[Help] SMART test failed, but drives list as OK

[cwendler]

I ran the long Smart test on my two drives, which they both failed. However, they both report as OK on the Overview tab. I'm guessing I'll need to replace them soon? I was able to create a storage pool (raid 1) as well as a volume on them and I can connect and copy files to them over on Windows. To be clear, the two drives are WD60EFRX, which are not on the compatibility list (the EFRX line is, just not the 6TB version). They were salvaged out of a WD My Book (which failed). Was hoping to repurpose them, but looks like I can't. Sorry if this has been discussed already, I'm brand new to the world of NAS and just wondering how to proceed before I starting copying over TB's of data. Thank you in advance for any input!

F4-424
WD60EFRX (6TB, x2)
WD Red SN700 500GB (Hyper Cache)

[TMzethar]

What version of TOS are you using? Please share the SMART anomaly indicators you see.

[cwendler]

Hello TMzethar. TOS version is 5.1.145. I've exported the SMART data for each drive, here are the links (Google Docs). Not sure what the info means, hopefully you do.


HDD 1
HHD 2

[TMzethar]

The Smart information does not show any obvious signs of anomalies. The detection failure may be due to other reasons and requires further investigation. You can log into the terminal via SSH/Telnet and use the following commands to test your two HDDs.

Code: Select all

smartctl -t long /dev/sda
smartctl -t long /dev/sdb

[cwendler]

Thanks for the tip about smartctl. Here are the results that I got...

sda
sdb

[TMroy]

From the S.M.A.R.T. report, I did not find any critical error. No idea at this moment.

[cwendler]

Ok. I will use them for a while to determine how stable they are. Thank you both for your help!

[Gremlin]

I wonder if 'smartctl' is actually returning a non-zero exit code, but the actual reason is not particularly critical: for example
(from one version of manual page for smartctl at https://linux.die.net/man/8/smartctl )

Bit 4:
We found prefail Attributes <= threshold.

And there do seem to be "pre-fail attributes".
So 'smartctl' might "fail" but the o/s does not consider that significant.