HI Folks

Have just started off with unraid, and thought I was having success.

Created a cache pool (2 x 4tb SSD)

Created a data pool (2 x 8tb hdd)

was in the process of transferring my media over and found the data transfer speeds slow - looking to speed up if possible

found threads talking about removing parity - so i take out my second drive from the data pool.

Errors get thrown, and i attempt to revert to two drives in the data pool after I find further info that says you have to pay the parity price at some time - I decide it may as well be now.

Have to reinstall the docker containers as everything breaks....

Successfully reinstall docker containers, and get plex running.

Data pool, despite my attempts however remains degraded.

Further reading indicates that I may have got the setup VERY wrong.

Before I pull all my data back off, upset Plex, delete everything and try again.

1 - Can anyone confirm that I have royally mucked up the setup - I don't have parity setup or my data HDDs (data array) in the correct section i think.

2 - Can I recover from here without moving my data back off the server to spare HDDs and without advanced skills/wizardry?

At this point I feel that my best option here is to recover my data back off the server onto some spare hdd I have, delete everything, and start again correctly. blergh!

campbellot-diagnostics-20260705-0904.zip

image of array running

image with array stopped

Edited yesterday at 03:47 AM1 day by dolphinspanker
cleaned up formatting

I do not see an array at all in those images. I see a pool named Cache and a pool named Data. Because there is no array, there is no parity drive. Suggest you watch a video or two from Spaceinvaderone or Ibracorp

https://www.youtube.com/@SpaceinvaderOne https://www.youtube.com/@IBRACORP

Digest a little bit of content then decide what works best for your next steps forward.

Could also post your diagnostics for folks to be able to see better what you are working with and offer suggestions.

You created two btrfs raid1 pools.

If you wanted to use the unraid array you should have instead added the disks to the array slots above the pool section when the array is stopped.

Btrfs goes into degraded mode when you pulled a disk, however it was likely in the process of rebalancing itself to single mode before you interrupted it.

Please refer to the docs: https://docs.unraid.net/unraid-os/using-unraid-to/manage-storage/cache-pools/#removing-disks-from-a-pool

To fix this you need to start the array with both disks assigned, let it rebalance back into RAID1 mode.

Once that’s done you can remove ONE of the disks, set the slot to “missing” and start the array. Then it will once again balance into SINGLE mode. It will display degraded until the process is complete. You can check the balance status by clicking on the pool name “data” or opening the console/terminal and typing in btrfs balance status /mnt/data

Edited yesterday at 01:49 AM1 day by MowMdown

I do not see an array at all in those images. I see a pool named Cache and a pool named Data. Because there is no array, there is no parity drive. Suggest you watch a video or two from Spaceinvaderone or Ibracorp

https://www.youtube.com/@SpaceinvaderOne https://www.youtube.com/@IBRACORP

Digest a little bit of content then decide what works best for your next steps forward.

Could also post your diagnostics for folks to be able to see better what you are working with and offer suggestions.

Ive been looking at a few of SpaceINvaderOne videos, its where i found the info that suggested id setup things wrong !

and i thought I had attached my diags above - attached again

campbellot-diagnostics-20260705-0904.zip

AS for rebalancing, i have attempted to do that several times - by forcing it using the rebalance button in the main menus, plus running scrub also.

And i have also let the server sit and run asking nothing from it overnight to let the scheduled rebalance occur.

No dice unfortunatly.

As you suggest, that data pool is in Raid 1 - that explains why i can still access the data even though its listed as degraded.

You created two btrfs raid1 pools.

If you wanted to use the unraid array you should have instead added the disks to the array slots above the pool section when the array is stopped.

I've been trying to understand how the Unraid parity setup is different from a Raid 1 pool.

What I am trying to achieve is fairly simple (I think).

Place to store data (mostly media) that has redundancy incase a HDD fails. Basically a Plex server at this point. Once I have that setup right, Ill look at other things - one step at a time.

Low write speeds are fine, will be using various hdd sizes, but using docker containers to run bits a bobs.

Containers at this stage will be Plex and torrent, considering the arr collection, krusader, reverse proxy for remote access, local cloud storage and a family planner of sorts for future use.

Looking about, it seems that the UnRaid parity setup would be the best option - hence why I think i have mucked this up.

In order to move the data pool I have in place currently over to the Unraid Array, with parity.

1 - move any data I want to keep off the data pool to external location ( currently doing - 4 hours to go)

2 - delete the shares I have created

3 - remove the disks from the pool and add back to the unraoid array, with the biggest disk (8tb) as the parity and anything else as data disks.

4 - recreate the shares, folder structure for media, and repoint Plex to the new media locations.

I think thats it. Have I missed anything?

Edited yesterday at 03:53 AM1 day by dolphinspanker
added in extra info

The pool currently has a missing device, and the GUI is not reflecting that. You should still be able to fix it and keep its data using the CLI.

You can then keep that disk as a pool or move it to the array as data if you prefer; I can post more detailed instructions if you like

1 hour ago, JorgeB said:

The pool currently has a missing device, and the GUI is not reflecting that. You should still be able to fix it and keep its data using the CLI.

You can then keep that disk as a pool or move it to the array as data if you prefer; I can post more detailed instructions if you like

Hi JorgeB,

I have , after much fluffing about,

managed to get all my data off the pool drives. Opening up all the directories /locations that i placed data shows nothing there.

Consolidated all my data across multiple hdds outside the server. (basically back to where it came from). The external drives with the data show everything located there.

And freed up one of the 8tb hdds i have pegged for the array, formatted as xfs. No parity at this stage until i move the pool disks over.

So next steps, as far as I can tell.

1 - dismantle the data pool. (2 x 8tb) - No need for this once i setup the array (which is what i meant to do the first time!)

2 - reallocate 1 as parity, and the second as data for the array.

This will yield 1 x 8tb as parity, and 2 x 8tb as data.

So - I attempt to remove the second hdd from the pool 'data' to force it to single drive mode to let it rebalance, and it throws this error, and wont let me progress.

So - Find the instructions to rectify this error - remove a pool and re-add it back. Have done this and now the second drive reports as 'Not installed"

I'm starting to think that I should just reallocate the second drive that is reporting as not installed into the array, reformat it as parity, and go from there.

The folder structure should appear as available in the array, so I can start to copy data back.

campbellot-diagnostics-20260705-1731.zip

Edited yesterday at 09:32 AM1 day by dolphinspanker
additional details

2 hours ago, dolphinspanker said:

I'm starting to think that I should just reallocate the second drive that is reporting as not installed into the array, reformat it as parity, and go from there.

If there is no data left on that drive, I would do that. No real point in fighting with a pool that you intend to remove, just remove it and assign that drive to the array.

Edit: and delete that pool.

Edited 22 hours ago22 hr by Veah

The pool is basically empty, so you could just recreate, but if you prefer to correct it, with the array running type

btrfs device remove missing /mnt/data

Then stop the array and reimport the pool with just that device; it should then only have the remaining device (which could also be used as an array data drive).

12 hours ago, dolphinspanker said:

I've been trying to understand how the Unraid parity setup is different from a Raid 1 pool.

Well unraid parity is more like RAID4 but without data striping of any kind. Each disk in the unraid array remains an individual filesystem. RAID1 is just a mirror of two drives which act as a single unit.

If you need more technical details I highly suggest reading the unraid documentation as it explains it in detail.

Edited 19 hours ago19 hr by MowMdown

Thanks to everyone for their help - I think i am on the right path now.

Have created the array, with parity - running the parity check now (I'm surprised that it has taken 14 hours, when there is nothing to check - oh well...)

Have removed the second drive from the 'Data' pool, and its happy running in single drive mode. This hdd will be removed eventually to join the array.

Found an article about turbo write. Once the parity has been setup correctly, will enable this and start sending data to the array directly (not to the cache first).

Feel like i am on the right path now. Thanks again.