Step 1: Set up a Hadoop Virtual Machine instance

The easiest way to get started sampling searches in Hunk is to install a Virtual Machine that comes preconfigured with Hadoop.

For this tutorial we are using using the Cloudera Quickstart VM for VMware. See System and Software requirements for the full list of supported Hadoop distributions and versions.

Setting up your Virtual Machine for this tutorial

The examples in this tutorial use a Cloudera Quickstart Virtual Machine. If you are using another VM with Hadoop instance, see that product's directions for installation and setup. You may also need to modify the examples for your particular configuration.

If you wish to try out Hunk using YARN, we recommend you try using the Hortonworks Sandbox 2.0 here.

If you using Cloudera Quickstart for VM:

1. untar the Cloudera Quickstart VM on your computer:

tar –xvf cloudera-quickstart-vm-4.3.0-vmware.tar.gz

2. Start and access the virtual machine.

3. Import the OVF file from VMware Fusion.

4. Start the VM and open the terminal to find the IP address of your virtual machine.

Using this tutorial with YARN

If you want to try out Hunk using YARN, we recommend you try out the tutorial using the Hortonworks Sandbox 2.0.

Note: You can also use one of the virtual machines provided by Hortonworks here: http://hortonworks.com/products/hortonworks-sandbox.

You might need to assign more memory to YARN that is defined by default. We recommend that YARN have at least 5120 MB assigned to it (how much you are able to assign to it depends on how much memory is assigned to your VM) You can do this by editing the following property in yarn-site.xml.:

<name>yarn.nodemanager.resource.memory-mb</name> <value>5120</value>