aws s3 driven adapter, generates a OOM when reading a file big enough.

[gabheadz]

Is your feature request related to a problem? Please describe.
aws s3 driven adapter, generates a OOM when reading a file big enough. Current implementation calls for returning a InputStream with AsyncResponseTransformer.toBytes which tries to "write all contents to a byte array" according to Aws SDK documentation.

Describe the solution you'd like
Provide a flag or configuration option to choose from the following 3 options available in the AsyncResponseTransformer parameter for method S3AsyncClient.getObject

1. AsyncResponseTransformer.toPublisher
2. AsyncResponseTransformer.toBlockingInputStream
3. AsyncResponseTransformer.toBytes

For reactive projects options 1 or 3 may be used, with 1 as default. And for imperative projects options 2 or 3 may be used, with option 2 as default. Developers should only consider option 3 for reading small files only.

Describe alternatives you've considered
None at the time of writing this issue.

Additional context
N/A

[gabheadz]

Ultimately it would end having only the AsyncResponseTransformer.toPublisher, returning a Flux of ByteBuffer's for reactive projects.

[github-actions]

🎉 This issue has been resolved in version 3.6.1 🎉

The release is available on GitHub release

Your semantic-release bot 📦🚀