Ensure that the Region used for training and developing models is the same as the one used for data. This approach helps minimize the time and cost of transferring data to the computation environment.
Implementation plan
-
Keep data and compute resources in close proximity - Amazon EC2 is hosted in multiple locations world-wide. These locations are composed of Regions, Availability Zones, Local Zones, AWS Outposts, and Wavelength Zones. Each Region is a separate geographic area. If you are launching a compute cluster, you should launch the cluster in close proximity to your data to get the best performance. Say, your Amazon S3 bucket is in the US West (Oregon) Region, you should launch your cluster in the US West (Oregon) Region to avoid Cross-Region data transfer fees.
Documents
Blogs
Videos
