OMV6 MergerFS: How to Properly Use MergerFS to Pool Drives

Hi Community,

I found about OMV and wanted to learn and set up my own personal NAS to store films, music, and documents. I decided to pull the plunge and buy/use the following devices:

• Raspberry Pi 4 running on Bullseye Lite, with OMV6 installed(using the OMV guide)
• Sabrent 5 bay hard drive enclosure, holding 3x10TB and 2x12TB hard drives
• Using RPI-4 USB 3.0 to USB-C communication to the bay.

I was able to properly wipe my drives in a clean state(Secure Wipe), create a new filesystem, mount the drives, and create a shared folder in a specific drive where my laptop was able to detect the SMB/CFS network share with the shared folder I created. So far so good.

My next step was to figure out how to combine specific drives as a single drive; combining 3x10TB and 1x12TB drives together to get 42TB in total. Upon researching, I found that MergerFS with SnapRaid would be the best solution to my setup.

However, I am having trouble understanding how to set it up properly with MergerFS. From researching online, I see that on OMV6(after installing the mergerfs-6.0.x plugin) I can navigate to the mergerfs tab.

My question is: how would I pool all the selected drives together? I am looking at Snapraid + mergers files copied to a single drive forum and I can enter specific drives/UUID in the "Path" section. However, when I manually copy and past the specific mounted drives and save it, I get an error where I get a 256 character limit. The following below is an example:

/srv/dev-disk-by-uuid-c05648de-9886-4234-a5d3-b8ec2cfb0ab8
/srv/dev-disk-by-uuid-bdcd7360-2359-478a-b374-9194ffc47bc9
/srv/dev-disk-by-uuid-2fd2c0a5-ee84-482d-8283-63fa56ff0e7e
/srv/dev-disk-by-uuid-c86cff6c-aac4-4583-beb4-1b36932cf06b

From looking at a different post, it says that I would have to shorten these down where it will accept it. Would this be the appropriate way to pool specific drives that I want? Below is the example:

/srv/dev-disk-by-uuid-[cb2][0df8]*/

Is this the proper way to enter in the mergerfs "Path" section when using the OMV webgui? Would love to hear someone from experience/expertise on how to get this properly working.

Hi everyone, just to give an update, I went with the path of using the following command to pool specific drivers using the wildcard syntax in the "Path" Section:

/srv/dev-disk-by-uuid-[cb2][0df8]*/

From here on, I was able to create my Shared Folders using the pooled/merge drives and able to share it out using SMB/CIFS and FTP.

Quote from noiro

With the latest mergerfs plugin you do not have to shorten the path..

I have similar setup.

I named the disks d1, d2, etc.

D1= parity, content

D2=content, data

D3=data

D4=data

Dx=..... etc.

Go to shared folders and share the D1, D2 etc.

In mergerfs you add the paths D1,D2 etc. make sure you setup a pool give it a name and from the drop down menu select the disks/filesystems

Choose a create policy ( I have most free space..)

Then go to services and share the Pool under NFS and/or SMB/CIFS then from the clients you can access the pool...

I hope it helps.

Display More

noiro chente Thanks for this advice, sorry for not looking back and responding sooner. Regarding Parity/SnapRaid, I should repeat the content of SnapRaid content in other drives when using the 12TB disk for parity? Sorry if this part is confusing, I am still learning how to properly setup my storage to use MergeFS and SnapRaid properly.

Quote from chente

Generally, the snapraid content is placed on each data drive and is not placed on the parity drive. I don't really know what the reason for this is, it can also be put in the parity unit without any problem as far as I know.

So with that being said and using noiro's post as an example to show my configuration, my SnapRaid configuration should be checked as such:

Drive_1   (12TB) = Parity
Drive_2   (12TB) = Content, Data
Drive_3   (10TB) = Content, Data
Drive_4   (10TB) = Content, Data
Drive_5   (10TB) = Content, Data

Drive_1 is my SnapRaid parity drive, and Drive_2 to Drive_5 are my merged/pooled drives they have content and data. Would this be the correct format?

Quote from Forssux

The only thing that threw me of was that the disks needed a filesystem with a fresh install and the disks needed to be wiped.
For a fresh install this can be done with fewer steps....

4*4TB drives are conten/data

2*4TB drives are parity

Mergerfs has a pool/volume of the first 4 drives

When first introduced to OMV, I just installed it on my RPI4 and didn't know I had to wipe my drives either. The thing that confuses me is that it gives either Quick Wipe and Secure Wipe. I wasn't sure what to choose so I chose Secure Wipe, even tho these hard drives were brand new, I did not know the difference. Is there a reference/forum where it tells the difference between the 2?

And the format you had for the drives for parity and content is exactly what I did. The next step is, what do I do for the scheduling? Is there a recommended scheduling when using SnapRaid?

Thanks for the doc! I'm looking into it right now. I'll be changing the thread to "Resolved" as it answered my question regarding MergerFS and how to to use it.