Puppet - Automated systems administration

Servers and networks are still surprisingly managed manually, even in large environments. This is not only uneconomical and inefficient, but also very prone to error. Managing manually, administrators often find themselves barely keeping up with the growth of their systems.

The magic word is automation: regular recurring tasks such as package installations, configuration changes or computer installations don’t need to be performed manually anymore, rather centrally set up and then rolled out. This way tasks can be carried out quickly and consistently once or a thousand times over in a cluster if need be. And this is where Puppet presents a valuable solution

Centralized configuration

With Puppet, configuration changes are made and stored at a central location in the network. Puppet then distributes these files to its clients and makes the desired changes on the surrounding servers, regardless of their operating systems. Distribution is determined by Puppet through a sophisticated definition language as well as service definitions and computer classification. Necessary configuration data is self-generated by Puppet, making it possible for example, to make a change to all DNS servers without knowing the server names or their IP addresses.

All networks big and small

Puppet is suitable for not only small and heterogeneous, but also for very large networks. In small networks Puppet works with local server access, in large data centers Puppet functions on a client-server setup. Communication is secured between hosts by a SSL based protocol. The central Puppet server distributes the configuration settings to its clients and receives return reports.

• Configuration management
• Configuration distribution
• Version management