The blog covers various articles and posts on Cloud, Big Data Analytics, Data Science, Machine Learning, DevOps, Full Stack Development, Java and Middleware Technologies
Saturday, March 2, 2019
Hadoop HDFS Admin - Network Topology Of Cluster
sudo vi /etc/hadoop/conf/core-site.xml
<property>
<name>net.topology.node.switch.mapping.impl</name>
<value>org.apache.hadoop.net.ScriptBasedMapping</value>
</property>
<property>
<name>net.topology.script.file.name</name>
<value>/etc/hadoop/conf/topology.sh</value>
</property>
sudo vi /etc/hadoop/conf/topology.data
10-10-10-10 /rack1
20-20-20-20 /rack1
30-30-30-30 /rack1
40-40-40-40 /rack2
--Topology script will be called when Namenode starts
./etc/hadoop/conf/topology.sh 10-10-10-10 20-20-20-20 30-30-30-30 40-40-40-40
sudo vi /etc/hadoop/conf/topology.sh
#!/bin/bash
HADOOP_CONF=/etc/hadoop/conf
while [ $# -gt 0 ] ; do
nodeArg=$1
exec< ${HADOOP_CONF}/topology.data
result=""
while read line ; do
ar=( $line )
if [ "${ar[0]}" = "$nodeArg" ] ; then
result="${ar[1]}"
fi
done
shift
if [ -z "$result" ] ; then
echo -n "/rack0"
else
echo -n "$result "
fi
done
--IP Address encoded with rack information
xx.xx.rackno.nodeno
50-50-50-50
Labels:
Hadoop