Which AWS service solution meets these requirements?
Amazon S3 buckets with lifecycle policies to delete old objects.
Amazon RDS PostgreSQL and a job that deletes rows based on age and file type columns.
Amazon EFS and a scheduled process to delete files based on age and extension.
An EC2 instance store synced on boot from a central Amazon EBS-backed instance.
Explanations:
Amazon S3 is ideal for storing large, durable objects, but it lacks fine-grained support for partial updates or conflict-free modifications. Additionally, S3 is object-based, and partial changes to objects may require re-uploading the entire object.
Amazon RDS PostgreSQL is more suited for structured relational data and may not be ideal for storing large binary objects (such as files) that range from 1KB to 300MB, especially when partial updates are frequent. It also requires managing the lifecycle of the data explicitly.
Amazon EFS (Elastic File System) provides a scalable and shared file system that can be mounted across multiple EC2 instances, allowing for consistent access to data. EFS supports file-level operations like partial changes and conflict resolution. A scheduled process can be used to delete files based on age and extension.
EC2 instance stores are ephemeral and are lost when the instance is stopped or terminated. Using an EC2 instance store does not provide the required persistence and lifecycle management for the data across multiple EC2 instances.
I figure that the answer is:
Amazon EFS and a scheduled process to delete files based on age and extension.