Upgrade from BETA2 to RC1 stuck on ZFS Pool import

pycvalade

Cadet
Joined
Oct 27, 2021
Messages
9
Hi guys,

I've been using TrueNAS Scale since BETA1 in a Proxmox VM + PCIe passthrough with no problems. I tried to upgrade the system from BETA2 to RC1 today and got stuck on ZFS Pools import at boot and found it's a problem coming back every few years on this forum.

Hitting Ctrl+Alt+Del reboots the system and choosing BETA2 in the boot menu brings the instance back up.

Would you guys advise that I:
1. just wait a few more weeks and try again
2. unplug my main storage pool from the VM, try booting from RC1 and if it works then reboot and manually import my pool back
3. unplug my main storage pool from the VM and fresh install + config import

Thanks!
 

morganL

Captain Morgan
Administrator
Moderator
iXsystems
Joined
Mar 10, 2018
Messages
2,691
So basically, there's no solution yet to this.

I have no problem being a ginea pig if someone requires one to fix the issue btw if someone reads this!
Thanks... do you also have SMB shares with the extended attribute feature mentioned in the other thread?
 

awil95

Dabbler
Joined
Apr 23, 2017
Messages
28
I am running into the same issue with trying to upgrade my 21.08 B2 server that I built on October 16th. Somehow my pools all show the attribute as enabled.
Code:
zpool get all Atlas | grep xattr_compat
     Atlas  feature@xattr_compat           enabled          local
zpool get all Cronos | grep xattr_compat
     Cronos  feature@xattr_compat           enabled         local
zpool get all Hercules | grep xattr_compat
     Hercules  feature@xattr_compat           enabled           local

Is there anyway around this issue or must I do a fresh install of 22.02 RC1 and just create new pools from scratch?
 

rmr

Dabbler
Joined
Sep 8, 2021
Messages
17
It looks to me like "enabled" on xattr_compat is not the root cause (as long as it doesn't read "active"). The problem seems to happen when you configure a pool for "Apps" (or by extension when you're upgrading from a system that has the apps pool configured). Upgrading isn't the only way to break the system, it (also) happens to me on a fresh test install with a new, single, very small and very fast pool as soon as I set the "Apps" pool. I'd recommend waiting for a few days (continue to use 21.08-BETA.2). You're not the only one affected, I'm sure they will figure it out.

EDIT: This seems to happen when the selected "Apps" pool is encrypted. If the "Apps" pool is unencrypted, it works.

I am running into the same issue with trying to upgrade my 21.08 B2 server that I built on October 16th. Somehow my pools all show the attribute as enabled.
Code:
zpool get all Atlas | grep xattr_compat
     Atlas  feature@xattr_compat           enabled          local
zpool get all Cronos | grep xattr_compat
     Cronos  feature@xattr_compat           enabled         local
zpool get all Hercules | grep xattr_compat
     Hercules  feature@xattr_compat           enabled           local

Is there anyway around this issue or must I do a fresh install of 22.02 RC1 and just create new pools from scratch?
 
Last edited:

rmr

Dabbler
Joined
Sep 8, 2021
Messages
17
It looks to me like "enabled" on xattr_compat is not the root cause (as long as it doesn't read "active"). The problem seems to happen when you configure a pool for "Apps" (or by extension when you're upgrading from a system that has the apps pool configured). Upgrading isn't the only way to break the system, it (also) happens to me on a fresh test install with a new, single, very small and very fast pool as soon as I set the "Apps" pool. I'd recommend waiting for a few days (continue to use 21.08-BETA.2). You're not the only one affected, I'm sure they will figure it out.

EDIT: This seems to happen when the selected "Apps" pool is encrypted. If the "Apps" pool is unencrypted, it works.
I manually applied 22.02-RC.2 PR: https://github.com/truenas/middleware/pull/7796 and it fixes the issue for me.
 

pycvalade

Cadet
Joined
Oct 27, 2021
Messages
9
Thanks... do you also have SMB shares with the extended attribute feature mentioned in the other thread?
I only use ACLs for TimeMachine. Pool is encrypted though.

As others said, I'll wait a few weeks and see if it helps!
 

morganL

Captain Morgan
Administrator
Moderator
iXsystems
Joined
Mar 10, 2018
Messages
2,691
I only use ACLs for TimeMachine. Pool is encrypted though.

As others said, I'll wait a few weeks and see if it helps!
The pull request quotes by @rmr does seems to fix a dataset encryption bug. Feel free to use.

We are looking at an RC1.1 to resolve the major issues like this that we have found... its a discussion topic for next week and we'll confirm the plans after those discussions.
 

pycvalade

Cadet
Joined
Oct 27, 2021
Messages
9
Got it.

Meanwhile, is there a way to update the UEFI bootloader without messing stuff up so that the default boot item gets reverted to BETA.2?
 

malco_2001

Dabbler
Joined
Sep 10, 2013
Messages
20

morganL

Captain Morgan
Administrator
Moderator
iXsystems
Joined
Mar 10, 2018
Messages
2,691

Kasazn

Explorer
Joined
Apr 17, 2021
Messages
60
There's an update here. https://www.truenas.com/community/threads/scale-22-02-rc-1-feature-flags-and-import.96461/

In short there are two separate issues:
1. xattr_compat feature flag - if this is "active" the pool needs to be backed-up and recreated (described in release notes)
2. there is a pool import issues with encrypted datasets and SMB -- that will be fixed in an RC1.1 planned for this week. Described in this bug report: https://jira.ixsystems.com/browse/NAS-113052

If I am understanding correctly, those with issues on the xattr_compat will not be able to save their data and will have to be rebuild? If so I will wait for RC1.1 to come and recreate a new pool then.
 

c77dk

Patron
Joined
Nov 27, 2019
Messages
467
2. there is a pool import issues with encrypted datasets and SMB -- that will be fixed in an RC1.1 planned for this week. Described in this bug report: https://jira.ixsystems.com/browse/NAS-113052
Can't wait - for some reason my RC1 chose to reboot last night and got stuck (I was making it work hard, but have to dig deeper when it's online again). So for now it's turned off and it's getting cold in the room :P
 

pycvalade

Cadet
Joined
Oct 27, 2021
Messages
9
Just a simple update on this: I performed the RC1-1 update 15 minutes ago on my BETA2 install and it booted flawlessly. No problem at all with anything!

Btw, I saw that you guys brought back the console menu for the RC, nice!
 
Top