Hadoop Clustering on Single machine

Can I install a Hadoop cluster on a single machine? and How I can try its scalability features?

Regards,
Mohamed

Yes, you can try a cluster on single machine. The good way to try out this is by using virtual machines.
I have tried it with VM of Centos O/S. The configuration I tried on my Windows 7 system is

Hardware: 4 gb ram @ 1600mhz, Interl i5 processor and 5400 prm hdd.

Software: Cloudera cdhv3, VMWare player v5, the virtual machine files for Centos O/S

Hadoop configuration: One Master node( Holds namenode, secondary name node and jobtracker), Three slave nodes (runs datanode and tasktracker on each)

This set up will help you to learn all the aspects of Hadoop technology.

Share if you find any difficuilties to try out it, am happy to help you!

Arun Vasu wrote:Yes, you can try a cluster on single machine. The good way to try out this is by using virtual machines.
I have tried it with VM of Centos O/S. The configuration I tried on my Windows 7 system is

Hardware: 4 gb ram @ 1600mhz, Interl i5 processor and 5400 prm hdd.

Software: Cloudera cdhv3, VMWare player v5, the virtual machine files for Centos O/S

Hadoop configuration: One Master node( Holds namenode, secondary name node and jobtracker), Three slave nodes (runs datanode and tasktracker on each)

This set up will help you to learn all the aspects of Hadoop technology.

Share if you find any difficuilties to try out it, am happy to help you!

Thank you for sharing this, did you face any difficulties in network configurations for this cluster of VMs?

Make sure you have given proper entries in /etc/hosts file in the form of "ip address<tab>host_name".
Edit /etc/sysconfig/network and change the HOSTNAME= line to reflect your machine’s name. And your machine has an internet connection.

These are the only necessary network configuration you needs.