JonathanM

As far as shares are concerned, there is no difference between a VM and a normal desktop, same rules apply. Is the VM getting an IP from the same subnet as Unraid from your DHCP server? Can you ping from the VM to Unraid and vice versa?

I suggest a belt and suspenders approach. For the day to day automatic backup, use software inside the VM appropriate to the OS, for example a windows VM could use acronis backup or whatever else you would normally use for a desktop with that OS. For disaster recovery, periodically shut down the VM, make a copy of the VM's raw XML, you can just copy and paste to a text editor and save it, then make a copy of the vdisk.img file and keep it someplace safe, preferably other than on the same server. That way, for typical restore you just roll back using the OS based software, but if the VM won't even start, you can create a new VM with the vdisk file and XML, then apply the latest restore from the OS.

Can you add dedicate a NIC port on your proxmox to a pfsense install? That way you could plug Unraid into the LAN port of pfsense and not into the shared switch. Pfsense as a VM in Unraid does work, I'm personally doing it, but not as an offsite hands off thing. It definitely wouldn't work for a trial install of Unraid, because of a chicken and egg issue. Unraid trial requires internet to validate and start the array, and the array must be started for VM's in Unraid to start. That's not an issue with a full license, as there is no phone home required. Even with a full license, you need access sometimes to troubleshoot issues and if the pfsense VM won't start, no access. How do you access the IPMI on your proxmox hardware? If you can do the same with the Unraid unit's IPMI, it might work out ok.

As you found out, once the stick is disconnected, you must reboot to get it reconnected, and it will trigger a parity check because the shutdown can't be properly recorded on the stick. Just keep regular backups of the USB stick, I wouldn't transfer just yet,

This is after rebooting Unraid?

For external drives you need to use eSATA without port multipliers, or SAS. USB isn't stable enough for parity array disks.

Soon™?

Sure. Parity is realtime, it tracks all bits as they are changed, moving files, formatting drives, whatever. Parity doesn't store data. If you want to use mc to copy all data from /mnt/disk1 to /mnt/disk3, and /mnt/disk2 to /mnt/disk3, go right ahead. What you must not do is mix /mnt/diskX with /mnt/user, because /mnt/user is just the combined list of all the files on /mnt/disk1,2,etc.

You will have to do a new config, because as you said, the drive ID's are going to change. There is a good chance the enclosure may also change how the drive geometry is mapped, causing the partitions to become unmountable. So, the safe thing is to copy the data from the external drives elsewhere before making the transition. You could copy internally from drive to drive using mc or similar, but with so little data it may be easier to just copy it to your desktop.

Yes. That kind of damage is typically caused by a loose fit, where the metal on metal doesn't firmly hold, causing only a small spot to touch, causing high resistance and temperature. Technically the power supply end is what is supposed to provide the clamping and spring force, but the board tolerances probably didn't help. The motherboard could possibly be salvaged, by thoroughly cleaning the burnt parts on the metal posts inside the connector, but that would be a serious pain to do, probably with a very small bit on a dremel.

I do. The only feature I have enabled is the USB backup and key recovery.

Are you going to file a feature request / bug report?

LOL. That was my intention. It's not a clear cut situation. I'll add another factor to the equation. Parity doesn't hold any sensible data by itself, it works in conjunction with all the data drives in the parity array. So assuming one parity drive, if one drive fails it is emulated by the rest of the drives, if 2 drives fail, you lose the data on both failed drives. If that happens, you are much better off if parity is one of the failed drives, as it doesn't have any data. So using that info, parity should be the lowest priority for reliability, highest speed for writing, and data drives should be high priority for reliability and read speed. Bottom line, it doesn't make a whole lot of difference in the large scheme, do whatever makes you feel best. Expected lifetime of a drive is a crapshoot not likely to be influenced by whether it's a parity or data drive. Personally I've had better luck with WD than Seagate, so my suggestion is put the Seagate as parity. Since I've only personally had experience with 100's of drives vs. hundreds of thousands manufactured, my statistical experience is meaningless, you do you.

Conversely, if you use your server as many do, where a file is written once and read many times, only the write effects parity, reads are done only from the data drive. So the only time the parity drive is read is during parity checks or emulation and rebuild operations. So, the question is, how are you going to use your server? The usage will determine which drives would be optimal for which positions.

Watercooling is for desktop showpieces that you sit beside while it's on, not servers that run unattended. Failed pumps and leaks, small or large, can be fatal if not caught quickly.

It's not terribly clear to me, but it sounds like they want you to be in the mailcow container to run the script, not on the host.

It's a linux thing, not specifically Unraid. Marvell still operates under the assumption that microsoft rules the computer world.

After you do what ChatNoir said with your boot USB, you will need to format your data and cache drives as well. Unraid does a really good job of trying to keep you from losing data, so you have to go out of your way to completely start over.

I'm not aware of a single controller chip that can manage that many ports. It would be interesting to see just how this is handled, whether it's with multiple controller chips or port multipliers.

Physically smaller drives are more prone to failure due to heat buildup.

did you remove vnc?

Unraid isn't designed for that.

1. Router hosted VPN. 2. If you use shared key SSH it's not too bad.

Different PCIe slot?