SMART Tests all stuck at 10% for several days

[jsylvia007]

I can't seem to catch a freaking break.

I've upgraded all my hardware to enterprise class hardware to do things the right way...

I purchased NEW drives and ran WDIDLE3 /D to disable the Idle3 timer on the WD Green drives I purchased.

Now, all my drives have been locked trying to finish a smart self test for days now. As we say here in the Navy, Whiskey-Tango-FUCK over??? What the hell am I doing wrong?

Anyone have any idea why the SMART tests are hung with 10% remaining? I'm waiting on a drive to resilver, and as soon as it does, I will be rebooting the NAS and seeing what happens.

I'm concerned, because if the SMART tests won't finish, I'm pretty sure that I won't have any visibility should a drive start to fail if they keep hanging.

 

[cyberjock]

You gave it away... you shouldn't do more than any one of the following at any time:

1. Resilver
2. Scrub
3. Run SMART tests

You're running a SMART and resilver. ;)

 

[jsylvia007]

Ok, excellent... I guess the issue was that the automated SMART tests just happened to start during the resilvering process. I canceled all of the tests with -X and will rerun after the resilvering has completed.

Follow-on question - If this should happen automatically in the future (say, during an automatic scrub), will the tests eventually finish once the scrub has completed, or do I need to come up with some complex schedule to be certain that a scrub never happens at a time that automated smart tests might run?

Also, any specific reason why you can't run smart tests while the other stuff is running? Does ZFS grab exclusive access to the drive or something?

 

[cyberjock]

Follow-on question - If this should happen automatically in the future (say, during an automatic scrub), will the tests eventually finish once the scrub has completed, or do I need to come up with some complex schedule to be certain that a scrub never happens at a time that automated smart tests might run?

It's your job as the admin to make sure that never happens. A complex schedule awaits you!

Also, any specific reason why you can't run smart tests while the other stuff is running? Does ZFS grab exclusive access to the drive or something?

Specific reason? It doesn't work. Duh!

But seriously, nobody really knows why it doesn't work. There's no reason technically why it shouldn't except that disk usage might be so high it interferes with the test. I really don't know. It's easier to just tell people not to do that. ;)

 

[jsylvia007]

Once again, thanks cyberjock!!

Do you EVER sleep??!

 

[cyberjock]

Yes.. I took a nap for 7 hours this afternoon. My sleep schedule is somewhat backwards right now, so I'll be going to bed early so I can get back to being awake during the day.

 

[jsylvia007]

Hahaa!! Good luck, and thanks again.