Tuesday, June 6, 2017


1. Setting prerequisites on linux server
2. Downloading and installing cloudera manager
3. Installing Hadoop

 Setting prerequisites on linux server:

1. Login to linux server as a root user
2. Stop iptables using below command
#chkconfig  --list iptables       
#chkconfig  --level 0123456 iptables off
3. Disable firewall
#service iptables stop
4. Edit the /etc/hosts file and add all the 4 nodes information in it
# vi /etc/hosts

5. Install jdk1.6 or jdk1.7 and set java home and path variables
6. Install python and all its packages
yum install python*

Note : Like above prepare all the 4 nodes

Downloading and installing cloudera manager:

1. Download the cloudera manager software by using wget command as below
2. Give full permissions to the downloaded software
3. Start the cloudera manager installation by running the command as below
4. In the cloudera manager read me screen click next

5. In the cloudera manager license screen click next

6.  To accept the license click yes

7. In the oracle binary code license agreement screen click next

8. To accept the license click yes

9. In the installing screen we can see the cloudera manager installation progress

10. In the next step screen find the cloudera manager login details and click ok

11. In the finish screen click ok to complete the installation

Installing Hadoop:

1. Now we are going to install hadoop using cloudera manager
2. Login to cloudera manager , default user name and password is "admin"

3. For cluster environment select cloudera enterprise trail or cloudera standard and click continue

4. Review the distribution screen and click continue

5. In the specify hosts for cdh screen, specify the hostnames and ip addresses of the nodes we want to add in the cluster

6. Add all the nodes and click on search to check whether nodes are ready or not and then click continue

7. Select the packages we want to install and click continue

8. Select the latest cdh4 release and click continue

9. Select none for impala and soir as they are going to consume lot of ram, we are not installing them now. If we have more ram we can select them also.

10. Leave default for cloudera agent and click continue
11. Provide the root user details and click continue

12. Cloudera hadoop cluster installation starts now

13. After successful completion of cluster installation click continue

14. In the next screen choose the service we want to install for cluster and click continue

15. In the next screen check which services will be install on which node and click continue

16. Select the database type and provide the database details and click on test connection

17. If all the connections got succeeded click continue

18. In the next screen review all the configurations and click continue

19. All the services will start now and it will take some time

20. When all the services were started successfully click continue

21. We will get the success screen now

22. Now login as user 'admin' in Cloudera manager and we can see our clusters running successfully in our environment

No comments:

Post a Comment

  Opatch reports 'Cyclic Dependency Detected' error when patching ODI   Issue: When applying a Patch Set Update (PSU) to WebLogic Se...