Just a quick note...
I was testing out the SanDisk Connect Wireless Stick today with a datalogger and while I am able to set it up to work, there is a MAJOR inconvenience issue with it. It will *NOT* allow access to the drive while it is connected in USB Mode to read/write data, say on a computer or datalogger.
That means, to be able to actually access the contents of the stick via WiFi, the datalogger has to be turned OFF manually each time. The only convenience here is that I don't have to physically unplug the stick. I can switch off the datalogger, copy over the log files through WiFi, and then switch the datalogger back on. Or whoever is using the machine has to remember to turn on/off the datalogger before/after each cycle.
This is kind of a pain in the arse. I don't know if this is peculiar only to the SanDisk Wireless stick or if most of these wireless memory sticks can only be in one function (USB Mode or Wireless Access) at a time. Presumably this is to avoid corruption of data or read/writing conflicts or getting partial files. I get it. But it would have been nice to be able to gain access to the stick on-the-fly with it always plugged in receiving data, even to view partial progress (as it is all simple text-based files anyways).
It's due to how regular (as in: non-network) filesystems work. Whether the stick is formatted in FAT32, NTFS or exFAT, you cannot guarantee a consistent filesystem view from multiple hosts (or multiple OSs) connected to one block device - at the block device level - at once. Imagine you work with the data on your computer's drive where something (rouge process, virus, etc.) is changing random sectors behind your (and you OSs) back. The only guarantee of filesystem consistency is when there is only one "entity" accessing and modifying the actual data and the metadata (be it FAT, MFT or other) in a strictly controlled and synchronised fashion.
You could argue there's no problem as long as only one host has a read-write access to the media and other(s) only access it in read-only way and that would be true in a sense that no filesystem corruption on the actual media could occur but, even in this scenario, you could not guarantee the integrity of the in-flight data as seen by the read-only host whilst in the process of being modified by the R/W host (i.e. the data logger in your case).
Then again, you could argue that "but there is SMB/CIFS, NFS, etc. where you could have many concurrent users both reading and writing the data". The problem is, the gizmo that is accessing the media (
USB stick datalogger) would have to be aware of that and have all the necessary server software stack implemented.
To sum up: with all my disdain for Western Digital (a SanDisk brand owner) that's not laziness on their part but rather the fundamental limitation of how "classic" disk file systems work.
That said, I'm actually in the process of developing a Raspberry Pi Zero-based product that uses the OTG functionality mentioned by David. Naturally, it won't overcome the limitation I was talking about but will attempt to mitigate it to a degree by implementing features like unmount-update-remount, shadowing the data to the staging area for later auto-sync, etc.
Home
Help
Search
Login
Register
