Deduplication: No doubt this is efficient and cost effective but I'm not sure if its really implemented by the popular cloud storage groups in this manner. By sure, no means of a scientific test, but just last week I uploaded a legitimately owned copy of Windows 10 downloaded from Microsoft into my Dropbox. The file wasn't renamed or changed in any way from the one that was downloaded. I think it is a reasonable assumption that I'm not the first to upload that file. The upload took exactly the length of time one would expect with my given Internet speed which would seem to indicate that deduplication was not occurring--at least to mitigate lengthy uploads.
That doesn't mean a deduplication wasn't occurring afterwards or as part of an ongoing scheduled maintenance/clean up. However, the size of the file database for all Dropbox users would be humungous and contain significant unique files (think of all the pictures and videos uploaded automatically by millions of smartphone users with the Dropbox app configured to do so).
Where I think the deduplication is most likely used is if I were to share my ISO with you on Dropbox and you choose the option to "copy to my Dropbox" (instead of just downloading it to your computer from my share). At this point, the very nature of "copying" the file tells the host "duplicate this file to user x." It really just needs a pointer to the original file. If you were to modify your copy of the file, a differential copy will take place (similar to VMware/Parallels snapshots). Easy to determine and easy to implement.
All that is well and good but not likely to be relevant to a Plex user uploading a movie into their own drive that's unshared with the public.
That doesn't mean a deduplication wasn't occurring afterwards or as part of an ongoing scheduled maintenance/clean up. However, the size of the file database for all Dropbox users would be humungous and contain significant unique files (think of all the pictures and videos uploaded automatically by millions of smartphone users with the Dropbox app configured to do so).
Where I think the deduplication is most likely used is if I were to share my ISO with you on Dropbox and you choose the option to "copy to my Dropbox" (instead of just downloading it to your computer from my share). At this point, the very nature of "copying" the file tells the host "duplicate this file to user x." It really just needs a pointer to the original file. If you were to modify your copy of the file, a differential copy will take place (similar to VMware/Parallels snapshots). Easy to determine and easy to implement.
All that is well and good but not likely to be relevant to a Plex user uploading a movie into their own drive that's unshared with the public.
Last edited:
