An easy way to use your hardware to compute MapReduce applications. qosh is the result of my CS degree thesis.
- Full thesis in Spanish, paper and book in English featuring qosh.
- Keynote presentation slides in English.
- Click the image just below to watch it in action on Youtube.
This piece of software is intented to sample a real world application. It is by no means suitable for production deployments.
- x86_64 architecture and fresh Fedora 17 installation.
- 10GB free HDD space and lots of RAM (4GB+)
- 1h+ of your time...
Install Fedora 17.
sudo yum install -y git
sudo yum update -y
sudo reboot
cd <destination_folder_parent>
git clone https://github.com/marcos-sb/quick-openstacked-hadoop.git
(quick-openstacked-hadoop folder will be created for you)
cd quick-openstacked-hadoop
sudo ./install.py
(It will install OpenStack Folsom, a Hadoop virtual machine, Fabric and Django)
./configure.py
(To set up Django settings)
cd quick-openstacked-hadoop/Alba/albaproject
python manage.py runserver
navigate to http://localhost:8000/albaproject/mapred
user: "udc"
password: "udc"
OpenStack native administration web interface is accessible at http://localhost/dashboard