JonathanM

NAT loopback, reflection, hairpinning. Search your router model number plus those search terms.

Plug your drive sizes and RAID level into this page and see what you get. https://carfax.org.uk/btrfs-usage/

At some point you will want a larger data drive, then it will be necessary to have the corresponding parity drive. When to make that transition is up to you. Drive prices trend downward, so the longer you wait, the cheaper it will be (theoretically). However, each occupied slot has a cost, so buying smaller drives when larger are available at a lower cost per TB seldom (never?) makes sense. I don't buy drives until I need the space, which I personally define as having less free space on the array than my single largest data drive. Your budget and priorities may differ.

I'm dodging the question, but for a reason. You typically want to interact with your VM's using remote access software hosted inside the VM such as RDP or nomachine, and only use Unraid's VNC access for basic management tasks to the VM's that can't use the aforementioned remote access. Your experience with VM's will be much smoother. BTW, I'm not sure from your wording, but NEVER expose Unraid itself directly to the internet. Any WAN access to Unraid's GUI or SSH should be done through a VPN, wireguard is built in and just needs to be set up. If what I'm saying is unclear to you, PLEASE search the forum and ask rather than assume. Allowing general internet access to your Unraid server management will most likely cause crashes, and possibly worse.

Do you have destructive mode enabled?

I am aware of what your final intentions are, but when you had it set up that way, it wasn't working. We need to find out at what point in the process the errors start happening.

I was trying to get you to put the drives back in but NOT add them to Unraid's array. That way they will be drawing power, but not actively involved in the data process. Good troubleshooting only changes one thing at a time.

Did you physically remove the drives before you started this new sync?

Sure, this parity run was just to find out if the computer could actually be stable under certain circumstances. The next steps would be to start adding drives back and see where it starts causing issues. The fact that it runs well with only 2 disks implies that the issue could be either power or SATA controller. I guess the next step would be to connect the rest of the drives physically, both power and data, but don't set a new config, instead just start the sync again with only the two drives assigned.

We'll have a result soon, since previous runs wouldn't go beyond half an hour.

tools, new config

Tools, New Config

Wouldn't hurt to try one of the recovery programs on it to see if it finds any content, but I suspect it would be quicker and easier to re-rip from source.

Yep. I want to start eliminating variables.

Can you ping out from the console?

Ok. I think it's time to take a different attack. Do you have any data on the server yet? In your screenshots it appears to be a totally new setup. If there is nothing really set up yet, I'd like you to physically unplug all the drives except for the USB Unraid boot stick, and the two parity drives. Set a new config, with one of the drives as parity 1, and the other as data disk 1. See if it will complete a parity build.

https://forums.unraid.net/topic/51808-support-linuxserverio-swag-secure-web-application-gateway-nginxphpcertbotfail2ban/?do=findComment&comment=895908

Different manufactures can use special signalling that may not have programs compatible with linux / Unraid. Unraid out of the box uses http://www.apcupsd.org/ but you can install a plugin that uses https://networkupstools.org/ Just make sure whatever models you are considering can talk to one of those two programs.

Unfortunately, the drive's content list is most likely permanently gone, so recovery is probably going to be at best a jumbled mess of unsorted blocks of data that the recovery program may be able to deduce may be part of a file. It's the logical equivalent of taking a well organized wall of filing cabinets with thousands of labeled folders with various chapters of books and documents and upending the pages in those folders and drawers all over the floor, removing all the folders and labels, and giving the pile of loose paper a good kick just to be thorough. Your data is almost all still there, but putting humpty dumpty together again is a monumental task. What types of files were stored on that drive?

Pretty sure that the VM requires an address contiguous chunk of memory, and your free memory is extremely fragmented at this point. I haven't searched, maybe there is a way to tell the kernel to defragment RAM?

Keep in mind that you are excluding the host (containers included) from using any RAM you give to VM's. Many times it's better to give the VM absolute minimum required resources for performance and allow the host to manage everything you can give it.

URBackup?

Pretty any path besides /boot. When you have disks assigned, then /mnt/* starts to get populated.

Try copying the executable to a location in RAM before running it. /boot is FAT32, and is set up to deny executables.

kill-a-watt, or similar product. Server boards typically have IPMI, which is available as soon as the motherboard has power. This enables you to connect to the server and power it on, all you need is a VPN connection to your router. How often does the power go out in your area? Around here, if the power is out for more than a minute, it's probably going to be several hours before it's restored. Also, while planning this out, keep in mind that your internet and network pieces need to be powered as well. If your server isn't in the same space as your router and switch(s) you will need another UPS to keep them up as well. The load on that type of equipment is typically very small, so a cheap UPS will work fine if all it's doing is keeping the modem, router, and switch running.