jonathanm

It checks for matches to the database, so it will detect anything in the database. At the file level, it's all just ones and zeros being checked. It's not like a full featured A/V inside your OS, that looks at critical run entries and such, all it does is scan the files.

All of you that have this as a frequent reocurring issue, could you check Settings, Global Share Settings, Tunable (enable Direct IO): and set it to No instead of Auto or Yes, and see if that changes anything?

Or, do a new config, save all, change the parity slot to the 10TB, and assign the 8TB to a data slot. Parity has to be rebuilt anyway, why not save the clearing time and just calculate parity based on the desired layout of disks.

I was also talking about the relationship between /mnt/user and /mnt/diskX. Since you said you were copying to disk1 directly and not using shares, it may be important to understand how they interact. If you are already up to speed, forget I said anything.

LOL. Well, that certainly explains your symptoms. How well do you understand the /mnt/user and /mnt/user0 structures? If you don't have a handle on how they work, it's pretty easy to mess up badly.

How are you sending data to unraid? There is a situation where allocation method gets circumvented, if the entire directory tree is created empty, and then filled with files, the already created folders are preferred over unraid creating new folders on a different disk. Also, you need to set your minimum free space to a sane value, for 8TB disks I'd suggest probably 80GB. When the minimum free space is reached unraid should create the needed folders and start using a different disk. The standard explanation says to set minimum free to slightly larger than the single largest file you are likely to copy, but I like to leave a little more free space on each disk in case the file system needs space for metadata or error recovery in the future.

Great, that means your proxy is working correctly! Just to be sure, is the network type on the docker configuration for both delugevpn and jackett the same?

Short answer, you can't. Long answer, you may be able to get some things back with file system tools or recovery software. How much time and money is this worth to you? Was it not backed up elsewhere? Regardless of your answer, I advise you to immediately shut down the array, because every additional operation after the deletion could make it more difficult to recover if you decide to pursue that.

Try configuring a browser to use the proxy, and see what happens. To see if it's working, try going to a site in that browser that tells you your external IP. You should be able to see your real external IP with the proxy setting OFF, and a VPN provided IP with the proxy setting ON in the browser.

After you physically swap the drive you may need to refresh the main page to allow you to select the new drive and assign it to the slot.

Does 6.4.1 support GUI upgrade to 6.7.x?

Would the P630 integrated graphics in the Xeon E-2176G work for passthrough to a VM? What about if I wanted to pass it through to Plex docker for transcoding (which I don't often do), could it not be passed through to both simultaneously? Depends on the motherboard and chipset for VM. I don't think I've seen a successful case. Definitely not both a docker and a VM. Normally people use a recent Nvidia card for VM passthrough.

Are you connected to the local wifi on the same lan as the server? Not a guest ssid or some other segment?

That's definitely not normal. My monthly parity checks take about 19 hours, and until the last month that I added the pause feature, that's been 19 continuous hours. Never had to reboot. What does 3am have to do with it?

AFAIK, the host only knows the current state of the VM, not the intended trajectory. If you need an event to fire before an attempted shutdown, it's going to have to come from inside the guest. If you can get a trigger from inside the guest, it's pretty easy to have it fire off a script in the host. I do have one suggestion, don't know if it will help. Monitor ping returns from your guest. When the ping dies, you know the guest is pretty much already shut down. Your scenario is light on timing details and intended actions, so it's hard to come up with suggestions.