Users of Amazon Web Services will soon be able to orchestrate workflows across different AWS services and their own internal resources, using a new orchestration engine called the AWS Data Pipeline.
Amazon Chief Technology Officer Werner Vogels introduced the technology at the company's Re:Invent conference, held this week in Las Vegas. The service is now available in limited beta preview, though Vogels did not say when it would be commercially available, nor what the price would be.
The service can "automate the movement and processing of any amount of data using data-driven workflows and built-in dependency checking," according to a blog post AWS issued that further explained the technology.
Amazon designed the service to automate the process of parsing large sets of data. For example, one pipeline can move log data from an AWS EC2 (Elastic Cloud Compute) instance to the AWS S3 (Simple Storage Service) once a day, and then, once a week, evoke an analysis job on the data on an AWS Elastic MapReduce cluster.
To set up a workflow pipeline, the user identifies some data sources and describes the steps that AWS should take to process the data. The user would also identify the destination for the processed data as well as a schedule for when the pipeline should be executed. Preconditions can also be established that the service will check before executing a job, such as checking if a file that is needed for the operation exists.
Pipelines can run across EC2, Elastic MapReduce clusters, and the user's own hardware. Pipelines can be set up in the AWS Management Console or by writing a script.
This is not the first workflow engine on AWS. The company also launched the Amazon Simple Workflow in February. However, AWS Data Pipeline is more focused on executing data-driven jobs.
The AWS Data Pipeline is one of a number of announcements Amazon made at the conference. The company also unveiled a data warehouse service and an auto-discovery service to ease the management of its ElastiCache. It has also cut the prices of some of its storage services and created two new EC2 instance types, for high-memory usage and large data usage.