SOLVED How do I identify which disk is which?

[Shiroi Kage]

Hello,

My disks were migrated once which caused a physical shuffle in my array (15 disks RAID-Z2). Now I do not know which disk corresponds to which disk in the system. I currently have a warning with one disk being offline for uncorrectable sectors. How do I know which disk it is? I have my disks in bays with indicator lights, but they are all powered on. Should I start a scrub/read a massive set of files to see which disk is not being accessed, or is there a more elegant way?

Speaking of errors, what is the correct procedure for replacing a disk? When I physically replace the disk, what do I do to have the system include the new disk in the array?

Thanks in advance.

PS:

Errors as they appear in the web interface:

CRITICAL: Nov. 10, 2018, 12:12 a.m. - Device: /dev/ada7, ATA error count increased from 0 to 15
CRITICAL: Nov. 11, 2018, 5:44 a.m. - Device: /dev/ada7, 792 Currently unreadable (pending) sectors
CRITICAL: Nov. 11, 2018, 5:44 a.m. - Device: /dev/ada7, 93 Offline uncorrectable sectors

It is also worth noting that when looking for the disk in "view disks" under "storage," I can see it just fine. Reporting also shows that it is being accessed and shows a normal graph that is in sync with other disks in the array. When looking at volume status, all disks appear to be online too and the volume is reporting 0 errors.

 

[danb35]

It is also worth noting that when looking for the disk in "view disks" under "storage," I can see it just fine.

In that view, you'll see the disk's serial number. Open the case and find the disk with the matching serial number. Done.

 

[Shiroi Kage]

In that view, you'll see the disk's serial number. Open the case and find the disk with the matching serial number. Done.

So the errors above mean I definitely have to replace the disk, right? What do I have to do after I install the new disk?

 

[danb35]

https://forums.freenas.org/index.ph...bleshooting-guide-all-versions-of-freenas.17/
https://forums.freenas.org/index.php?resources/replacing-a-failed-failing-disk.75/

 

[CraigD]

15 disks RAID-Z2

You have fun now

 

[Shiroi Kage]

https://forums.freenas.org/index.ph...bleshooting-guide-all-versions-of-freenas.17/
https://forums.freenas.org/index.php?resources/replacing-a-failed-failing-disk.75/

Thank you very very much! The SMART log seems to be enough to show the drive's failure. Good thing I have a backup drive. Will see if I can get an RMA of some sort.

Much appreciated.

 

[Shiroi Kage]

You have fun now

I'm trying. What did you mean by this? That it will take me forever to find the offending disk?

 

[sretalla]

What did you mean by this?

I think he's referring to the recommendation of keeping RAIDZ2 VDEVs under 12 disks (best at 8 and under).
The rebuild is likely to take a long time.

 

[CraigD]

I was,15 wide will work until this happens

I do really hope your rebuild goes well

I have a couple of questions:

How long does a scrub take?
How long did the rebuild take?

Have Fun

 

[Shiroi Kage]

I think he's referring to the recommendation of keeping RAIDZ2 VDEVs under 12 disks (best at 8 and under).
The rebuild is likely to take a long time.

I see.

The resliver is already done. It took about 8 or 9 hours with the array working fine. Similarly, I was able to import the array after a boot partition failure and the data was fully accessible within 30 or so minutes, so I have not had any problems with things taking too long yet.

 

[Shiroi Kage]

I was,15 wide will work until this happens

I do really hope your rebuild goes well

I have a couple of questions:

How long does a scrub take?
How long did the rebuild take?

Have Fun

Apologies for not addressing this immediately.

I do not know how long scurbs take if I'm being honest. They finish within a single night as far as I know, but it could be because my is at 36% fullness (built it two years ago when I had much more time and thought I would fill it faster than this). This resliver was done overnight and I was really sleepy by the time it started, but I think it actually took something like 6-7 hours (I think I overestimated how long it took me to get everything set).

Thanks for the link by the way. Interestingly, I already have errors reporting on another disk, so the batch failure is happening to me. At some point, hopefully next year, my situation will change for a temporary few years, meaning my current server will be some sort of offsite backup. At that point I will have a good opportunity to build a new server with smaller arrays of high density disks. I also noticed the performance degradation on my pool.

Speaking of degradation, whatever happened with anyone trying to implement defragmentation outside of the licensed applications? Is it possible to mount the array in a different OS (like a Linux distro or something) and defrag it there using third party software?