Pool offline afte several reboots

[fdigrego]

Due service energy issues i have 2/3 reboots, counsecuence pool is offline now. I dont know what to do to get online again.

if i follow this steps to get back disks online as follow

i see this option that offer to wipe actual data

in the next image my disks

and finally, everithing is running on hp proliant ml10 gen 9 on proxmox, as follows the disk conf

hope this can help to get suppert. thanks in advance!

 

[Whattteva]

How are these drives presented to the VM? And what disk controller are you using?

 

[Arwen]

What does the following command show?
zpool import

 

[fdigrego]

Whatteva, in proxmox the drivers and controller are as follow:

Arwen, that command show this:

im newbie n this matter and will apreciate a lot any help (i have many info collected among time in this pc).

I have similar issue days ago, but i can finally get drivers back online (exportin/importing) zpool. you see in my first post, image disks related show "homeNAS(exported)", im not sure what does mean that.

Thanks for you support!

 

[fdigrego]

More data, i see the next alert messages:

and i see this outputs on several commands (zppol statux, import and passing -f param):

 

[NugentS]

Get a UPS?
Not that that helps with your current problem

 

[Arwen]

It appears you lost your pool.

TrueNAS & ZFS require very specific setup for reliability. ZFS was designed to have direct access to the disks. Using any virtualization of the disks, like you did, can cause severe corruption.

ZFS wants it's disk writes to be performed in exactly the order they were sent. Hardware RAID controllers or VM software may not do so, causing a pointer to what ZFS thinks is good data, when it is corrupt data. Generally this out of order writes does no harm. But, on the rare occasion, like with a power fail, that can go out the window.

Note that ZFS was specifically written to handle un-expected power fails. Even thousands, without data loss or pool corruption. (Except of course, any write in progress during the power fail.) However, out of order writes is the killer of pools, when combined with power fails at the same time.

The correct method to virtualize disks, is to pass the entire disk controller that has the disks attached. Then TrueNAS can see and perform work directly on the disks as ZFS wants. See this;

Resource - "Absolutely must virtualize TrueNAS!" ... a guide to not completely losing your data.

[---- 2024/01/16: Still relevant. Virtualization really doesn't change much. Updates made as appropriate. ----] [---- 2018/02/27: This is still as relevant as ever. As PCIe-Passthru has matured, fewer problems are reported. I've updated some...

www.truenas.com

Where you go from here, might be backups, extreme pool recovery techniques or a recovery service.


Others might have more suggestions, or help with the extreme pool recovery methods.

 

[Whattteva]

Whatteva, in proxmox the drivers and controller are as follow:
View attachment 71051

Well then, that's your issue.
You do NOT want to use virtualized or individual disks. Here's how my VM on Proxmox looks like.

Notice that there is only one disk (the boot drive). And that PCI Device at the bottom there is a directly passed through HBA with all the disks, so Proxmox do not even see the disks. You need to read the resource that Arwen posted. Fail to follow those guidelines at your own peril, as you are kinda' finding out the hard way now.

 

[fdigrego]

Thanks for your support, appears I will lean in the hard way. I will fíx it and take in account those steps. Probably this week end I wiĺl wipe disks and start again. Love if any way to save the info stored in those disks. Thanks again.