What do you propose that we do, realistically?
We do not have that model of NAS to test against and the problem seems specific to that model of NAS and its configuration.
Have you complained to Synology that, depending on the mode their NAS is in, the speed you can copy to it with two different programs is drastically reduced? "If I was making the slower NAS I would look into it because performance difference is so drastic, 50 vs 100 MB/s."
Below is some more background here, that was in a separate support thread, copied here for the benefit of anyone else who finds this forum thread:
[quote="I"]There's some background reading on tuning network filecopy performance here, about changes in Vista RTM which were subsequently almost all undone in Vista SP1:
blogs.technet.com/b/markrussinov ... 26167.aspx
So even Microsoft, who make the OS, the network drive protocol and the filesystem and network drivers that implement all of it, don't really know what they are doing here, tried to do something more clever, then ended up reverting back to simple code similar to Opus in almost all cases.
From that article, the only documented differences I can see in how Explorer and Opus copy the data is whether or not the client-side enables caching when copying to/from a network server, and whether or not SMB2 pipelining is used.
The buffering should only affect client-side memory usage, not overall speed (unless so much memory is used that the client starts paging). [And leaving buffering on should allow the OS and filesystem drivers to make more informed choices about how much data to read/write over the network at once, than what Opus could do by guessing, at least in most situations.]
SMB2 pipelining is not widely supported by NAS devices, and is documented as being useful with "high bandwidth, high latency" networks, which is unlikely to apply to a locally attached NAS.
Beyond that, there are all sorts of things that could be going wrong. Some devices are sensitive to buffer sizes, where one buffer size works great with one device and poorly with another. Some devices work great if the file data is written via a particular API or mode and terribly in any other mode. Perhaps Explorer is using slightly different buffer sizes or APIs to Opus and one works well with your NAS (when not in iSCSI mode) while the other works poorly. Unfortunately, without access to the source-code to Explorer or your NAS, we can only guess what the difference is.[/quote]