Engine Yard powers SOA for the cloud
Engine Yard, which has specialized in Ruby on Rails application-hosting, is introducing Wednesday a platform to extend SOA to the cloud. The company also is extending its Rails stack to the Amazon Web Services (AWS) cloud platform, for quicker deployment of Engine Yard customer applications.
Technologies being unveiled include Vertebra, an open source SOA platform for developing and managing secure cloud applications, and Solo, a deployment service to run the Engine Yard Rails stack on Amazon's system.
( InfoWorld'sTest Center examines the trend of more and more services entering the cloud through Amazon in " Can Capgemini convince enterprises to trust the cloud?" | See also: Burton Group analyst Anne Thomas Manes wrote an "obituary" for SOA. )
Vertebra is intended to orchestrate communication and coherent operations among autonomous agents in many clouds, the company said. The service "future-proofs" for the clouds of tomorrow, according to Engine Yard. Featured is a framework for building cloud-specific applications.
"Vertebra is a way of breaking down applications into small components that can run across many, many servers, and the work can be dispatched across those servers," said Tom Mornini, CTO at Engine Yard. Applications can be sent to servers responsible for the data related to those applications, he said.
Distributed, real-time applications can be built via Vertebra, Engine Yard said. The platform can embrace differences of many clouds and automate processes and application management. The service is initially built to manage customer applications running on the Engine Yard cloud.
Features include XMPP (Extensible Messaging and Presence Protocol) infrastructure, which supports instant messaging (IM). Also highlighted are a security and discovery agent to manage security policy and a process automation agent to orchestrate operational tasks involving machines and people.
An analyst lauded the IM technology in Vertebra.
"I think [Vertebra is] very innovative and it's an interesting use of IM for machine-to-machine communication," said Jay Lyman, open source analyst at The 451 Group. The open source nature of Vertebra also could help it, he added. Engine Yard said it is using XMPP for distributed communications.
Other features include a system provisioning registry for applications to become self-organizing and a federated design enabling applications to operate seamlessly similar to e-mail. Distributed auditing/logging and job control are offered as well.
Vertebra initially supports Ruby and Erlang, but other languages will be added. An early release of Vertebra is available for download on Wednesday, licensed under the Lesser GNU Public License. Engine Yard intends to offer commercial management tools for Vertebra.
With Solo, users can access Engine Yard's capabilities for deploying Ruby on Rails applications and Amazon's cloud infrastructure. While Solo deploys applications on a single virtual machine, a planned follow-up tentatively named Flex would expand deployments to additional virtual machines. Solo enables Engine Yard to deploy applications faster; it is priced at US$129 per month for a single virtual machine.Â
Currently, the Engine Yard stack runs on Engine Yard's own stack and has more than 400 customers.
"On Wednesday, we are allowing our customers to choose to use [the] Engine Yard Rails stack and a new fully automated deployment model based on AWS resources," Mornini said.
"[Amazon has] a lot more scale than we do and can provide more resources at a lower cost point than we can internally," he said. But customers still can leverage the Engine Yard stack for Ruby on Rails deployments, Mornini explained.
Engine Yard's Ruby on Rails capabilities include Web site functionality around application deployment as well as support. Featured is an application-centric viewpoint that accommodates dependencies on other applications, such as PHP applications.
While Engine Yard offer customers perhaps 100 virtual machines, Amazon can offer thousands, Mornini said.
Solo will be available on the Amazon Elastic Compute Cloud (EC2) on January 28 for $129 pre month as the starting price. Plans call for eventually extending the offering to other cloud platforms also.