1 hour ago, gnollo said:

taking a lot longer than a couple of hours

On 1/20/2026 at 7:43 AM, Frank1940 said:

about two hours per TB

1 hour ago, gnollo said:

Current position: 364 GB (3.6 %)

Estimated speed: 211.9 MB/sec

That is fast, as usual for the long, outer tracks. It will be slower as it gets to the short, inner tracks. Long circumference has more data than short circumference, but RPM is constant.

Event: Unraid Disk-Clear
Subject: Notice [TOWER] - Disk-Clear finished (0 errors)
Description: Duration: 17 hours, 27 minutes, 41 seconds. Average speed: 159.1 MB/s
Importance: normal

Disk is unmountable

Found the format button at the bottom of the webpage, ticked the box and now it is formatting and creatign a file system. Disk is XFS.

Edited January 23Jan 23 by gnollo

Just out of curiosity, what is the advantage of using rsynch instead of just moving all the files from one drive to the other with midnight commander?

rsync is the premium copying/backup tool for both Unix and Linux and it has been ported to virtually every other OS there is. It is virtually bulletproof in its operation. It can be configured to do about any type of copy or backup job. If the transfer is interrupted, it will resume flawlessly on restart. And it is freeware. The only downside is that is has to be run from the command line.

1 hour ago, gnollo said:

just moving all the files from one drive to the other with midnight commander?

If you want to do the conversions with a minimum of risk to your data, you do NOT want to move the files. You want to wait until you are sure that the copy operation has successfully moved all the files to the new Drive. Then when you change the format to XFS, the files will then be 'deleted' as a result of the formatting operation.

1 hour ago, Frank1940 said:

If you want to do the conversions with a minimum of risk to your data, you do NOT want to move the files. You want to wait until you are sure that the copy operation has successfully moved all the files to the new Drive. Then when you change the format to XFS, the files will then be 'deleted' as a result of the formatting operation.

Copy is also faster than move, since move is copy from source to destination (a write operation) then delete from source (another write operation). And all write operations in the array also update parity.

You should exclude the destination disk from user shares in Global Share Settings, then include it again after you reformat the source disk.

Thank you, Rsynch it is.

Stopped array. Settings / Global Share settings

Enable disk shares: Yes

Enable user sharres: Yes

Included disks: all

Excluded disks: disk 12

Permit exlusive shares: No

Tunable (support hard links): Yes

Tunable (enable direct IO): Auto

Number of fuse filee descriptors: 40960

Apply / Start Array

So I will run the emby docker over the weekend and start the procedure on Sunday late night.

1) click on the terminal icon on the unraid web interface

2) as my other largest (10TB) drive in the array is disk4, type: rsync -avPX /mnt/disk4/ /mnt/disk12/

3) prompt should return at the end with no errors showing

4) as I am paranoid, will check integrity with typing: rsync -avhPX/mnt/disk4/ /mnt/disk12/

5) if all goes well it will just return to prompt

6) stop the array

7) Click on Tools, then New Config, then Preserve current Assignments

8) select All, then check Yes I want to do this, click Apply then Done (if unmountable disk(s) appear

then I need to select the correct file system for those disk(s) whichh in my case would be rfs for all but one disk)

9) I go back to the Main page and click on the dropdown for disk12 (the xfs drive) and unassign it (I click on "unassigned" or "no device")

10) I click on the dropdown for drive 4 (the one I already copied), and unassign it (make a note of the drive details) so I can then reassign it as disk 12

11) I click on the dropdown for the drive 12 and reassign it as the physical rfs drive I have just copied across

12) Click on each drive name and swap also the file system format of the drive - if it's ReiserFS change it to XFS, if it's XFS change it to ReiserFS [this was not possible on my version of the OS as reiserfs is grayed out as an option, thankfully all drives were recognised as rfs except the one I just converted to XFS)

13) All array disks are with blue icons, on the main Tab scroll to the bottom and click on the check box for Parity is already valid, then click on start

14) stop the array and click the disk 12, and change the format to XFS, then click Apply and Done

18. Start the array; at the bottom of the Main tab, you will get a warning Unmountable disk present: tick the Yes I want to do this box under the message "Format will create a file system in all Unmountable disks." A warning box will popu up, click on OK then click on the Format button for disk 12

Edited March 3Mar 3 by gnollo
added steps

42 minutes ago, gnollo said:

So I will run the emby docker over the weekend and start the procedure on Sunday late night.

1) click on the terminal icon on the unraid web interface

2) as my other largest (10TB) drive in the array is disk4, type: rsync -avpx /mnt/disk4/ /mnt/disk12/

3) prompt should return at the end with no errors showing

4) as I am paranoid, will check integrity with typing: rsync -rcvpx /mnt/disk4/ /mnt/disk12/

5) if all goes well it will just return to prompt

6) stop the array and swap drive assignment

7) Click on Tools, then New Config, then Preserve current Assignments

8) select All, then check Yes I want to do this, click Apply then Done

What you are doing by making the checklist is what I did when I did the conversion. I actually had a table which listed the actions and parameters needed for each step. I then checked off each step as it was completed. Checking was my way of keeping track of where I was in the conversion of each disk. Many of the steps require hours to complete and the table with its checked off steps verified exactly where I was in the process at all times.

Yep I like lists, I hope it helps others that will go through this process

Quick question. Stats shows 2% data on disk12 is that right? Have not started moving anything yet.
Also I tried reaching the drive to see what files are on it, with \\tower\ in windows explorer but I cannot see disk 12.
Also tried \\tower\disk12, no can do.

Is that because Export is set to NO under SMB securities settings for that drive, I guess that is because I excluded the drive from the shares?

12 minutes ago, gnollo said:

lso I tried reaching the drive to see what files are on it, with \\tower\ in windows explorer but I cannot see disk 12.
Also tried \\tower\disk12, no can do.

Rather than looking at it from Windows, Use the built-in File Manager. You can open by clicking on this icon in from of the disk.

Next thing. When you format a disk, the process adds the file system overhead onto the disk. This takes up disk space...

2 hours ago, gnollo said:

Is that because Export is set to NO under SMB securities settings for that driv

That would stop the drive being visible over the network.

2 hours ago, Frank1940 said:

Rather than looking at it from Windows, Use the built-in File Manager. You can open by clicking on this icon in from of the disk.

Next thing. When you format a disk, the process adds the file system overhead onto the disk. This takes up disk space...

Thank you, I got to browse it that way

I found

disk12\appdata\EmbyServer

with the following folders labelled as <SHARE> under the size column

1) Cache

2) Data

3) Logs

I guess written by the emby docker before I removed the drive from the shares (116KB when I selected and clicked on calculate)

disk12\unraid\emby windows server backups\Emby Server Databases - 2026-01-24 00.11.10

also I guess written by the emby docker (306MB)

On 1/7/2026 at 2:35 AM, Frank1940 said:

One thing to realize is that after you finishing copying one disk, you have a window of opportunity to permit family viewing of 'their' movies and shows. Then start on the next disk. (I also believe you can speed up copying by going to SETTINGS >>> Disk Settings and change this parameter as shown:

Interesting, thank you I just changed that setting. Any drawbacks from doing that?

23 minutes ago, gnollo said:

Thank you, I got to browse it that way

I found

disk12\appdata\EmbyServer

with the following folders labelled as <SHARE> under the size column

1) Cache

2) Data

3) Logs

I guess written by the emby docker before I removed the drive from the shares (116KB when I selected and clicked on calculate)

disk12\unraid\emby windows server backups\Emby Server Databases - 2026-01-24 00.11.10

also I guess written by the emby docker (306MB)

You should really stop all Docker containers and VM's while doing this process. I have the feeling that you have a fair number of large capacity disks to convert and I know this may be painful due to WAF, It really causes a problem when the disk you are copying from gets updated after rsync has finished that part of the directory tree on that disk. If you can't do this, run rsync again after the copy has finished and see if it has had to update/add files to the destination disk. (If no action is required, this only takes a couple of minutes.) Obviously, you have to repeat until no action is required!

Just now, gnollo said:

Interesting, thank you I just changed that setting. Any drawbacks from doing that?

Yes, all the disks on the array have to be spun up rather than just those actually involved— parity disk(s), source disk and destination disk.

2 minutes ago, gnollo said:

Interesting, thank you I just changed that setting. Any drawbacks from doing that?

The drawback is that any writes to the array requires all drives to be active so they are much less likely to spindown thus potentially increasing power consumption of the server.

4 minutes ago, Frank1940 said:

You should really stop all Docker containers and VM's while doing this process. I have the feeling that you have a fair number of large capacity disks to convert and I know this may be painful due to WAF, It really causes a problem when the disk you are copying from gets updated after rsync has finished that part of the directory tree on that disk. If you can't do this, run rsync again after the copy has finished and see if it has had to update/add files to the destination disk. (If no action is required, this only takes a couple of minutes.) Obviously, you have to repeat until no action is required!

Yes absolutely, will turn all dockers off when doing this, 11 drives to move, 52TBs of data. MIght even start the server in safe mode. If I do so can I still connect from another PC to it?

3 minutes ago, itimpi said:

The drawback is that any writes to the array requires all drives to be active so they are much less likely to spindown thus potentially increasing power consumption of the server.

OK so once I have finished this project I will set it back to auto

I guess I can expect 3-4 Hrs per TB? Bottleneck here being the writing speed as it is all happening inside the server?

Also with XFS format can I fill the disks to the brim in the future? Currently all of my drives are at 89 or 90% utilisation.

On 1/24/2026 at 5:07 AM, gnollo said:

Enable disk shares: Yes

I always recommend NOT sharing disks on the network.

19 minutes ago, gnollo said:

with XFS format can I fill the disks to the brim in the future?

It is always a good idea to leave some space free for check filesystem to work with. Also, you should set Minimum Free for each of your User Shares, of course.