Trying to write to HDFS on a remote machine using Faunus

Question

I am trying to pull data from a Titan-Cassandra graph database and write it to a single Hadoop node using Faunus. The Hadoop node is running on a remote machine. So, the machine on which Faunus is running acts like a source from which data is streamed and this has to be written to a remote single Hadoop node.

Inside titan-cassandra-input.properties, I specify that the output be written to remote HDFS by specifying the output location:

faunus.output.location=hdfs://10.143.57.157:9000/tmp/foutput

I changed the Hadoop configs:

core-site.xml

<configuration>
    <property>
        <name>fs.default.name</name>
        <value>hdfs://10.143.57.244:9000/</value>
    </property>
    <property>
        <name>dfs.permissions</name>
        <value>false</value>
    </property>
</configuration>

mapred-site.xml

<configuration>
    <property>
        <name>mapred.job.tracker</name>
        <value>10.143.57.244:9001</value> 
    </property>
</configuration>

I have added the source IP to /etc/hosts

10.143.57.244  hadoop2

But when I try to start Hadoop with ./start-all.sh, I see that the NameNode is not getting started. When I see the NameNode logs, I see the following error:

ERROR org.apache.hadoop.hdfs.server.namenode.NameNode: java.net.BindException: 
Problem binding to master/10.143.57.244:9000 : Cannot assign requested address

I am not able to make out why it is trying to bind to the source IP. Is it treating the source IP as another node in the Hadoop cluster?

I do not want to setup a cluster. I just want the Hadoop node to listen to any connections from the source IP. How do I configure this? Please help.