JonathanM

Automated port scans. VPN server hosted on your router or similar is a better option.

Change the auto start so the array doesn't come up automatically.

Create an archive file on the source end so you only need to transfer 1 large file.

Preclear has NEVER been necessary. Ever. Originally, many many years ago, unraid wouldn't fully start and allow you to use it while a new drive was being cleared in preparation to participate in the parity calculation. That created a need for the ability to clear the drive before you added it, and pre-clear was born. It matured into a drive testing tool and continues in that role today. Now, when you add a drive to a parity protected array, Unraid starts fully and allows you to use all the normal array functions while the newly added drive is cleared in the background. When it's done clearing, you will be given the option to format it and start using it to store data. Pre-clear isn't necessary now, and it wasn't necessary before, just a convenience. However... because unraid requires the full rated capacity of ALL drives, even empty ones, to be perfectly readable to accurately reconstruct a failed drive, it is prudent to test drives before trusting them in the array. This can be accomplished with the drive manufacturers tools, third party drive testing utilities, or preclear. It's just one option, a very thorough option, to test drives. If a drive is precleared successfully, then as long as it remains in that state with no changes, format, partition or otherwise, then when you add the drive to the parity protected array unraid detects the special preclear signature and trusts that the drive is clear and immediately offers to format it instead of clearing it. However, the preclear utility takes WAY longer to do its job than the built in clear which doesn't do testing and verification does, so the fastest option is indeed just to plug in the new drive and add it to the array.

You should have it set for cache prefer, that way if the files weren't open mover would have automatically put them on the cache where they belong. Turn on the help on the share settings page for explanation of the cache modes.

No, set the http proxy field in jackett to the IP address of the delugevpn container (typically the IP of your Unraid) along with the port defined in the delugevpn privoxy field, typically 8118. Like 192.168.1.10:8118

I don't use jackett, but a quick google leads me to believe there is a setting in the GUI for a http proxy entry.

Only a fraction of Unraid users read this forum, and only a fraction of those post. There is no guarantee that someone clueless enough to leave the server open is clueful enough to come here for help.

Every time you see the machine is online and unsecured, shut it down. Least harm possible for best benefit. It can't be hacked by anyone else if it's off.

@Doftorul, by proven, he means it works and more importantly doesn't screw up any existing functionality. Last thing we need is a change that breaks unraid for others. Does it pass SMART information and commands properly? Does it support spin down? Does it pass the serial numbers without modifying them with added or truncated ID strings?

Did you google the error? Second hit: https://www.redhat.com/archives/virt-tools-list/2017-September/msg00008.html

Yes. But it would be easier if you keep a current backup of your USB, so you won't have to figure out which disk goes in which slot. Backup is easy to do, Main GUI tab, Flash, Flash Backup.

Sorry, but multiple simultaneous drive failure is pretty bad for a storage server. Devastating, even. Something you may want to avoid replicating.

If the ambient temperature gets below 10 C I'd disable spindown. The temperature shock of spinning up a cold hard drive can have devastating consequences. Ask Tom @limetech about his garage experiences.

Look up "wax motor".

You have it easy. See my previous reply. Seal the heat in until it's too warm, then vent with a thermostat.

Heh. As long as it's below rated maximums, warm isn't bad. It's the extreme swings that will get you. In a garage situation, I'd invest in an insulated box with greenhouse type ventilation. Only allow outside air to circulate when the temperature is above normal room temp.

7 years is fine for server grade stuff as long as it's kept well ventilated and hasn't overheated from dust bunnies. The only reason I can think of to mess with a working system like that would be lack of CPU power along with high power draw. However... you can buy a WHOLE lot of electricity for the upgrade cost difference, so unless you are running out of CPU power I'd keep running as is. BTW, if you are in a heating heavy climate, with little or no need for A/C, then the power is not bad either. The power drawn by those servers is exactly the same economy wise as keeping a space heater running year round. Paying for A/C to cool your servers means you are paying twice the power, so may be more of a concern.

You are confusing 2 different things. The core technology, and the application using that technology. https://openvpn.net/faq/what-is-openvpn-access-server/

Those two facts lead me to bad RAM. Has this new machine had at least 24 hours running memtest with a clean result? If it's using ECC RAM, I think you will need to create a boot USB with the new proprietary memtest program to get accurate results. Perhaps as a test temporarily remove half the RAM and see if the symptoms change, swap with the removed set of RAM and repeat.

I can see the usefulness of this, however, I can see a bunch of confused support requests when the drive displays xfs before the array is started, and unmountable after it starts. Changing the entire meaning of a column based on the state of the array seems like poor UI. Just my opinion.

How would you propose handling the situation where the filesystem is set for auto instead of a specific type?

I don't use any USG's, I only have wifi radios in my sites. Sorry to mislead you, I didn't see any mention of a USG in your post, and didn't think to go back in your post history to see which devices you had.

Problem is, people in general have no clue how parity works in unraid, and even less of a clue that formatted is NOT clear. If this was implemented I guarantee there would be data loss within the first week, when someone put a freshly formatted drive in, told unraid it was clear, and then had a drive failure.

Using the override with FQDN works fine internally for me. Are you sure your NAT reflection / loopback is working properly in your router? I have 3 external sites and 1 internal site managed with this container, all working fine, with all devices using FQDN.