Gragorg

If you dont have a backup or copy of your key you will have to contact Unraid to get the key back. You would need a new key anyways with a new USB. Even without Connect you can run a manual USB stick back up. Sounds like you have your order of the drives in your array.

Were you using legacy boot before? Did you try uefi boot?

This can be caused by a dusty/oxidized PCIe slot. I would reseat the LSI card and blow out slot if necessary. Also verify your cabling.

Yes in fact some say its more secure at you can close the port forwards to NPM. Just be sure to check on "Always us HTTPS" on cloudflare under SSL/TLS then edge certificates. Make sure you see the padlock locked beside your url name in the brower.

Start by posting your diagnositcs for one of the resident experts. You likely have a cabling or disk issue as it shouldn't take 12 hours to write 23% of a 12 TB drive.

You can move the data off one of the data drives and shrink the array. Once complete you reassign the empty data drive at parity 2. On the bottom right corner of the GUI click on manual and look up "shrink array"

Really weird. Maybe try turning off everything in NPM except "block common exploits" and "force ssl" and restart it. So everything runs fine local?

Have you tried leaving the advanced tab in NPM blank.

Are you using 80 and 443 for other dockers. Dont change if you already moved Unraid off those ports.

Do not use 80 or 443 as unraid uses them. Change 80 to 1880 and 443 to 4443. Then in your router forward external port 80 to internal 1880 and external port 443 to 4443 and give that a try. You only need to do TCP protocol

In the NPM template how are those ports setup? If memory serves docker port 8080 is set as 1880 internal port and docker port 4443 is set as internal port 18443. Is this the case for you?

I need the Nginx Proxy manager one.

I assume you setup from unraid app store with default settings?

Do not forward port 8096. Which NPM docker are you using? Did you change any port settings in the docker template? I assue you setup the cname using cloudflare. I have found that the one from Djoss works best for me.

What ports did you forward on your router? Your external port:8096 should not go through. How did you set your ports in the NPM docker template? Jellyfin should work fine with the advanced tab empty. All you should need is block common exploits and force ssl.

If you can't load the local GUI then it has to be an issue with Vaultwarden setup. I would focus on getting Vaultwarden running standalone without NPM first. Some people have had issue with bonding being enabled Unraid Network settings so you could try and disable that and see.

In the admin General Settings did you enter your https cloudflare URL under Domain URL?

If you open NPM and you click on your vaultwarden proxy hosts name does it load you to the login? Also did you use http or https in NPM for vaultwarden? Make sure your server time and date are also corrrect.

I would post diagnostics for experts for anything obvious. I assume your new internet is running the same ip range as your old provider and your server is on the same ip it was before.

At one time there was talk about an Unraid hardware database where users could enter their hardware builds and any issues that arose. Is this still in the works so that people building new systems can avoid any pitfalls?

You should setup syslog server and post diagnostics. Since its crashing so much setup "mirror to flash" is easiest. Memtest would also be worth checking.

Post your diagnostics for the experts

You can use its own just make sure the port forwards are correct.

Your port forwards look correct but you only need to forward tcp only. Plex is usually setup as HOST. Most create a custom docker network for arrs and Nginx. Nginx normally shares an ip with the server. If you want to create a custom docker network just type the following in terminal docker network create "name of network (without the quotes)" This network name will then show on "Network Type" on docker templates. Yes

Ya my update took way longer than normal also. I was thinking it locked up also. Not sure why it was so slow.