Failed S.M.A.R.T Test Detail?

[Alkaid_c]

Hi.

I built a system several days ago with 5 brand new hard drives (and two SSD cache drives). Yesterday I found a warning sign under the storage tab saying that there are failed S.M.A.R.T. tests.

I clicked view all tests, and found that all five drives failed once.

I am sure I did not schedule such a test. Does the system do it automatically? And it is so weird that all new drives failed at the same time. These drives belong to 2 different models and were purchased in 3 different places, and the system is attached to a UPS. So I did following things and have several questions:

First, I manually ran a long test. All drive passed. So I do suspect that the failed test might be caused by some arbitrary reason - for example, I rebooted the system the day before, and it may interrupt a test. My questions are: (1) if all drive passes the test, does it mean that my drives are okay and the pervious failure is ignorable? (2) if they are indeed ignorable, how to eliminate the orange warning sign so I won't ignore the warning sign when a real thing happens in the future?

And I went to shell and executed smartctl command to gather more detail about the failed test. However, it reports that my drives do not support SMART and reports zero temperature. However the other part of the GUI tells me my drive supports SMART and there is no reason they don't support it. Why does this situation happen?

Thank you guys very much,

 

[danb35]

What happens if you run the smartctl command without -d scsi? And what model are the disks? From the serial numbers, sda looks like it's Western Digital, which should support SMART.

 

[Alkaid_c]

What happens if you run the smartctl command without -d scsi? And what model are the disks? From the serial numbers, sda looks like it's Western Digital, which should support SMART.

Ah, I see. I used the wrong command. I did get the smart info after removing -d scsi. And the failed test is indeed "interrupted." Does it really mean "being interrupted (by my reboot)"?

 

[Ericloewe]

As it says, "host reset" and doesn't actually report any failures.

Also, please copy/paste console output into a CODE block instead of taking a screenshot. It's easier to read, consumes less storage, less bandwidth, it's easier to render and it looks better.

 

[danb35]

And the failed test is indeed "interrupted." Does it really mean "being interrupted (by my reboot)"?

Yes, that's what it means. This is a bug in 22.12.1, that interrupted tests are being reported in the GUI as failed.