Can’t import pools on new system after motherboard burnt on power up

[serfnoob]

Hello community,
Im in a real pickle.
My motherboard ( AsRack D1541D4U-2T8R ) made zappy sounds and burnt electrical smell yesterday as I was powering it on. So I pulled the power straight away. On inspection I can see a fried IC or Capacitor type thing.

TrueNAS Core 13
2 pools.
TANK (2 x 4 disk) onboard LSI
TANK3 (1 x3 disk) onboard sata
Mirrored (SSD boot pool) onboard sata

1: I don't have other hardware that can host that many disks. I do have a tiny ProxMox Box with all 4 sata ports available.

2: Clean install of TrueNAS Core on proxmox VM - TrueNAS > Storage > pools > add, import pools... nothing in the list
If I select new instead it show the 3 disks....

3: Then found an old config from September > uploaded config > reboot > re-configured the network in shell (this was the limit of my CLI) Storage > pools > add. ‘import pools’ had my 2 old pools with big red export/disconnected... still couldn't add my TANK3 pool.
4: Pulled the mirrored SSD pair from the old box, removed all the proxmox M.2s and tried to boot from the SSDs, nothing. just end up in BIOS with no option to select anything to boot from

I did a lot of reading here but found nothing conclusive for a luddite like myself to follow.
I'm afraid of CLI, I can copy and paste but really don't know what any of it does or means. I'm without the language to even understand the basics.

I'm in full freak out mode, not thinking straight. Wife is sick, dog just had surgery... now I feel like i'm in a deep hole I don't have the brains to get out of.

I would love to be able to mount that 3 drive pool... I might get some sleep tonight if I can.

Thank you in advance for any help.

Is red dead?

Nothing in the drop down

System can see the 3 x 10TB drives from TANK 3

 

[Arwen]

We sort of need the output from the command line of;
zpool import
That will tell you if the pool is importable. And if not, it will help with the diagnosing of the problem.
The command will not harm anything. Nor actually import the pool.

For example, my local backup pool on a USB drive looks like this;

Code:

# zpool import
   pool: usbbackup
     id: 9996583267698845780
  state: ONLINE
status: Some supported features are not enabled on the pool.
    (Note that they may be intentionally disabled if the
    'compatibility' property is set.)
 action: The pool can be imported using its name or numeric identifier, though
    some features will not be available without an explicit 'zpool upgrade'.
comment: Last backup taken on 2022/08/28
 config:

    usbbackup   ONLINE
      sda6      ONLINE

 

[serfnoob]

Thanks for the help Arwen,
I made a clean instal on an old SSD, as I can get my boot mirror to mount.

Code:

root@truenas[~]# zpool import
   pool: tank3
     id: 14637536508062001946
  state: UNAVAIL
status: The pool was last accessed by another system.
 action: The pool cannot be imported due to damaged devices or data.
   see: https://openzfs.github.io/openzfs-docs/msg/ZFS-8000-EY
 config:

    tank3                                           UNAVAIL  missing device
      raidz1-0                                      ONLINE
        gptid/91174145-adb4-11ec-96ec-d05099c17b37  ONLINE
        gptid/9135e790-adb4-11ec-96ec-d05099c17b37  ONLINE
        gptid/91285b56-adb4-11ec-96ec-d05099c17b37  ONLINE
    logs   
      gptid/ea401ab7-2810-11ed-98fd-d05099c17b37    UNAVAIL  cannot open
root@truenas[~]# 

It's taken me a while just to figure out how not post the formatted code.Little wins.

Do you need the bot pool results too?

Thank you so much, I imagine it must be frustrating dealing with this issue from over and over again fromm under skilled units like myself.

 

[serfnoob]

tank3 UNAVAIL missing device

AMAZING. That line reminded me that I had m.2's in that box. I took both out of the old case and put the first in and rebooted...
Ran zpool import again but it now said "no pools"... I almost died.But I went to dashboard to see what its thoughts were now. To my amazement it had auto imported, alive again.

Unfortunately in my excitement I managed to roll over the other M.2 with my chair, that one is now ruined.
Will this mean the other pool can never be able to be imported?
I cant remember if the m.2 were for read or write cache, does it matter?

Thanks again, I might be able to get some real sleep today.

 

[spuky]

If you used the m.2 in the same way as with tank3 then you used it as log device which is used for the intend log not really a write cache more a kind of write buffer...

you should be able to import the other pool with the help of the cli

zpool import

has a -m option

Allows a pool to import when there is a missing log device. Recent transactions can be lost because the log device will be discarded.

so you would loose up to 5 seconds of transactions on tank

 

[jgreco]

not really a write cache more a kind of write buffer.

SLOG is neither a write cache nor a write buffer. It is an intent log, used to protect synchronous writes. See

Some insights into SLOG/ZIL with ZFS on FreeNAS

What is the ZIL? POSIX provides a facility for the system or an application to make sure that data requested to be written is actually committed to stable storage: a synchronous write request. Upon completion of a sync write request, the underlying filesystem is supposed to guarantee that a...

www.truenas.com

If you have not specifically purchased an SSD with power loss protection, or an Optane, it is probably doing nothing useful for you other than just suffering excessive wear and tear on its way to an early burnout death. People throw random SSD's into their system thinking "cache! faster!" without noticing the conspicuous warnings all over the place about using a PLP SSD.

 

[serfnoob]

Thank you Spuky.
It will be a while until I can test the big pool. Need to wait for my board to be RMA'd or look at a temporary solution.

and thanks jgreco. yes I'm a People that throw random SSD into their system type noob. I had them lying around... curiosity and a cat.
If I get this up and running again I will not put them back in.

Thanks again,
Will report back when I have had a shot at mounting TANK.

 

[Arwen]

AMAZING. That line reminded me that I had m.2's in that box. I took both out of the old case and put the first in and rebooted...
Ran zpool import again but it now said "no pools"... I almost died.But I went to dashboard to see what its thoughts were now. To my amazement it had auto imported, alive again.

Unfortunately in my excitement I managed to roll over the other M.2 with my chair, that one is now ruined.
Will this mean the other pool can never be able to be imported?
I cant remember if the m.2 were for read or write cache, does it matter?

Thanks again, I might be able to get some real sleep today.

Glad I could help.

As a note, you can remove a SLOG on a pool after it is imported.

We almost need a Newbie / Noob guide to success. Something that says, don't use L2ARC, SLOG, De-Dup, Special Meta-devices, USB, hardware RAID, and other things we see here. After they are no longer Newbies / Noobs, they will then understand what some of those are and when to use / not use them.

 

[HoneyBadger]

zpool import

has a -m option

That's the ticket here, and if the second pool was using the M.2 in a similar manner, it will be importable. If it was using it as L2ARC ("read cache") then it should have auto-imported with the device missing, so I suspect you also had the other M.2 attached to the pool as an SLOG.

Whether or not it actually needs that device is an academic exercise, and we can sort that out after you get it successfully mounted.

 

[spuky]

That's the ticket here, and if the second pool was using the M.2 in a similar manner, it will be importable. If it was using it as L2ARC ("read cache") then it should have auto-imported with the device missing, so I suspect you also had the other M.2 attached to the pool as an SLOG.

Whether or not it actually needs that device is an academic exercise, and we can sort that out after you get it successfully mounted.

He has not tried the other pool since he has not hardware to attach the 8 disks to...

 

[HoneyBadger]

He has not tried the other pool since he has not hardware to attach the 8 disks to...

Well, yes, but I'm simply stating that if the broken M.2 drive was an SLOG, the pool hasn't gone to the big /dev/null in the sky forever, and once hardware is available again, there's a method to import it, with all but the very latest data intact.

But if it'll be awhile for that RMA to be fulfilled, perhaps I'll come back to this thread for the academic discussion on the extra vdevs when I have a full keyboard.

 

[serfnoob]

Finally put together a replacement system that can spin that many drives, 2nd hand parts take a while in australia.
Good news is the big pool mounted, didn’t care about the missing m.2 at all.
Thanks to everyone who was there to help in my freaked out state.

All the best.