In its great expanse of servers, Facebook has deployed Opscode to help scale the social network’s infrastructure. To do this, Facebook is using Private Chef, the web-scale automation technology that Opscode updated today.
Facebook additionally worked with Opscode to deliver Chef 11, which has been rewritten from the bottom up with the Erlang programming language and PostgreSQL database. Opscode 11 delivers an automation platform that can scale up to 10,000 nodes from a single Chef server.
Without getting too deep in the weeds, Facebook needs ways to scale its infrastructure without hiring armies of IT people. Private Chef is Opscode’s offering that it has customized from the Chef “cookbook,” for large enterprises to:
- Manage servers by writing code that automatically keeps servers configured.
- Integrate tightly with applications, including databases and LDAP directories.
- Configure applications that require knowledge about the entire infrastructure.
Facebook, like Google and other Internet services, faces steeper scaling issues than enterprise companies but the comparisons are relevant as loads increase. These are companies that have what Opscode executives like to call “Carl Sagan,” size deployments. For those unfamiliar, according to Wikipedia, Sagan was a well-known personality in astronomy and natural sciences. His teachings helped people grasp the enormity of space.
With the amount of data flowing through these networks , scale has multiple dimensions. There are more servers, more configurations and the need for people with the technical capabilities to manage it all.
Data is at the heart of what makes Opscode one of those companies that is helping create a new infrastructure for enterprise companies. Puppet Labs is one of Opscode’s biggest rivals. Last month, Puppet raised $30 million from VMware.