SPECS Enabling Platform -- Cluster Launcher

This repository contains the files necessary to setup the SPECS Cluster Launcher.

This is an open-source project released under the Apache 2.0 license (see the "Notice" section below).

This project has been developed under SPECS Project (

Table of contents


A running Linux environment with the following dependencies:

Initial environment setup

On a clean OpenSUSE 13.1 environment, with all the dependencies setup, the following steps are required to prepare the environment (note: replace with your custom domain instead):

    zypper ref
    zypper install apache2 mercurial
    a2enmod proxy
    mkdir -p /srv/www/
    cd /srv/www/
    hg clone .
    cp .apache-vhost.d/ /etc/apache2/vhost.d/
    chown -R +x *
    service apache2 restart

How to use mOS Cluster Launcher

In order to use mOS Cluster Launcher an end user (EU) must have:

  • valid authentification credentials for SPECS Testbed (Eucalyptus) or Amazon EC2;
  • a valid token to be used with mOS Cloud Resource Allocator;

One the EU obtained the credentials it can start using the Cluster Launcher via the web interface. The web interface is available at (replace with the configured domain).

The Cluster Launcher interface is splitted in three sections: Provider, Cluster formation and Nodes.


In Provider section the EU must add specific cloud provider information. After choosing a cloud provider, information about Credentials or Connection must be filled in. Credentials reffers to the cloud provider authentication information, like Access key/Private key pair. Moreover, some connection attributes must be set like client SSH key name (used for passwordless authentication) and security group (a list of ACLs used to restrict the access to the VMs). These are provider specific information and the methodology to obtain it is different. Please consult the cloud provider documentation on how to obtain this information.

Cluster formation

In Cluster formation the EU must define information regarding the cluster identification and what repositories (Chef repositories) need to be loaded from which the nodes will be deployed and configured.

  • Cluster ID : represents a unique string that will identify the cluster; when the EU enters an id the Cluster Launcher will check if the chosen id is not already in use;
  • Mechanisms Chef Repository : represents the Chef repository URL from which the Mechanisms chef recipes will be downloaded;
  • Enabling-Platform Chef Repository : represents the Enabling Platform chef recipes repository; by clicking Validate the Cluster Launcher will test and load the chef recipes descriptors into the web interface;


In Nodes the EU creates its cluster configuration in terms of VMs. The EU can add multiple VMs (using Add node button) or delete unwanted extra nodes. Moreover for each node the EU can choose the list of recipes to be applied at boostrap phase and what type of VM to acuquire (in terms of hardware specification).

Launch the cluster

After defining all the information the EU can now launch the cluster to be deployed. Before launching the Cluster Launcher will provide to the EU the JSON descriptor of the selected cluster instance configuration.

Monitoring the cluster instances

The EU has a special dashboard, called Sessions, from which she/he can monitor the status of the cluster initialization.


This product includes software developed at "Institute e-Austria, Timisoara".


  • Silviu Panica ( / )


  Copyright 2010-2016, Institute e-Austria, Timisoara, Romania

  Licensed under the Apache License, Version 2.0 (the "License");
  you may not use this file except in compliance with the License.
  You may obtain a copy of the License at:

  Unless required by applicable law or agreed to in writing, software
  distributed under the License is distributed on an "AS IS" BASIS,
  See the License for the specific language governing permissions and
  limitations under the License.