UFAL/S3 - The problem to download a big file - it does not show the progress

[milanmajchrak]

The user is "stucked" on the /download page without any download information.

Note:
The S3 direct download must be disabled by updating the s3.download.direct.enabled

[milanmajchrak]

Implementing a custom input stream is risky, and since DSpace uses AWS SDK v1, I haven’t found an appropriate solution.

The chatgpt answer:
You're correct in observing that using AWS SDK v1's TransferManager to download large files can lead to issues such as prolonged waiting times and unresponsive behavior during the download process. This is a known limitation, especially when handling large files, as TransferManager downloads the entire file to a temporary location before making it available, which can cause significant delays and resource consumption.

Recommendation:

To mitigate these issues, consider implementing a custom solution that downloads large files in manageable chunks. This approach involves dividing the file into smaller parts and downloading each part sequentially or in parallel, depending on your system's capabilities. By processing each chunk individually, you can provide immediate feedback to the user and avoid the long waiting times associated with downloading the entire file at once.

7.6.1. and also main

        <dependency>
            <groupId>com.amazonaws</groupId>
            <artifactId>aws-java-sdk-s3</artifactId>
            <version>1.12.261</version>
        </dependency>