DEFAULT SMART Scheduling

[dkusek]

Preface.... This may be a dumb question....

Ok, I have used 9.2.1.5, 9.2.1.8, 9.3, and am now running the most current version of 9.10.1-U4. My hardware is beyond recommended specs. I have looked for a few days now for this exact question and cant seem to find a specific answer (again, maybe because its one of those DUH questions...).

What I would like to know is if I don't actually schedule SMART test, but the service is on and my disks support it, when do these test run?

Some notes... When I have a pool added and S.M.A.R.T. services are on, I can go into /usr/local/etc/smartd.conf and get something like this:
/dev/da0 -a -n never -W 0,0,0 -m root -M exec /usr/local/www/freenasUI/tools/smart_alert.py

Now, that is all I have really ever done. I cant seem to find where the default settings are located for when these run EVEN THOUGH I have not scheduled the test in the GUI. If this is impossible and someone can straight out tell me SMART does not run unless it is scheduled, I will sleep soundly. But for some reason, I have gotten SMART notifications via email from time to time with ZERO WebGUI scheduled tasks. I can even run smartctl -a /dev/da1 (for example) and get output that show short and long tests have been run. I have been combing through the filesystem and can't find a clear cut answer to if there is a default (again without scheduling) or not.

Any help is appreciated.

Thank you,
-Devin

 

[Jailer]

Smart test are not cofigured by default, you have to set them up. The documentation explains exactly how to do this.

 

[dkusek]

Understood about the configuration. I am purely asking specifically a scheduling question. Maybe I am missing the terminology somewhere...
Here is an example on a system that I did a fresh install about 6 hours ago (see image New Install 9.3)

And here is a better example of what Im talking about with a system that has been running for a longer period of time that I have updated and is currently on 9.10.1-U4 (see image System 2....).

As you can see, both systems have run either a background long or short test but NEITHER are actually Scheduled (no SMART tests in Tasks)...

One thought is that the disks, which are SMART capable, are consistently gathering SMART data and when there is a fault or something like that, SMART actually projects the pre-gathered data?

I am pretty deep in this rabbit hole and may be missing something here...

Again, thanks to anyone for guidance here.

 

[joeschmuck]

What I would like to know is if I don't actually schedule SMART test, but the service is on and my disks support it, when do these test run?

All this means is the OS will check the SMART data, basically the SMART Health Status. This is self-run basic testing while the drive runs such as checking the drive electronics and then any write/read failures. This is not conveyance, short, or long testing. You should read up on what SMART is to fully understand it. Basically if you wait for a SMART reported failure without running routine SMART tests then you can expect the drive to warn you of a catestrophic failure that has already happened or will happen in 24 hours or less.

SMART short and long tests run in the background by nature but must be commanded to run, they will not run automatically and if you find one that actually does run automatically, I'd like to know about it because that would be something worth investigating.

When you post data like SMART outputs, please post all of the data, it feels like we are missing quite a bit here. I'm not sure why you took an image vice taking the text abnd pasting it. You obviously have a failing drive in System 2 9.10-1U4.

But for some reason, I have gotten SMART notifications via email from time to time with ZERO WebGUI scheduled tasks.

That is becasue your drive has failed.

So I see you having two problems right now:

1. You have a hard drive failure which you need to identify and replace.
2. You need to setup routine SMART Short and Long tests.

If you need help with the failing drive, in my signature is a guide I made to help people with hard drive troubleshooting.

I'd recommend you setup daily short tests and weekly long tests.

Good Luck!

 

[dkusek]

THANK YOU!

This is exactly the information that I was looking for. We are/have been aware of the failed drive. That system with the failed drive has test information on it so not a big deal haha. Your help gives me a better understanding of what is happening behind the scenes on a drive specific basis in terms of SMART activity.

Again, thank you for the insight on SMART and why we get reports without having schedules in place. Truly grateful for the recommendations and will enforce a short/long testing policy with our systems.

Thanks again!

-Devin