ejabberd-contrib VS docker-swag

Also a fan of XMPP, ejabberd.im is a formidable piece of software that pretty much just works once you have it all set up and configured.

Yes, work out-of-box, and in case, you like to deny OMEMO in your server, there is even a ejabberd-contrib module for that.

I have a NAS on .0.181 and a swag container (on a different port than nginx) on .0.180 that points to my public facing services. For obvious reasons, I don't want my public domain to point to any other ports/addresses on my home network. Additionally, as elegant as swag is, it requires authentication and so won't work for simple local DNS. I now have one local domain for each server and an nginx instance on each that resolves to my different services on each.

OP is even linking the Github... https://github.com/linuxserver/docker-swag

I have an UnRAID server with a few services (Jellyfin, Nextcloud, etc.) running on it behind Linux Servers' SWAG reverse proxy container, which is built on Nginx and Let's Encrypt. This is pointed to a DuckDNS link, which is then pointed at my domain with a CNAME. So I can access Jellyfin, for example, at jellyfin.mydomain.com. A few weeks ago, due to seemingly unrelated issues, I got a new modem/router, an Arris SURFboard G34. For the first few weeks, everything was working as before. But now, when on my LAN, I can't get to my services at the proxied domain. It times out every time. There are no errors in SWAG's logs, nothing seems amiss in the router's web interface, and the services are available both at their IP:port address and, when not on my LAN, I can access them at the domain no problem.

Reverse proxy, probably? I use Docker SWAG, setup here, with DuckDNS and it works really well for me. There are of course many ways to reverse proxy, as I linked to earlier.