Hoopster

For the most part, this is true. Unraid runs on just about any hardware. It does not support ARM processors, but it runs without issue on most Intel or AMD based hardware. The Unraid license is tied to the GUID of the USB flash device which also holds the OS files (loaded into RAM on each boot) and the Unraid server configuration. This makes Unraid "portable" if you make hardware changes. I have personally changed motherboard, CPU, RAM, etc. five times on my main server while still using the same USB flash drive to boot Unraid.

No need, really. The appdata and system shares are used by docker containers and VMs. If you are using neither Docker nor VMs (they are disabled), nothing will be written to appdata so it doesn't really matter where it is. If you decide to use docker containers or create VMs in the future, appdata and system will be on cache which is really where they should be.

If you use the same USB flash drive with your current Unraid license with new hardware (assuming you are keeping the same array/pool/cache disks and configuration) it is pretty much a seamless update with no changes. Unraid is not hardware specific in most cases. The only thing that could be a potential gotcha is VMs if you are doing any hardware passthrough with current hardware. That would need to be reconfigured with hardware changes.

I have changed hardware 4-5 times (motherboard, RAM, CPU, sometimes PSU) all with the same USB flash drive as it contains your Unraid configuration in the config folder.

As mentioned above, everything related to your current configuration is in the config folder on the root of your USB flash drive. The really important things to keep are the *.key (Unraid license) file, e.g Pro.key, and the super.dat which is your current disk configuration. With many configuration files, deleting the file from the config folder will cause it to be recreated with the defaults on the next reboot. Essentially, this is like starting over with that configuration. If you delete network.cfg, for example, you will get the default network configuration with DHCP enabled on the next reboot.

There are also subfolders in config for certain things like shares. This is where your current shares configuration is located. If you delete those files, you wont lose the shares, as they will be recreated with defaults on reboot; but, you will lose how those share are configured.

The quickest way to "start over" with the Unraid config is to make a backup copy of the current config folder, recreate the flash drive, and then copy your *.key file and super.dat from the backup back to the config folder. This will boot on the new hardware with the current license and disk configuration and everything else will be back to defaults. If you are using the same array/parity disks with the current data, shares will get recreated automatically with default settings as top level folders on array disks automatically result in shares of that name in Unraid. If there are other configurations you know are good and you don't want to redo them, you could copy those .cfg file back into the config folder as well.

Unless you use the --delete option, files that are deleted in the source are not deleted in the destination location

The rsync command uses the --delete option to remove files from the destination directory that are no longer present in the source directory. This ensures that the destination mirrors the source, including deletions.

You would need to add the --delete option to your rsync command, so, rsync -avh --delete --progress /mnt/disk3/Photos/library "/mnt/disks/4TB Easystore/Immich"

mnt/diskX path references a specific disk. If you want to reference the share, which can span disks depending on how you set it up, replace the /mnt/diskX/folder path with mnt/user/Photos.

A share can exist across multiple disks. When you setup a share you can limit it to one or more disks if you wish. How is your Photos share configured. Also, do you have the share set to use a cache drive? If so, any files created in the share will actually exist o the cache disk until the mover moves it to the array. Just some things to check with your Photos share.

I do not have the folder view plugin installed. Since updating the Unraid Connect plugin yesterday, the Docker tab buttons have disappeared. They flash briefly at the top of the Docker tab when it loads and then disappear altogether.
I also updated the Unraid Connect plugin to the latest version available today (2025.09.08.1804) and the problem persisted. I removed the Unraid Connect plugin and the Docker page displayed normally with the function buttons at the bottom of the page and functional.
I reinstalled Unraid Connect (2025.09.08.1804) and the Docker page problem returned. This time the buttons don't completely disappear but they are stuck at the top of the page and not functional. Just ghosts.

Yes. Unraid tracks disks by serial number not by SATA port or slot location. The only exception would be if the HDDs were connected to some kind of RAID controller (not in RAID mode) that messes with serial numbers.

Don't think you can limit drives used by subfolders. I could be wrong, but, I don't know a way to do it; maybe something "under the covers".  Drive assignments can be done by share.  I have created separate shares for media types (movies, TV, photos, etc.) because I like to manage them that way in Plex,  I can limit shares to using particular drive(s).

The fix is for docker containers in general running host mode with Tailscale and not Plex specific.  See the 7.0.1 release notes for more details.
 
With respect to running Plex in a mode other than Host, I have done it two ways; Custom: br0 with a static IP address on same subnet as the host and on a VLAN I created for docker containers on a different subnet than the host (this showed up as Custom: br0.3 as it was on a xxx.xxx.3.xxx subnet.  Both worked fine with no problems with direct and remote streaming. 
 
The only problem I had with Custom: br0 was macvlan call traces. These would crash my server every few days to a couple of weeks. This used to be a big issue and one of the reasons ipvlan was implemented.
 
Because of the macvlan call traces on br0, I created a VLAN (br0.3) and all my problems went away.  This was years ago. I have not tried going back to br0 although I did switch to ipvlan. Having Plex on a VLAN with Unraid host and the content on a different subnet was not a problem because my router automatically passes traffic between "corporate" subnets.

Isn't what you want in Network type br0?  That or a Docker VLAN is the way I do it.
 

That is just the space taken up by the XFS file system. There is no data on the disk.  The file system supports a lot of new features now that take up some space when the disk is formatted.

Been posted a few times.  Here is the fix, it worked for me.
 
 

Found the problem.  Somehow Disk 5 was excluded in global share settings. Nothing I remember doing, but, that was the case.

Because I can 😀 
 
Actually, I just discovered the Site-to-Site VPN capabilities of my routers and set it up just last week.  It was super easy and is based on WireGuard.  I do not have a lot of experience with it yet but I m about to move my backup server down to our second home this week and I will be trying it out a lot more.  I intend to do unattended Unraid to Unraid server backups over Site Magic utilizing User Scripts and the IPMI capabilities of the servers.
 
WireGuard on the 24x7 Unraid server is what I have been using for years for remote LAN access to Unraid and other resources on the same LAN.  I also included my docker container VLAN subnet (192.168.3.0/24) in the allowed IP addresses for the WireGuard client devices (phones, laptops, PCs, tablets) from which I access the LAN and VLAN remotely.  It works great.  WireGuard on a Raspberry Pi in the LAN is my backup entry point in case there is an issue with it on the Unraid server.
 
For this setup, I have two ports forwarded in my router, one for WireGuard on Unraid and one for WireGuard on the RPi.
 
I am not as familiar with Tailscale use as I just started using it recently but it is also based on WireGuard.

USB stick should remain inserted in server once the system boots up.  Unraid loads into RAM using the system and configuration files located on the USB stick but any configuration changes are written to the stick.  Unraid will complain if the USB stick is removed once the system boots.
 
Yes, every boot into Unraid requires the USB stick to be present in the server.
 
The licesne is just a key file.  A particular version of Unraid does not come with the licence.  The license can be used with any version of Unraid you install to the USB stick.
 
 
Changing boot drive in the BIOS should allow you to choose to boot Unraid (USB stick) or Windows 11 (SSD).  Nothing in any verison of Unraid should (or could) prevent this as far as I know.
 

A UDMA CRC error is a problem with the transfer of data between host and disk.  It often comes down to poor or loose disk cables.  If the system detects a transfer problem it retries sending the data which often works just fine.  UDMA CRC errors become a concern when they grow rapidly in quantity or occur frequently on the same disk.
 
Usually, the first thing to check is the cabling to the disk.  Is it a secure connection?  Does swapping out the SATA/SAS cable eliminate the errors?
 
I once "replaced" a disk that accumulated 1109 UDMA CRC errors very rapidly.  I say replaced in quote marks because I rebuilt the disk back onto itself.  It was not a new disk.  The UDMA CRC errors disappeared on that disk and have not returned.  It may have just been the act of removing the disk from the hot-swap cage and reinserting it that fixed the problem as there could have been a bad connection.  I rebuilt it just to be safe.
 
UDMA CRC errors can technically result in data corruption if multiple data resend tries do not fix the issue, but, that is rare.

The license being tied to the USB stick GUID does not affect access to your data at all.  No data is stored on the USB stick and the license does not control access to data.  The only things there are the Unraid server configuration, the license key and the files needed to unpack Unraid OS into RAM each time the server boots.  If you have to create a new USB stick because of the failure of the previous stick, you need to have a copy of prior USB stick for the server configuration.  You could then run in trial mode until you can get the license transferred to the new USB stick.  You can do this one time per year automatically through the GUI.  If you need to do it more than once a year, contact Support directly and they will help you with the transfer if you explain the situation to them.

Often, flash devices that show as incompatible in the Flash Creator can be successfully used with the manual installation method.

No, not normal.  My 8TB parity takes a little over 16 hours to build.  A reasonable rough estimate is about 2 hours per TB.  Your 12TB parity should rebuild in approximately 24 hours if everythng with your server is as it should be.  Clearly, there are some problems.
 
You can use your server during rebuild altough that can slow down the rebuild depending on how heavy the use.  You should minimize the use.  Another option is to use the Parity Check Tuning plugin and spread a parity build over several low-use periods such as at night.  I typically do my rebuild in two roughly 8-hour (midnight to 8am) sessions.

They probably have blank variables, which, up until version 6.12.8 of Unraid, was not a problem.  This happened to me as well.  A docker container disappeared and could not be added again even through Previous Apps.
 
You likely need to install the Docker Patch 6.12.8 plugin which will fix this and let docker containers with blank variables be installed.  You can then decide if you need to fill in these variables.  This, of course, assumes you are running version 6.12.8 of Unraid.

I don't know what 5G service you are planning to use in your home, but, the T-Mobile 5G gateway, for example, has an Ethernet port to which a switch or router can be connected.  A lot of third-party routers have 4-port switches built in.  In either case, it should be possible to connect your Unraid server to an Ethernet port if you go with a 5G Internet service that has a gateway with an Ethernet port.
 

No, shutting down does not trigger the Mover.  Mover can be run manually before shutdown if you wish, and, of course, the schedule can be changed to any time you wish.  3:40am is just the default. 
 
How long mover takes depends on how much data needs to be moved from cache to the array.  You can control by user share if it writes data first to cache (moved to array by Mover) of if it writes directly to the array.  I don't have any shares currently writing first to cache so the Mover is not an issue in my case.

You have export in share security settings set to No.  This hides the share from Windows.  You need to set it to public or another setting with secure access if you want a share visible in Windows.

It does a lot more than that.
https://unraid.net/connect
 
Have you provisioned a myunraid.net certificate?  https://docs.unraid.net/connect/help/#manage-your-server-from-within-the-connect-ui

In the settings for the drive in Unassigned Devices, did you enable sharing?
 

You have macvlan call traces.  Try switching the docker network type to ipvlan.  Settings --> Docker --> docker custom network type
 
Prior to Unraid 6.12.x (don't know what version you are running since you did not include full diagnostics) usually macvlan issues only occurred on systems where custom IP addresses were assigned on br0.  Now, (6.12.0+ as mentioned in the release notes) with some hardware, macvlan problems can appear even without custom docker container IP addresses.