Squid

You've got 4 gig of memory, and with everything running it boils down to there might not be enough available. 4G is the absolute bare minimum of memory for the OS, and 8G would be the realistic functional minimum. I'd upgrade the memory and try again. Alternatively, you could also do the https://docs.unraid.net/unraid-os/getting-started/manual-install-method/ , but make a copy of the /config folder first and then after you set up the flash, copy your backup back onto it.

Can you drop to the terminal with a locally attached keyboard & monitor and get the diagnostics from there before you reboot. And this is after Connect was uninstalled? Also, have you tried running a memtest from the boot menu for at least a pass or two if only to rule out memory issues

Can you show a screenshot of Tips & Tweaks with your settings and what it's saying is the current set (on the right)

Before you cleared the CMOS, did you happen to notice if the Date / Time on the server was completely wrong (by a large amount)?

Question @whwunraid Is the docker.img file really old, and Emby installed into it a long time ago? (IE: have you ever recreated the docker image on say Unraid 7.0+, or is this still the original you created when you were running say 6.1 / 6.2 and Emby installed at that time?)

May 20 09:04:26 Tower publish: curl to session failed This is when those errors started. You logged in just prior. What did you do when logging in? Also, nginx had some weird issues after your reboot and it started the containers. What did you do at that point to get the UI back up and running (or did you notice it at all)

Was there any improvement? Also, besides connect, are you running docker compose plugin or something like Portainer or NextCloudAIO in docker?

The log file is composed of whatever is outputted to stdout in the script. Can't remember OTOH if it also catches stderr

Which plugin?

I'll check it out. But you do have to also bear in mind that what this plugin does is execute the find command in an attempt to keep the directory cached in memory. Million things can cause the cache to drop it, and with 2M files I wouldn't be surprised that it is.

I remember when powerdown was originally going to be removed from the OS and complained (it remained in lol). Muscle memory dictates that I always Reboot: powerdown -r Shutdown: powerdown

appdata.backup.plg - 2024.11.21 (Up to date) This is probably set to backup at that time and you've told it to shut down the containers prior to the backup

@grayraven I did replicate this on 7.1.2 using a private share. But, after I added [GLOBAL] map archive = no to smb extras in the WebUI, this behaviour did disappear. (I didn't test adding it as a via a [SHARE] section in smb-extras, but the attribute is usable within a share according to the man page) How did you test and find that map archive (or store dos attributes) didn't fix for you? (I haven't as of yet tried store dos attributes)

It's deprecated over semantics. It simply calls the command that isn't deprecated.

Reload the page and try again. It should give an option to clear the flag. But, there's also an issue that I saw where because the original you have installed is no longer present, but the replacement is CA thinks you've already got the replacement present. If you can't get the flag to clear, a reboot will work. Then uninstall from either CA, Installed Apps or the Plugins Tab and reinstall file activity from CA

Maybe this, and add to the template appropriate path mappings to your shares?

IDK. Looks like you installed un-get from @ich777 which is the replacement "nerd pack", but purely advanced users only and definitely caveat emptor software (hence why he won't publish it to CA). Personally, I'd uninstall that plugin. Also, FWIW you've got unbalanced plugin installed. That should also be uninstalled and reinstall unbalanced2

Start the array, let things get up and running for a bit then post a new set of diagnostics, stop the array and pull the ethernet cord out / shut off the server

Assuming you've got NerdPack / NerdTools installed, from its GUI you should be able to select to not install it. Alternatively, you'd find the package probably within /extra on the flash drive. Failing all else, post your diagnostics

Wait for @JorgeB to check this out. He's the uber expert

docker compose support IS coming... Promise

Not hacked per se, but every script kiddie is trying to log in. Is your server in your router's DMZ? Are you forwarding ports 80 / 443 / 22 to the server from the router?

FWIW, under OS version 7, the help text automatically disappears when you change pages in the WebGUI (IE, the "?" doesn't stay selected when switching pages). You're not the first person who that tripped up

Can you uninstall the connect plugin, reboot and then try again. Reason for the unresponsive gui is a file that was required is missing... Not quite sure how though. If that fixes the unresponsive, then see if the docker tab is working. If / when it fails, and you can live with it until I notice a response from you, then I might have some ideas, or troubleshooting ideas that might help with what's going on...

Can you open a terminal and post the output of ps -auxf Make sure that you already are seeing the "du" processes running