S3 Async Client PutObject results empty Content-Encoding

[trung]

Hi,

I'm using S3 Async Client to perform putObject with stream body.

s3AsyncClient.putObject(
            PutObjectRequest.builder()
                .bucket(fileUploadBucket)
                .key(objectPath)
                .contentLength(contentLength)
                .contentType(MediaType.APPLICATION_PDF_VALUE)
                .build(),
            AsyncRequestBody.fromPublisher(body))

The file is uploaded successfully in S3, if I download it from S3, the content looks good.
However, I noticed that the metadata section of the file in S3 has empty Content-Encoding

This causes problem when using Cloudfront signed URL, the file content is not viewable (varies by browsers and versions).

if I take away the .contentLength in the PutObjectRequest. Content-Encoding is not in the metadata and Cloudfront signed URL works. However, it's no longer a streaming upload as the file content is now buffered internally.

Interestingly, I can't even set the Content-Encoding to be an empty value using the AWS Console.

I can set Content-Encoding with some random value in PutObjectRequest so to force this value to be available in S3 object metadata instead of empty, but it doesn't sound right.

IMO, this is nothing to do with SDK, just that S3 server doesn't do the right thing when setting the Content-Encoding to empty.

Appreciate any suggestion.

[trung]

For those who are searching for answers. It turns out to be the CRT-based S3 Async Client is the culprit. I swapped with the default one then it works. Content-Encoding is no longer populated with empty value

[trung]

Opened an issue with CRT: awslabs/aws-crt-java#729

[github-actions[bot]]

Hello! Reopening this discussion to make it searchable.