We’ve encountered a problem with the File manager, specifically when trying to download files larger than 8 GiB.
A client of ours pointed out that a backup file with a size slighty over 8 GiB, was truncated to (size – 8 GiB). So of the actual 8,23 GiB, when clicking on the file in the File Manager, the resulting downloader window listed a filesize of 234 MiB.
After verifying the downloaded file, it is actually only 234 MiB that’s downloaded, and the file is naturally corrupt and unusable.
We’ve replicated this behaviour by generating a zero-filled file on a random location on one of our DA servers, and downloading it from within the File Manager. In every instance, the resulting filesize is a multiple of 8 GiB less than the actual filesize.
For instance, a file generated with an actual size of 16,5 GiB, will be downloaded as approx. 500 MB, missing the first 2 blocks of 8 GiB.
This behaviour only occurs with downloads from within the File Manager interface, FTP’ed transfers are complete and functional.
I would very much like to know where this limit has its origin, and if there’s a way to circumvent this, other than advising to use other means of download such as FTP.