With eurucamp 2012 being just under 3 weeks away, we wanted to start introducing our sponsors.
Already a eurucamp 2011 sponsor, Scalarium were generous enough to join us once again in 2012, this time as a top-level Kastanien sponsor.
Scalarium has been on the hosting scene for a while now and their Berlin-based team is still growing. They are always on the look out for talented developers – so if you <3 automation and clouds – apply for an open position online or get in contact with the Scalarium team during eurucamp.
As for their story? We'll let them tell explain: –
It all started with the guys at Peritor working hard on consulting gigs and helping people move their applications to Amazon EC2. Doing this is complex and involves a lot of repetition. Also Amazon EC2 lacks some fundamental capabilities like automatic replacement of dead hosts, easy configuration, or application deployment support. So Scalarium was born in mid 2009 to fill that gap. Scalarium now hosts some really big and popular sites and games around the globe - we could drop webscale or big data here – but we don’t ;).
Scalarium is a management layer on top of EC2 that automates the hell out your machines and gives you:
- Configuration with Chef
- Healing and replacement of failed hosts
- Auto scaling by time of day or load metrics on your machines
- Build-in deployment and logging
- Extensive monitoring and tracking
The most important component of Scalarium is the local agent running on every machine. It communicates the server state up to Scalarium and responds to changes by executing Chef runs. Scalarium uses Chef to form servers into your desired state. What differentiates Scalarium from a simple Chef server setup is that Scalarium generates life-cycle events (like setup, deployment or configuration state change) and pushes them down to all machines rather then relying on Chef server to pull. By having multiple life-cycle events you can tell Scalarium what to do when servers come or go or how to respond to a new application server (e.g. by adding it to the load balancer!).
You model your architecture by using build-in roles (like Rails app server, MySQL server, load balancer or Node.js app server) or by creating custom definitions. Thanks to the power of Chef, Scalarium can automate anything scriptable.
Using the provided auto scaling mechanisms many clients scale down machines during the night and boot them over the day as needed. This saves a lot of money as you only pay for what you are actually using.
Our team is based in Berlin and primarily works with Sinatra/Rails for the API and Frontend, Nanite, RabbitMQ, CouchDB, Redis, Chef, and of course Ruby for the backend.