Ah, well, then perhaps I will monitor it.

For internal use I just monitor everything with zabbix. What Ive been wanting is (as I said) a public “status screen” that my few users can hit just to verify if things are in fact down or if it’s just them.

Ok, you might have finally gotten me to consider a “dashboard”. I’ve been wanting a simple public facing service status page and this sounds like a nice solution.

Haha, I’ve never had to deal with something quite that high pressure but I’ve definitely been a little looser than standard during at least a couple emergencies.

Kernel upgrade WHILE you’re out for beers!

Someone just posted their own short reviews of a slew of wiki options in this community so maybe go take a peek at that.

Personally I’m finding I like Otterwiki quite a lot though I’ve not yet dug deep into it.

I use portainer extensively and am quite fond of it. I normally live on the CLI so picking a GUI tool over cli management is unusual for me but I’ve found portainer largely just makes typical management easier and doesn’t get in my way at all.

For your other questions I have no answers. I self host everything so to me “what is worth running” is not a question that makes sense. I run what I need and my needs therefore define what I run.

I stick to IRC over matrix.

500Mbps isn’t a measurement of electrical consumption

To each their own but I think I prefer to stick to Nextcloud and just continue to keep things organized the old fashioned way for the most part.

Though for documents those all get fed to paperless-ngx

That sounds fine. I would only say don’t use Syncthing to actually make your backups. My preference and recommendation is restic, possibly combined with the helper utility autorestic.

Lidarr isn’t really meant for that so you’re always going to fighting things I unfortunately

NPM likes to eat the let encrypt requests which is what I’m assuming is breaking the cert gen inside the container. I believe you can work around this, but honestly I’d recommend just moving to a more advanced but more flexibile proxy solution.

Personally I recommend Traefik. There isn’t a friendly gui to help you but once you wrap your head around it things just work. It also allows for defining proxy parameters right in your compose file via labels so it takes out the need to log into NPM and manage proxy entries there. Just deploy you’re compose fils and you’re off.

As far as making what you’ve got just work, you can either try to get NPM to stop intercepting the LE cert requests or hack up the signal-tls-relay container and jam the NPM certs into it. I wouldn’t recommend either of these options though. I’ve been in a similar scenario and it’s this among other reasons why I moved off NPM. I started with NPM because I thought it would be simple and easy and it is, right up until you want to do a thing even slightly outside of its fairly limited box.