ADVSUS04-BP01 Use batch processing for data cleansing and enrichment to create customer profiles
Use batch processing for data cleansing and customer profile enrichment in advertising workloads. Schedule the batch jobs during periods of lowest carbon consumption to minimize resource usage and environmental impact.
Implementation guidance
-
For workloads like privacy-enhanced data collaboration that involve data cleansing, enrichment, and customer profile creation, implement batch processing architectures to minimize resource usage.
-
Use AWS services like AWS Batch
and AWS Step Functions to queue up and schedule these batch jobs during periods when the carbon intensity is lower, such as times when more renewable energy is available or when demand is lower. -
Consider using AWS Graviton
-based instances if supported, for batch processing workloads, if as they offer energy-efficient compute capabilities. -
Sample data sets when possible, to reduce compute, analytics, and data transfer needs.
Key AWS services
-
AWS Instance Scheduler
(for scheduling batch jobs during low-carbon periods)
