I am not able to start the Data Node in the Hadoop cluster after installation completed on the Ubuntu operating system. Getting below error in the command line.
DataNode Error:
HTTP ERROR 500 in the browser while accessing the data node
Can't browse the DFS since there are no live nodes available to redirect to Caused by: java.io.IOException: Can't browse the DFS since there are no live nodes available to redirect to at org.apache.hadoop.hdfs.server.namenode.NamenodeJspHelper.redirectToRandomDataNode(NamenodeJSpHelper.java:789) at org.apche.hadoop.hdfs.server.namenode.nn_005fbrowserdfscontent_jsp._jspService(nn_005fbrowsedfscontent_jsp.java:80) at org.apache.jasper.runtime.HttpJspBase.service(HttpJspBase.java:90) at javax.servlet.http.HttpServlet.service(HttpServlet.java:760)
Solution:
Step 1: First, stop all services using the below command:
bin/stop-all.sh
Step 2: Remove your username from Hadoop tmp folder
rm -Rf /app/tmp/hadoop - username/*
Step 3: After that format Hadoop Namenode
hadoop namenode -format
Step4: Change the permission for hdfs folder
chmod -R 777 /home/hdfsuser/hdfs
Step5: After completion of the above steps then start all Hadoop services.
bin/start-all.sh
Step 6: To check all services using the below command:
jps
Summary: I have installed Hadoop single node cluster on Ubuntu operating system. Then I started all Hadoop services using “start-all.sh” command. After that checked all daemons using “jps” commands it showing everything but I accessed data node WebUI, it’s not working so I tried the above steps then stated the services. In Cloudera to stop all services using Cloudera Manager WebUI then start the services one by one. Coming to MapR distribution also stop all running services using the MapR Control System(MCS).