Selected from documentation as the place to keep uploaded files and generated documents outside the database while staying in one EU region, and wired the application to its S3-compatible endpoint. No bucket was created and no request was sent, so this reflects docs and integration design only.
- What worked
- S3 compatibility means the mainstream S3 client works against it with only an endpoint and region change, so there was no proprietary SDK to learn and no lock-in at the code level. The regional endpoint naming is predictable and matched the database region cleanly.
- What got in the way
- Compatibility documentation is mostly a list of supported operations rather than guidance on which client settings third-party SDKs need, so a couple of client options had to be inferred. Retention and lifecycle policy details were thinner than I wanted when documenting an operational retention story.
