When a drive dies or is removed, all data regarding that disk is inaccessible in the UI (except for model and serial number being shown in the alert). In the “Storage → Manage Devices” it shows up with some arbitrary ID as device name and the device state listed as missing.
Since Community Edition does not support SES it’s a massive pain in the rear to find out which drive actually died. We’re using the disk’s description field to record which bay it’s installed in and then find the failed drive by checking the description of all the remaining drives.
So I guess my question is: is there an easier way? Either, while the disk is still alive, somehow find the disk’s ID that would be shown once the drive were missing? Or, once the drive died, recover the Description field (which obviously isn’t lost since it can be recovered by replacing the drive) before actually replacing the drive?
Yes! In the disks panel, enter the serial number of the disks. SN are unique for each device, so you can easily identify each one.
We’re using the disk’s description field to record which bay it’s installed in and then find the failed drive by checking the description of all the remaining drives.
This is good strategy. Optionally, you can label the disks with some stickers, and enter these labels in the description.
Yes! In the disks panel, enter the serial number of the disks. SN are unique for each device, so you can easily identify each one.
Disk panel only shows online disks, and even searching for the serial listed in the Alert won’t bring it up, so I’d have to compare each of the remaining disk’s serials to my physical documentation and hope that the documentation is actually correct. Not really a decent way to go in anything bigger than a 4 bay system.
label the disks with some stickers, and enter these labels in the description
Still won’t help me if the GUI doesn’t show the description of a dead drive. Whatever I put in the description only shows up for online drives. So I’d still have to compare the remaining online drives’ descriptions with the stickers to find the one dead one
This is all I see when a drive is unavailable.
So the question still remains:
a) Where’s the original disk info of the dead disk stored? I know it’s stored somewhere because it can be transferred to a replacement drive via a checkbox in the GUI
b) if A is not easily available, where can I find the IDs (34690xxx in my screenshot) for the disks that are still alive so I can document these as a workaround?
Do either of these commands help you out?
lsblk -bo NAME,LABEL,MAJ:MIN,TRAN,ROTA,ZONED,VENDOR,MODEL,SERIAL,PARTUUID,START,SIZE,PARTTYPENAME
sudo ZPOOL_SCRIPTS_AS_ROOT=1 zpool status -vLtsc lsblk,serial,smartx,smart
What hardware are you running in particular HBA?
lsblk only lists stuff about drives that are still online but nothing about the missing drive
zpool status shows what I assume is the PARTUUID of the missing drive’s data partition, but still nothing about the disk’s Description. I’m rather hesitant to document partition UUIDs since - technically - they could change without me knowing
Currently experiencing this on a HP ProLiant DL380 Gen9 with HP SmartArray 840 controller in passthru mode (ie running as dumb HBA, not as a RAID controller). Not sure how that’s relevant, tho, because the problem existed on every hardware I’ve ever been running TrueNAS on
Using the tools the HBA provides to identify disks and slots can often be very helpful. Unfortunately I’m not familiar with yours but if you were using an LSI HBA for example then tools like sas2ircu and sas3ircu can do everything you’re after including the ability to turn lights on a location in your chassis.
Storcli is another option for newer LSI HBAs.
1 Like
I’d much rather have a hardware independent solution. The info is kept somewhere within TrueNAS, there’s gotta be a way to get to it
