Hey there,

 

Sorry if this is a basic question but i haven't been following the Unraid updates very closely lately so i don't know exactly what functionality is available and how it all works. For a couple years now i have been usingan 8TB disk with the unassigned devices plugin. The disk is mounted and i use Nextcloud to perform backups to it from an external Synology NAS.

 

Now the disk is getting full because the data on the Synology is growing. I'd like to add another disk and pool it together with the 8TB one, so let's say i add a 20TB disk, i want to have a pool of 28TB of backup space available, it should be completely seperate from my main array, and not use my cache drives either. And hopefully be accessible through Nextcloud so i can keep my current backup configuration working. It would be nice if the current backup data on the 8TB disk can be kept as well so i don't have to reupload all that.

 

Is this possible nowadays in Unraid? I know something with pooling was added but i'm not entirely up to date on what it is and how it works. Most of what i can find is about multiple cache pools, but that's not what i want??

 

Thanks.

You can create a new pool with both disks, but can only add the new disk to the other one while keeping the data if it's using btrfs. (zfs also, but would need to be done manually)

3 hours ago, JorgeB said:

You can create a new pool with both disks, but can only add the new disk to the other one while keeping the data if it's using btrfs. (zfs also, but would need to be done manually)

 

I'm in luck, it is using btrfs. Is there a guide somewhere on how to do this by any chance?

You just need to add the new device to that pool, by default Unraid will create a mirror, so cancel the balance and then then convert to single profile.

Hi @JorgeB thanks a lot for the help.

 

I think i have managed to do it, this looks good but i didn't actually do anything to convert it to a single profile. However the 16TB free space and about 8TB used seems like it is acting as a single profile, no? Is there anything i should check about this to make sure it is all okay? My existing backup data still seems to be there (The 8TB)

 

 

But how do i actually use this pool now to perform backups to? Previously i had mounted the unassigned 8TB drive which gave me a share of just the drive. I added this share to Nextcloud and performed backups to it like that. How do i do something similar with this pool? And while keeping the existing backup data of course.

 

Edit: I found out there is now a /mnt/nas-backup/ folder. I think i can map this to my Nextcloud container, and then configure it as "local storage" in Nextcloud. Would this be the right way to go about things? Like this (taken from Nextcloud documentation):

Edited June 19, 20242 yr by YourNightmar3

Post the diagnostics, that will show the pool profile in use.

 

You can backup using the pool as target, if it's an external pool, that is not always connected, you could also use UD.

52 minutes ago, JorgeB said:

Post the diagnostics, that will show the pool profile in use.

 

You can backup using the pool as target, if it's an external pool, that is not always connected, you could also use UD.

Alright i have attached the diagnostics.

 

The pool is always connected so i will use the method of mapping it to my Nextcloud container and configuring it in there as "local storage".

 

I can see in the diagnostics for the nas-backup pool it says diskFsProfile = "raid1", is this supposed to say "single"?

If i do "convert to single mode" and press "balance" nothing changes in the diagnostics when i download them again.

tower-diagnostics-20240620-1033.zip

Edited June 20, 20242 yr by YourNightmar3

Profile is OK, metadata and system are raid1, but the data profile is single, that's what you want for this case.

Alright thanks a lot, that makes sense now with what i see in the screenshot as well.

 

One last question, a seperate pool like this does not use your regular array cache, does it?

If you mean using a different pool for the initial writes, no it's not possible, it will be with v7.0

45 minutes ago, JorgeB said:

If you mean using a different pool for the initial writes, no it's not possible, it will be with v7.0

Alright thanks, that is what i meant and i don't want it in this case so my current situation is good

 

Only i'm now having problems where my Nextcloud container can't write to my /mnt/nas-backup/ folder. I mounted it in my Nextcloud like this:

 

It has read access but no write access it seems like. Here is the command ls -la from my unRaid system in case that helps:

 

This is the console of my Nextcloud container:

 

And ls -la from my Nextcloud container:

 

I have limited Linux knowledge so i provided the above information so hopefully you can see if something is wrong, i do not know what most of the command output means. How can i diagnose/fix this?

 

 

 

Can't really help with that, suggest asking for help in the appropriate container support thread:

 

Actually now that i did some more tests it doesn't seem to be Nextcloud related at all. When opening a terminal straight from my unRaid UI i also cannot write to the pool.

 

Anything i can try to check to see why this would be the case?

 

i have attached my diagnostics.tower-diagnostics-20240623-1801.zip

Edited June 23, 20242 yr by YourNightmar3

Pool filesystem appears to be corrupt, reboot and post new diags after array start.

 

P.S. also seeing macvlan call before rebooting change the docker network to ipvlan.

I have restarted the server and here is a new diagnostics:

 

tower-diagnostics-20240624-1526.zip

 

The pool is still readonly. The terminal touch test has the same result.

 

Something odd i noticed is right after starting the array the pool was at 7.87TB Used and 0TB available, with the little bar in red. Then a couple seconds later it suddenly jumped to 7.87TB used and 16.0TB available, which is what i'd expect it to be. That seemed a bit odd to me.

 

Also thanks for the ipvlan tip, i haven't done it just yet because i don't know how (yet) but i'll look into it.

 

 

Edited June 24, 20242 yr by YourNightmar3

The pool is trying to balance to raid1 and failing because there isn't enough space, didn't you intend to use the single profile for that pool?

Yes that is correct, i don't know why it is balancing to raid1, i don't want it to. It should be a single profile.

Hi @JorgeB, or anyone else really

 

Do you still have ideas left for what i can do to diagnose / fix this issue?

Sorry, missed your last post, something/someone started a balance to raid1, with the array stopped, try this:

mkdir /x
mount -t btrfs -o skip_balance /dev/sdg1 /x
btrfs balance cancel /x
umount /x

 

Then start the array and post new diags.

 

 

Hi @JorgeB, no problem.

 

The second command gives an error:

 

This is above my Linux knowledge so all i can really do is post this error and hope you know what to do, i think. The only thing i noticed is that /dev/sdg1 is a 4TB disk in my array, is that correct? Is this supposed to be one of the pool disks?

 

I have attached another diagnostics file in case it's nessecary. tower-diagnostics-20240626-1104.zip

 

 

Edited June 26, 20242 yr by YourNightmar3

42 minutes ago, YourNightmar3 said:

Is this supposed to be one of the pool disks?

It is, I forgot to mention it can change with a reboot, use sdh, so:

 

mount -t btrfs -o skip_balance /dev/sdh1 /x

 

I accidentally closed the terminal window after executing the commands so unfortunately i can't copy paste the output But when doing that one i got a warning about metadata and system being RAID1.

 

 

The pool is not readonly anymore so that seems good 🙂

 

I have attached a new diagnostics tower-diagnostics-20240626-1203.zip

 

Is it all good now?

Edited June 26, 20242 yr by YourNightmar3

15 minutes ago, YourNightmar3 said:

Is it all good now?

Not yet, you need to free up some space to balance the metadata, delete/move about 50G from that pool elsewhere, then wait around 5 minutes and post new diags.

This sounds like voodoo magic I moved a 72G folder off the pool, then waited at least 9 minutes (its been more but i lost track)

 

tower-diagnostics-20240626-1323.zip

Now type:

 

btrfs balance start -dusage=10 /mnt/nas-backup

 

Then post new diags.

 

P.S. this happened because the original device was too full, with COW filesystems you should avoid completely filling them up.