For those of us with more curiosity than brains. . . .
I decided to do a tear-down of my pristine My Cloud 6TB (second generation) just to see what was inside it, and what interesting things I could find out by looking at the hard drive.
First - you have to get the little beastie apart.
To do this, you take a credit-card sized piece of plastic and slip it between the back of the case and the edge of the cover - like this:
Having done this, you grab the left and right sides with your fingernails and use your thumbs to push the white cover forward. It’s a bit tricky, but you should eventually get the top-edge to slide back.
Note: Avoid pushing on the feet! If they come out, they’re darn near impossible to get to stick back in. (Unless you have some double-sided tape.) Go ahead, ask me how I know. . .
Once that is done, you turn the thing over, 
and repeat on the other side.
(Yes, I know, I get a SMACK for my “the other side” image!)
Once you slide the white cover off, you get this:
and this:
If you look carefully at the left-hand side of the second image - showing the PCB - you will note that the entire electronics and HD assembly lifts out from the left.
Once removed, it looks like this.
Hard drive side:
and the electronics side:
You remove the three screws attaching the mainboard to the hard drive and slide it off the SATA connector. Note that there are three spacers that the screws go through to attach the board to the hard drive.
Once that’s done, you have the mainboard available to view.
Top:
. . .and bottom:
Here’s the top of the board from the other edge, (with the activity LED lens removed):
Two things are obvious, and interesting.
- There is a place for a 90 degree header on the left that smells like it should be a serial console header.
- There is a ten-position surface-mount header on the right (labeled “JP1”) that could be either a JTAG on-board programming header, or a place where a five-position dip-switch would be located.
Another interesting find is the metal striping with slotted holes all around sections of the PCB like someone had envisioned metal shielding for the mainboard. Maybe it passed RFI emissions testing without it? Dunno. It is curious though. And no, the shielding, if used, would not have interfered with the hard-drive as there is plenty of space between the PCB and the hard drive when assembled.
I also noticed that there was a lot of flux residue - which I promptly cleaned - left over from manufacturing.
I took the hard-drive and placed it in a SATA drive dock and connected it to a Linux system running Mint 17. Here is what I found.
GParted shows seven partitions.
What I should have done, (and did not do. . .), was to make an image of the first six partitions using “dd” or something like that. This way if - for whatever reason - the hard drive goes to Hell in a Hand-basket, I would be able to recreate the system parts of the disk.
What I did do was to take a look at each-and-every partition to see what was in it. I used “tree” to create a printed “directory tree” listing of every folder and file in the system.
Note: These have been converted from “Unix/Mac” format to “DOS” format.
dev-sdx2_Listing.txt (7.6 KB)
dev-sdx3_Listing.txt (439 Bytes)
dev-sdx4_Listing.txt (9.7 KB)
dev-sdx5_Listing.txt (453 Bytes)
dev-sdx6_Listing.txt (101 Bytes)
dev-sdx7_Listing.txt (4.8 KB)
The second partition, (/dev/sdx2) is the partition where the actual file shares are located, and I have noted within the listing where the file store is located. One interesting thing I noticed there is that the file store is represented twice, once as itself, and once again as file metadata within a hidden folder. Note that I “snipped” the data to keep the file listings from becoming excessively long.
I’ll leave it to the embedded systems geniuses out there to look at the documents and tell the rest of us what it all means.
Enjoy!
Jim (JR)
