There are instances where that can happen, for example:
- If multiple small read requests come in to the disk, and all of those requests fall into the same chunk, and none of them are already in the cache, they will issue separate provider read I/O in parallel (assuming there are available read threads).
- Since the minimum read unit size is 1 MB for drives employing chunk verification, it will look like the same set of data is being downloaded multiple times at once.
Now you may ask, why do this? If you are already downloading that 1 MB unit, just hold off on the other reads and let the first one finish, then give the other ones the data that they need. This is exactly what the chunk cache is for. Its job is to optimize reads (and coordinate with writes) in such a way as to avoid redundancy. It's usually enabled by default, unless you change that when creating a new drive.
So the first thing to check it whether the chunk cache is enabled. That information is shown in the tooltip that comes up when you hover your mouse over the drive size in the UI.
Now, with that said, it has been reported that multiple reads for the same data are still happening when the new minimum download size option is enabled. That's still an open issue and I will look into reproducing and fixing that, if that's the case.