Install Apache Ambari server & clients on Digitalocean

This project installs ambari-server and also prepares droplets serving as clients for Ambari. It also adds FQDN for both server and clients.

This project assumes that you use digitalocean and have setup a domain under digitalocean. This project is tested using Ubuntu 14.04 droplets.

First, put your credentials in config.rb

Edit the following file accordingly:

#!/usr/bin/ruby

@docean_key       = 'your token api v2 key' // read from env `ENV['DIGOCEAN_TOKEN']`
@region = 'fra1'
@sshkey = XXX
@domain = 'zouzias.org'  # You domain name
@image  = 'ubuntu-14-04-x64'


#
## Important : Droplet RAM must be >= 8gb (if use hive/hdfs/spark)

@client_size   = '8gb'

SSH key for fabric

Change the following line in fabfile.py to point to the ssh key associated with the above VMs env.key_filename = '~/.ssh/digitalocean_rsa' Also update the domain name.

Update hostname for clients and server

Update ambari-server.txt to your desired name for ambari-server.

Similary, change ambari-clients.txt to the desired hostname for all ambari clients. Each line will correspond to a new droplet!

How to setup Apache ambari server

ruby createAmbariServer.rb
fab init                    # install ambari repo info
fab install_ntp             # install network time protocol
fab install_ambari_server   # install ambari-server package
fab ambari_start            # Start ambari-server process

Make sure that the ambari-server process is running (connect to server droplet using ./connect.sh). The last step fab ambari_start might fail, you might need to connect to the droplet and start the server manually.

How to setup Apache ambari clients

ruby createAmbariClients.rb
fab clients install_ntp  # install network time protocol

Then login on ambari-server.your_domain_name.com:8080 with username/password admin/admin

Requirements:

• Install Ruby and Ruby gems
• Install droplet_kit (gem install droplet_kit)
• Install python >= 2.7.
• Install fabric, i.e., pip install fabric (http://www.fabfile.org/en/latest/).

Optional - Advanced

TODO : Cleanup deployment, i.e., provide clean up functionality after destroy of all droplets.