Backround short SMART failed - But why?

About a week ago my secondary data cluster (12 x 8TB SAS RAIDZ2) finished a resilver, and then the night after, a disk failed showing read and write errors.

Easy enough, I pop in a new drive and it starts the 7 day resilvering process. About a day into it, I get a notification that a completely different disk, da9, has failed its SMART test. I don’t touch it, as its in the process of resilvering

Now the resilver has finished, 0 read/write/checksum errors across the array, but da9 is still failing its daily SMART check

All of these drives are old used 8TB SAS disks, and from past experience running them in a Windows box, there is a SMART attribute which isn’t well explained and is essentially “old age” and can be ignored

Why is it that as good as TrueNAS is, it can’t just show me the SMART output? Its telling me its failed, but won’t tell me what its seeing. Is there a way to get this data? I don’t want to swap a perfectly healthy drive and go through another 7 day resilver

Are you guys listening to SMART? Part of me wants to just turn it off. If there are no read/write errors then what do I care?

I’m running CORE 13.0-U6.1

Check it in shell.
smartctl -a /dev/sdx

smartctl -a /dev/da9 on CORE of course :wink:

1 Like

Especially with SAS drives I would recommend smartctl -x /dev/da9 since SAS drives technically do not use SMART and instead have their own ‘standard’; -x should get you all availably info.

1 Like