I left Badblocks running last night. I estimated about 4 hours to complete the 3 TB based of the first few %. Strangely this AM it was not complete. After 10 hours it was only 60% done. I am concerned that the explanation for the slow down might be that it ran into "bad" or "weak" areas of the disk that were recovered. So far no errors have been identified by BADBLOCKS. The only other idea I had was that perhaps it was a power saving issue, maybe it was asleep?
Ok a few things.
i)
The smartctl -a output you posted shows two interesting things:
a) The current pending is 5, which means there are 5 LBAs which don't pass the read ECC, but which *have not* yet been reallocated. They will be when new data is written to those LBAs. Then they may or may not show up as reallocated blocks. WD firmware is inconsistent in this; I have some drives which behave as expected, others which just quietly remap the LBA. So you have already gone from 4 to 5. Not looking great for that disk.
197 Current_Pending_Sector 0x0032 200 200 000 Old_age Always - 5
b) The Synology was running SMART short test every 24 hours.
SMART Self-test log structure revision number 1
Num Test_Description Status Remaining LifeTime(hours) LBA_of_first_error
# 1 Extended offline Completed: read failure 90% 2739 9437194
# 2 Short offline Completed: read failure 90% 2730 9437194
# 3 Short offline Completed: read failure 90% 2706 9437194
# 4 Short offline Completed without error 00% 2682 -
# 5 Short offline Completed without error 00% 2658 -
# 6 Short offline Completed without error 00% 2634 -
Item #4 and older passed. Items # 1, 2 ,3 show errors (consistent with the Current Pending_Sector). So the NAS correctly told you the disk's SMART show failing. The extended long self test was probably you using the WD tools.
ii) As noted above the WD test and the NAS SMART self tests are non-destructive.
iii) You don't need a partition table to have a filesystem on the disk, if that filesystem uses all the disk. Don't know if Synology does this or not, but it certainly is possible.
iv) By running badblocks you have certainly now have destroyed the filesystem.
v) Assuming a 100MB/s average rate for the WD Red (close enough) it will take at least 8 hours to write and 8 hours to read the disk. Badblocks does this 4 times; so expect this to take 67 hours.
vi) badblocks will write to the "bad" blocks, and should cause the 197 counter to go to zero. I don't know whether the blocks get reallocated and show on Attribute 5 of the SMART data; none of my Red's has thrown an error yet. Certainly most of the 'Greens' I have do not, but all the Blacks I have do.
So good luck, and watch the SMART output, it conatins all the info on the health of the drive.
