Installing the IBM Spectrum Symphony client using the .rpm file (Linux root only)

If you have root permissions on a Linux® host, you can install using the default settings.

Procedure

Define the cluster administrator account by setting the CLUSTERADMIN environment variable. Set the CLUSTERADMIN variable to any valid OS user account, which then owns all installation files.
For example:
```
setenv CLUSTERADMIN egoadmin
```
If you are want to use the MapReduce framework supported by IBM® Spectrum Symphony Advanced Edition, configure environment variables required to run MapReduce workload.

Note: This step is optional before installation. You can choose to skip this step now and configure environment variables (as described in step 4) after installing the IBM Spectrum Symphony client.
1. Configure JAVA_HOME to the directory in which you have installed Java™. The MapReduce framework requires Java Development Kit (JDK) 1.6.0_20 or higher.
  For example, on csh:
```
setenv JAVA_HOME /usr/java/jdk1.6.0_25
```
2. Configure HADOOP_HOME to the directory under which Hadoop has been installed.
  For example, on csh:
```
setenv HADOOP_HOME /opt/hadoop-0.20.2
```
3. Configure HADOOP_VERSION to the Hadoop 2.7.2:
  For example:
  - (bsh) export HADOOP_VERSION=2_7_2
  - (tcsh) setenv HADOOP_VERSION 2_7_2
Run the .rpm file:
```
rpm -ivh install_file.rpm
```
If you want to use the MapReduce framework and you did not complete step 2, configure the environment variables required to run MapReduce workload:
1. Go to the mapreduce/conf directory where the IBM Spectrum Symphony client was installed.
  
  For example, if you installed the IBM Spectrum Symphony client in the /opt/ibm/spectrumcomputing/client732 directory, go to the /opt/ibm/spectrumcomputing/client732/mapreduce/conf directory.
2. Configure the JAVA_HOME environment variable to the directory in which you installed Java. Java Development Kit (JDK) 1.6.0_20 or higher is required for the MapReduce framework to run:
  1. In the pmr-env.sh file, set JAVA_HOME to the Java installation directory.
    For example:
    export JAVA_HOME=/usr/java/jdk1.6.0_25
  2. Configure JAVA_HOME again in cshrc.pmrclient or profile.pmrclient file:
    - For csh, edit the cshrc.pmrclient file and change the following line to the directory in which you installed Java:
      setenv JAVA_HOME $JAVA_HOME
      For example:
      setenv JAVA_HOME /usr/java/jdk1.6.0_25
    - For bash, edit the profile.pmrclient file and change the following line to the directory in which you installed Java:
      JAVA_HOME=$JAVA_HOME
      For example:
      JAVA_HOME=/usr/java/jdk1.6.0_25

What to do next

Optional: Customize your installation

To install as root without using the default database path or installation directory, use the --dbpath and --prefix options.

Define the cluster administrator account by setting the CLUSTERADMIN environment variable. Set the CLUSTERADMIN variable to any valid OS user account, which will then own all installation files.
If you are want to use the MapReduce framework supported by IBM Spectrum Symphony Advanced Edition, configure environment variables required to run MapReduce workload as described in the previous section.
Run the .rpm file:
```
rpm -ivh --prefix /install_dir --dbpath dbpath_dir install_file.rpm
```
Note: If you install the .rpm file without options, the IBM Spectrum Symphony client is installed under the /opt/ibm/spectrumcomputing/client732 directory.