How can I migrate between Amazon S3 storage classes without disrupting applications?

Resolution

Before you begin, review the requirements and considerations for replication.

To migrate between Amazon S3 storage classes without disrupting your applications, complete the following steps:

1. Use S3 Batch Replication for the initial data transfer. S3 Batch Replication operates in the background and doesn't affect your application's access to the source data.
2. Configure Cross-Region Replication (CRR) to automatically replicate new objects. CRR makes sure that your applications continue to write to the source bucket without interruption.
3. Update your application to read from the new storage class only after the initial replication is complete and you set up CRR. For more information, see Mount an Amazon S3 bucket as a local file system.

To move data to the Amazon S3 Standard-IA storage class, you can use a lifecycle policy to transition the storage. To transfer data to the Amazon S3 Standard storage class, you can run Amazon EMR S3DistCp on an EMR cluster to copy all objects.

S3 Glacier storage classes

For Amazon S3 Glacier Deep Archive or S3 Glacier Flexible Retrieval storage classes, direct sync isn't possible because file retrieval can take hours. Before you sync, transition the files to the S3 Standard storage class. For more information, see Automatically copy your Amazon Glacier vault archives to an S3 bucket and storage classes.

Related information

Amazon S3 pricing

AWS Pricing Calculator

What is Amazon S3?