Additional Configuration for a MapR Hadoop Distribution

For a MapR Hadoop distribution, you must do the following:

If you run the Hadoop tracer script on MapR and are running on a Windows client, you must manually add the following property to the mapred-site.xml file on the SPD Server machine. Otherwise, an error occurs when you submit a program to Hadoop.
```
<property>
   <name>mapreduce.app-submission.cross-platform</name>
   <value>true</value>
</property>
```
In the spdsserv.parm parameter file, modify the java.security.auth.login property in JREOPTIONS to point to the mapr.login.conf file. Normally, the mapr.login.conf file is installed in the MAPR_HOME/conf directory.
Here is an example:
```
-jreoptions (-Djava.security.auth.login.config=/opt/mapr/conf/mapr.login.conf)
```
On the Microsoft Windows operating system, you must add the group identifier and the user identifier properties to the mapred-site.xml and yarn-site.xml Hadoop cluster configuration files. You do this to describe the account that is responsible for SPD Server. Here is an example. Replace the property attributes in italics with the correct information at your site.
```
<property>
    <name>hadoop.spoofed.user.uid</name>
    <value>4658</value>
  </property>
  <property>
    <name>hadoop.spoofed.user.gid</name>
    <value>100</value>
  </property>
  <property>
    <name>hadoop.spoofed.user.username</name>
    <value>spdsmgr</value>
  </property>
```
On the Linux operating system if the Hadoop cluster supports Kerberos, you must add the following commands to the rc.spds script, which is located in the InstallDir/site directory:
```
MAPR_HOME=/opt/mapr
export MAPR_HOME
```
If the Hadoop cluster supports Kerberos, the MapR ticket expiration and renewal expiration times must be at least as long as the Kerberos ticket expiration and renewal expiration times. SPD Server uses the Kerberos ticket expiration and renewal expiration times to manage MapR tickets. They are renewed and re-created at the same time as the Kerberos ticket.