speed-of-thought, interactive analysis on all three platforms. Oracle R Connector for Hadoop is designed to work independently, but if the enterprise data for your analysis is also stored in Oracle Database, then the full power of this connector is achieved when it is used with Oracle R Enterprise.
Individual connectors may require that software components be installed in Oracle Database and the Hadoop cluster. Users may also need additional access privileges in Oracle Database.
See Also:My Oracle Support Information Center: Big Data Connectors (ID 1487399.2) and its related information centers.
Enterprises are seeing large amounts of data coming from multiple sources. Click-stream data in web logs, GPS tracking information, data from retail operations, sensor data, and multimedia streams are just a few examples of vast amounts of data that can be of tremendous value to an enterprise if analyzed. The unstructured and semi-structured information provided by raw data feeds is of little value in and of itself. The data needs to be processed to extract information of real value, which can then be stored and managed in the database. Analytics of this data along with the structured data in the database can provide new insights into the data and lead to substantial business benefits.
MapReduce is a parallel programming model for processing data on a distributed system. It can process vast amounts of data in a timely manner and can scale linearly. It is particularly effective as a mechanism for batch processing of unstructured and semi-structured data. MapReduce abstracts lower level operations into computations over a set of keys and values.
A simplified definition of a MapReduce job is the successive alternation of two phases, the map phase and the reduce phase. Each map phase applies a transform function over each record in the input data to produce a set of records expressed as key-value pairs. The output from the map phase is input to the reduce phase. In the reduce phase, the map output records are sorted into key-value sets so that all records in a set have the same key value. A reducer function is applied to all the records in a set and a set of output records are produced as key-value pairs. The map phase is logically run in parallel over each record while the reduce phase is run in parallel over all key values.
Note:Oracle Big Data Connectors do not support the Yet Another Resource Negotiator (YARN) implementation of MapReduce.
Apache Hadoop is the software framework for the development and deployment of data processing jobs based on the MapReduce programming model. At the core, Hadoop provides a reliable shared storage and analysis systemFoot 1 . Analysis is provided by MapReduce. Storage is provided by the Hadoop Distributed File System (HDFS), a shared storage system designed for MapReduce jobs.
The Hadoop ecosystem includes several other projects including Apache Avro, a data serialization system that is used by Oracle Loader for Hadoop.
Cloudera's Distribution including Apache Hadoop (CDH) is installed on Oracle Big Data Appliance. You can use Oracle Big Data Connectors on a Hadoop cluster running CDH or the equivalent Apache Hadoop components, as described in the setup instructions in this chapter.
For conceptual information about the Hadoop technologies, the following third-party publication:
Hadoop: The Definitive Guide, Third Edition by Tom White (O'Reilly Media Inc., 2012, ISBN: 978-1449311520).
For information about Cloudera's Distribution including Apache Hadoop (CDH4), the Oracle Cloudera website at
For information about Apache Hadoop, the website at
To download from Oracle Technology Network:
Use any browser to visit this website:
Click the name of each connector to download a zip file containing the installation files.
To download from Oracle Software Delivery Cloud:
Use any browser to visit this website:
Accept the Terms and Restrictions to see the Media Pack Search page.
Select the search terms:
Select a Product Pack: Oracle Database
Platform: Linux x86-64
Click Go to display a list of product packs.
Select Oracle Big Data Connectors Media Pack for Linux x86-64 (B65965-0x), and then click Continue.
Click Download for each connector to download a zip file containing the installation files.
Oracle SQL Connector for Hadoop Distributed File System (HDFS) is installed and configured on the system where Oracle Database runs. It is also installed and runs on the system where Hive is installed, if Hive tables are used as data sources.
Before installing Oracle SQL Connector for HDFS, verify that you have the required software.
Oracle SQL Connector for HDFS requires the following software:
Java Development Kit (JDK) 1.6_08 or later. Consult the distributor of your Hadoop software (Cloudera or Apache) for the recommended version.
Oracle Database release 11g release 2 (188.8.131.52 or 184.108.40.206) for Linux.
To support the Oracle Data Pump file format, an Oracle Database one-off patch. To download this patch, go to
http://support.oracle.com and search for bug 14557588.
The same version of JDK on the Oracle Database system as your Hadoop cluster.
Hive 0.7.0, 0.7.1, or 0.9.0 (required for Hive table access, otherwise optional)
Oracle SQL Connector for HDFS works as a Hadoop client. You do not need to configure Hadoop on the Oracle Database system to run MapReduce jobs for Oracle SQL Connector for HDFS. However, you must install Hadoop on the Oracle Database system and minimally configure it for Hadoop client use only.
Install and configure the same version of CDH or Apache Hadoop on the Oracle Database system as on the Hadoop cluster. If you are using Oracle Big Data Appliance, then complete the procedures for providing remote client access in the Oracle Big Data Appliance Software User's Guide. Otherwise, follow the installation instructions provided by the distributor (Cloudera or Apache).
Ensure that Oracle Database has access to HDFS:
Log in to the system where Oracle Database is running by using the Oracle Database account.
Open a Bash shell and enter this command:
In this command,
$HADOOP_HOME is the absolute path to the Hadoop home directory. You should see a list of directories. If not, then first ensure that the Hadoop cluster is up and running. If the problem persists, then you must correct the Hadoop client configuration so that Oracle Database has access to the Hadoop cluster file system.
The Oracle Database system is now ready for use as a Hadoop client. No other Hadoop configuration steps are needed.
Complete this procedure to install Oracle SQL Connector for HDFS on the Oracle Database system and, for access to Hive tables, on the Hadoop cluster.
Download the zip file to a directory on the system where Oracle Database runs.
Unzip orahdfs-version.zip into a directory. The unzipped files have the structure shown in Example 1-1.
The hdfs_stream script is the preprocessor script for the HDFS external table.
Run hdfs_stream from the Oracle SQL Connector for HDFS installation directory. You should see this usage information:
$ ./hdfs_stream Usage: hdfs_stream locationFile
Read and execute permissions on the
$ ls -l OSCH_HOME/bin/hdfs_stream -rwxr-xr-x 1 oracle oinstall Nov 27 15:51 hdfs_stream
If you do not see these permissions, then enter a
chmod command to fix them:
$ chmod 755 OSCH_HOME/bin/hdfs_stream
Read permission on
SQL> CREATE OR REPLACE DIRECTORY osch_bin_path AS '/etc/orahdfs-2.0.0/bin'
To provide support for Hive tables, install Oracle SQL Connector for HDFS on the Hadoop cluster. If you only plan to access HDFS files, then you can omit this procedure.
To install Oracle SQL Connector for HDFS on the Hadoop cluster:
Download the zip file to a directory on the system where Hive is installed.
Add the Hive JAR files to the
HADOOP_CLASSPATH environment variable.
The unzipped files have the structure shown in Example 1-1.
EXECUTE on the
OSCH_BIN_PATH directory created during the installation of Oracle SQL Connector for HDFS. Do not grant write access to anyone. Grant
EXECUTE only to those who intend to use Oracle SQL Connector for HDFS.
Example 1-2 shows the SQL commands granting these privileges to
Before installing Oracle Loader for Hadoop, verify that you have the required software.
A target database system running one of the following:
Oracle Database 11g release 2 (220.127.116.11) with required patch
Oracle Database 11g release 2 (18.104.22.168)
Note:To use Oracle Loader for Hadoop with Oracle Database 11g release 2 (22.214.171.124), you must first apply a one-off patch that addresses bug number 11897896. To access this patch, go to
http://support.oracle.comand search for the bug number.
To install Oracle Loader for Hadoop:
oraloader-version-1.x86_64.zip: Use with CDH3 and Apache Hadoop 1.0
oraloader-version-2.x86_64.zip: Use with CDH4
Unzip the appropriate archive into a directory on your Hadoop cluster.
jlib lib examples
$KVHOME/lib/kvstore-2.0.22.jar to the
HADOOP_CLASSPATH variable, if you are reading data from Oracle NoSQL Database Release 2.
Installation requirements for Oracle Data Integrator (ODI) Application Adapter for Hadoop are provided in these topics:
To use Oracle Data Integrator Application Adapter for Hadoop, you must first have Oracle Data Integrator, which is licensed separately from Oracle Big Data Connectors. You can download ODI from the Oracle website at
Oracle Data Integrator Application Adapter for Hadoop requires a minimum version of Oracle Data Integrator 126.96.36.199.0.
Before performing any installation, read the system requirements and certification documentation to ensure that your environment meets the minimum installation requirements for the products that you are installing.
The list of supported technologies and versions is available on Oracle Technical Network:
To set up the topology, see Chapter 4, "Oracle Data Integrator Application Adapter for Hadoop."
Oracle R Connector for Hadoop requires the installation of a software environment on the Hadoop side and on a client Linux system.
Oracle Big Data Appliance supports Oracle R Connector for Hadoop without any additional software installation or configuration. However, you do need to verify whether certain R packages are installed. See "Installing Additional R Packages."
However, to use Oracle R Connector for Hadoop on any other Hadoop cluster, you must create the necessary environment.
You must install several software components on a third-party Hadoop cluster to support Oracle R Connector for Hadoop.
Install these components on third-party servers:
Complete the instructions provided by the distributor.
Hive 0.7.1 or 0.9.0
Sqoop for the execution of functions that connect to Oracle Database. Oracle R Connector for Hadoop does not require Sqoop to install or load.
Java Virtual Machine (JVM), preferably Java HotSpot Virtual Machine 6.
Oracle Loader for Hadoop to support the OLH driver (optional). See "Oracle Loader for Hadoop Setup."
Note:Do not set
HADOOP_HOMEon the Hadoop cluster. CDH4 does not need it, and it interferes with Oracle R Connector for Hadoop when it checks the status of the JobTracker. This results in the error "Something is terribly wrong with Hadoop MapReduce."
Sqoop provides a SQL-like interface to Hadoop, which is a Java-based environment. Oracle R Connector for Hadoop uses Sqoop for access to Oracle Database.
Note:Sqoop is required even when using Oracle Loader for Hadoop as a driver for loading data into Oracle Database. Sqoop performs additional functions, such as copying data from a database to HDFS and sending free-form queries to a database. The OLH driver also uses Sqoop to perform operations that Oracle Loader for Hadoop does not support.
To install and configure Sqoop for use with Oracle Database:
Install Sqoop if it is not already installed on the server.
For Cloudera's Distribution including Apache Hadoop, see the Sqoop installation instructions in the CDH Installation Guide at
Provide Sqoop with the connection string to Oracle Database.
$ sqoop import --connect jdbc_connection_string
sqoop import --connect jdbc:oracle:thin@myhost:1521/orcl.
Hive provides an alternative storage and retrieval mechanism to HDFS files through a querying language called HiveQL. Oracle R Connector for Hadoop uses the data preparation and analysis features of HiveQL, while enabling you to use R language constructs.
To install Hive:
Follow the instructions provided by the distributor (Cloudera or Apache) for installing Hive.
Verify that the installation is working correctly:
$ hive -H usage: hive -d,--define <key=value> Variable subsitution to apply to hive commands. e.g. -d A=B or --define A=B --database <databasename> Specify the database to use -e <quoted-query-string> SQL from command line -f <filename> SQL from files -H,--help Print help information -h <hostname> connecting to Hive Server on remote host --hiveconf <property=value> Use value for given property --hivevar <key=value> Variable subsitution to apply to hive commands. e.g. --hivevar A=B -i <filename> Initialization SQL file -p <port> connecting to Hive Server on port number -S,--silent Silent mode in interactive shell -v,--verbose Verbose mode (echo executed SQL to the console)
If the command fails or you see warnings in the output, then fix the Hive installation.
Alternatively, you can download R from a Comprehensive R Archive Network (CRAN) website at
ORCH is the name of the Oracle R Connector for Hadoop package.
To install the ORCH package:
Set the environment variables for the supporting software:
$ setenv JAVA_HOME /usr/lib/jdk6 $ setenv R_HOME /usr/lib64/R $ setenv SQOOP_HOME /usr/lib/sqoop
Unzip the downloaded file:
$ unzip orch-version.zip Archive: orch-linux-x86_64-2.0.zip extracting: ORCH2.0/ORCH_2.0_R_x86_64-unknown-linux-gnu.tar.gz inflating: ORCH2.0/ORCHcore_2.0_R_x86_64-unknown-linux-gnu.tar.gz inflating: ORCH2.0/OREbase_1.3_R_x86_64-unknown-linux-gnu.tar.gz inflating: ORCH2.0/OREstats_1.3_R_x86_64-unknown-linux-gnu.tar.gz
Change to the new directory:
$ cd ORCH2.0
Install the packages in the exact order shown here:
$ R --vanilla CMD INSTALL OREbase_1.3_R_x86_64-unknown-linux-gnu.tar.gz $ R --vanilla CMD INSTALL OREstats_1.3_R_x86_64-unknown-linux-gnu.tar.gz $ R --vanilla CMD INSTALL ORCHcore_2.0_R_x86_64-unknown-linux-gnu.tar.gz $ R --vanilla CMD INSTALL ORCH_2.0_R_x86_64-unknown-linux-gnu.tar.gz
Your Hadoop cluster must have
libpng-devel installed on every node. This procedure uses the
dcli utility to execute the commands. See the Oracle Big Data Appliance Owner's Guide for the syntax of the
To install libpng-devel:
Log in as
root to any Oracle Big Data Appliance server.
libpng-devel is already installed:
# dcli rpm -qi libpng-devel bda1node01: package libpng-devel is not installed bda1node02: package libpng-devel is not installed . . .
If the package is already installed on all servers, then you can skip this procedure.
If you need a proxy server to go outside your firewall, then set the
HTTP_PROXY environment variable:
# dcli export HTTP_PROXY="http://proxy.example.com"
Download the yum configuration file:
# cd /etc/yum.repos.d # wget http://public-yum.oracle.com/public-yum-el5.repo
Open public-yum-el5.repo in a text editor and make these changes:
Copy the file to the other Oracle Big Data Appliance servers:
# dcli -d /etc/yum.repos.d -f public-yum-el5.repo
Install the package on all servers:
# dcli yum -y install libpng-devel bda1node01: Loaded plugins: rhnplugin, security bda1node01: Repository 'bda' is missing name in configuration, using id bda1node01: This system is not registered with ULN. bda1node01: ULN support will be disabled. bda1node01: http://bda1node01-master.us.oracle.com/bda/repodata/repomd.xml: bda1node01: [Errno 14] HTTP Error 502: notresolvable bda1node01: Trying other mirror. . . . bda1node01: Running Transaction bda1node01: Installing : libpng-devel 1/2 bda1node01: Installing : libpng-devel 2/2 bda1node01: Installed: bda1node01: libpng-devel.i386 2:1.2.10-17.el5_8 ibpng-devel.x86_64 2:1.2.10-17.el5_8 bda1node01: Complete! bda1node02: Loaded plugins: rhnplugin, security . . .
Verify that the installation was successful on all servers:
# dcli rpm -qi libpng-devel bda1node01: Name : libpng-devel Relocations: (not relocatable) bda1node01: Version : 1.2.10 Vendor: Oracle America bda1node01: Release : 17.el5_8 Build Date: Wed 25 Apr 2012 06:51:15 AM PDT bda1node01: Install Date: Tue 05 Feb 2013 11:41:14 AM PST Build Host: ca-build56.us.oracle.com bda1node01: Group : Development/Libraries Source RPM: libpng-1.2.10-17.el5_8.src.rpm bda1node01: Size : 482483 License: zlib bda1node01: Signature : DSA/SHA1, Wed 25 Apr 2012 06:51:41 AM PDT, Key ID 66ced3de1e5e0159 bda1node01: URL : http://www.libpng.org/pub/png/ bda1node01: Summary : Development tools for programs to manipulate PNG image format files. bda1node01: Description : bda1node01: The libpng-devel package contains the header files and static bda1node01: libraries necessary for developing programs using the PNG (Portable bda1node01: Network Graphics) library. . . .
Whereas R users will run their programs as MapReduce jobs on the Hadoop cluster, they do not typically have individual accounts on that platform. Instead, an external Linux server provides remote access.
To provide access to a Hadoop cluster to R users, install these components on a Linux server:
The same version of Hadoop as your Hadoop cluster; otherwise, unexpected issues and failures can occur
The same version of Sqoop as your Hadoop cluster; required only to support copying data in and out of Oracle databases
Mahout; required only for the
orch.ls function with the Mahout ALS-WS algorithm
The same version of the Java Development Kit (JDK) as your Hadoop cluster
ORCH R package
To provide access to database objects, you must have the Oracle Advanced Analytics option to Oracle Database. Then you can install this additional component on the Hadoop client:
Oracle R Enterprise Client Packages
You must install Hadoop on the client and minimally configure it for HDFS client use.
To install and configure Hadoop on the client system:
Install and configure CDH3 or Apache Hadoop 0.20.2 on the client system. This system can be the host for Oracle Database. If you are using Oracle Big Data Appliance, then complete the procedures for providing remote client access in the Oracle Big Data Appliance Software User's Guide. Otherwise, follow the installation instructions provided by the distributor (Cloudera or Apache).
Log in to the client system as an R user.
Open a Bash shell and enter this Hadoop file system command:
$HADOOP_HOME/bin/hadoop fs -ls /user
If you see a list of files, then you are done. If not, then ensure that the Hadoop cluster is up and running. If that does not fix the problem, then you must debug your client Hadoop installation.
Complete the same procedures on the client system for installing and configuring Sqoop as those provided in "Installing Sqoop on a Hadoop Cluster".
When you are done, ensure that users have the necessary permissions to connect to the Linux server and run R.
You may also want to install RStudio Server to facilitate access by R users. See the RStudio website at
Complete the procedures on the client system for installing
ORCH as described in "Installing the Software on Hadoop".
To support full access to Oracle Database using R, install the Oracle R Enterprise Release 1.3 or later client packages. Without them, Oracle R Connector for Hadoop does not have access to the advanced statistical algorithms provided by Oracle R Enterprise.
See Also:Oracle R Enterprise User's Guide for information about installing R and Oracle R Enterprise
Footnote LegendFootnote 1: Hadoop: The Definitive Guide, Third Edition by Tom White (O'Reilly Media Inc., 2012, 978-1449311520).