How To Replace Disk in Mirrored Boot Pool

[zenon1823]

I feel like I must have missed this in the documentation so feel free to give me a black and blue eye for this question.
I'm running SCALE 22.12.0 on a Dell R210ii with a pair of 240GB SSD's on the Motherboard's Sata ports.

Last week one of those drives show showed write errors and was labelled as faulted and the pool degraded. I rebooted the server and I was surprised when the error count went back to 0 and the pool was healthy again. Then after a few days more write errors showed up and after the system hung for 20mins or so it reported the drive as Faulted and the pool degraded.

I see the button to replace the drive in the boot pool, but no option to offline the disk like for that storage pools. So how am I supposed to swap this drive? Just a few mins ago i shut down the server and swapped the drive i went to boot it up expecting the pool to be degraded missing a disk with the new disk available to in the replace menu. But instead the system failed to boot with a no bootable device error. I put the old "faulted" drive back in and rebooted and thankfully the system came back up and again the error counter is zero and the pool "healthy" again.

I know this drive is faulted. how do i got about replacing it?

---
TrueNAS-SCALE-22.12.0
Dell R210ii w/ Xeon E3-1220v2 & 12GB DDR3 ECC Udimm
Dell H200e (sas2008) flashed to IT Mode
Dell MD1200 Disk Chassis
2x 240GB ssd - mirrored boot pool (internal server sata controller)
4x 8TB Ironwolf - mirrored vdevs data pool (disk shelf)

 

[winnielinnie]

I see the button to replace the drive in the boot pool, but no option to offline the disk like for that storage pools. So how am I supposed to swap this drive?

On Core the procedure through the GUI is either:

• "Replace" (preferable)
• "Detach" -> "Attach"

The first method is preferable.

I'd assume it's the same terminology and process with SCALE.



The wording is different between the Storage Pools page and the "Boot" menu. (Keep in mind that the GUI is not a word-for-word overlay of zpool and zfs commands. The GUI obfuscates certain things.)

But instead the system failed to boot with a no bootable device error.

Your BIOS might be configured to boot from a particular device/port, without falling back to an alternative device. (Check your BIOS settings.) That's why you need to have the "failing" drive plugged back in.

 

[zenon1823]

On Core the procedure through the GUI is either:

• "Replace" (preferable)
• "Detach" -> "Attach"

The first method is preferable.

The wording is different between the Storage Pools page and the "Boot" menu. (Keep in mind that the GUI is not a word-for-word overlay of zpool and zfs commands. The GUI obfuscates certain things.)

I'd assume it's the same terminology and process with SCALE.

I assumed the same, but i cant find any detach or offline buttons in the gui for the boot pool. They exist in the gui for the regular data pool, unless they are just "hiding" in front of my face for the boot pool...

As i don't have any more bays available to do a replace. My only option will be to power the new ssd from a external power adapter and attach sata to the motherboard. Do the replace option then physically swap the drive. That should work right?

Your BIOS might be configured to boot from a particular device/port, without falling back to an alternative device. (Check your BIOS settings.) That's why you need to have the "failing" drive plugged back in.

sorry let me elaborate (i should have been more specific) - whats odd is that it listed 4 truenas options that failed to boot - it was something like truenas0, truenas1, truenas0 & truenas1 (i assume my boot environments, there are 2, on each drive), and then that no bootable device was found. So it wasnt just the generic no boot device found message. It looks like GRUB failed to boot off the single drive. I'll try and get a pic of it, if I end up booting it in that scenario again.

 

[sretalla]

System Settings | Boot | Click the button "Boot Pool Status"... then you have the 3 dots on the right of each element (mirror and disks).

 

[zenon1823]

System Settings | Boot | Click the button "Boot Pool Status"... then you have the 3 dots on the right of each element (mirror and disks).

The only option I see there is to replace, not to offline or detach a disk?

 

[sretalla]

The only option I see there is to replace, not to offline or detach a disk?

Interesting...

I see the option to replace and attach with the single boot device I have (so I didn't see that the Mirror item has no 3 dots option as I was expecting it to have).

Looks like a bug to me, so feel free to submit a bug report.

 

[winnielinnie]

I see the option to replace and attach with the single boot device I have (so I didn't see that the Mirror item has no 3 dots option as I was expecting it to have).

We're on Core, while @zenon1823 is on SCALE. Maybe they want to simplify the GUI for SCALE users. "Fewer options is better! Just ask Apple!"

 

[sretalla]

We're on Core, while @zenon1823 is on SCALE.

I had noted it and was comparing my SCALE systems, but don't have mirrored boot on any of them.

 

[zenon1823]

Interesting...

I see the option to replace and attach with the single boot device I have (so I didn't see that the Mirror item has no 3 dots option as I was expecting it to have).

Looks like a bug to me, so feel free to submit a bug report.

ok will do. Glad it wasn't just me being blind.

Can you advise if this is a suitable workaround?

I don't have any drive bays available to install and new ssd to use for a replace operation. But i could power the new ssd from a external power adapter and attach it to an open sata port on the motherboard. Do the replace option then physically swap the drives. That should work right?

 

[Michael Kepler]

Replacing a faulted drive in your mirrored boot pool on SCALE 22.12.0 is straightforward, despite the confusing behavior you've encountered. Skip pre-emptively shutting down or swapping drives; instead, directly use the built-in "Replace" option within the TrueNAS GUI:

1. Navigate to System Settings > Boot > Boot Pool Status.
2. Locate the faulted drive and click its three dots menu.
3. Select "Replace" and choose the new drive.
4. Wait for the replacement process to complete, which can take several minutes.

The pool will remain online throughout the process, minimizing downtime. This method automatically handles offline procedures and integrates the new drive seamlessly. Don't worry about the occasional error counter resets; focus on replacing the drive using the designated "Replace" option for a smooth transition.

Remember, a healthy pool shouldn't have faulted drives, so keep an eye out for future errors and repeat the replacement process when necessary. Good luck!

 

[SavanFlou]

I am currently in this situation, but with a very annoying complication.

So I installed Truenas Scale, something like a year and half ago. Following some youtube videos, I installed Truenas on a SSD and a flashdrive.
Recently, this key started to be presented as "degraded". And yesterday, it is now "unavailable".

So now, I can access the Truenas, all my different datasets, but all the apps disappeared, which is... not good let's say :).
As I cannot simply remove, I tried to replace it.

So I plugged another flashdrive, chose "replace", then my flashdrive, the process started... and then it seems like the webpage got refreshed, I have merely the time to see the connection page that I am back to the "boot pool" status page, but the "boot replace" work has been stopped.

So I am currently stuck, part of my system has disappeared but I am not able to fix it by the simple way.
So is it possible to do it by another way?

Thank you for your help.

System= TrueNAS-SCALE-22.12.4.2

 

[SavanFlou]

It seems I cannot edit :
So... I need to take some holidays.
During the week-end, I put for some reason an automatic refresh on this page. I don't understand why that affects Truenas, but problem solved. For this, I started to replace by another flashdrive, then I will buy and put another SSD.