Selfhosted

I agree with the folks saying reverse proxy of some kind + WAF. That way end users don’t have to deal with the VPN, but your home system is not directly exposed.

I’ve been doing something similar with SSH local port forwarding and a $5/month VPS. Haven’t come anywhere close to my network quotas, and performance has not been an issue for home use with 2-5 concurrent users most of the time. I forward the local caddy ports to unprivileged ports/user on the VPS, then use the firewall on the VPS to forward that port to 443 and lock down the rest.

if theyre close, add them to your tailscale, if not and you have a web serve, use a reverse proxy.

for tailscale, you'd probably have to walk them through setting it up but then its one and done

best practice states security through obscurity is not to be relied upon, but compare ssh logs after one year on the default port vs a non-standard port and you'll immediately see why you want to use a non-standard https port for non-professional services. it cuts 99.9% of the noise/attempts.

call me crazy but I just use a couldflare tunnel since my domain is already provided through them. They make it really easy to add sub-domains attached to specific ports on the local machine and automatically adds the proper DNS entries. Additionally they have a ton of domain security/anti-bot measures you can set up.

Depends on what you mean by "secure." My personal setup is Jellyfin LXC on proxmox --> Wireguard to VPS -> Nginx reverse proxy on VPS.

This setup relies somewhat on Jellyfin's auth, but I'm comfortable with that risk. The LXC is blocked from sending local traffic on my network by firewall rules. Yes, someone could exploit a vulnerability in Jellyfin (though looking through the CVEs I'm not overly worried about that), then escape the LXC and fuck with my server. But that's a lot of work for no profit.

For more protection (in sense of reducing traffic that even interacts with your server), I'd recommend getting a wildcard cert for the domain so that the actual subdomain jellyfin is on is undisclosed to anyone not using your service.

Security isn't about making everything impregnable, it's about making attacks more trouble than they're worth. Otherwise, we'd all live in fortified bunkers surrounded by landmines. 🙃

Tailscale or cloudflare

I agree with reverse proxy + middleware. I’d also suggest something like Jellyswarm as the front end. That way I can connect to other friends’ servers too.

If they could be using web UI (I'm not sure how a client would work with auth like that) then it kind of depends on how much hassle they can live with and how secure you want to be

Simplest would be to use https://doc.traefik.io/traefik/reference/routing-configuration/http/middlewares/basicauth/
You set up username and password and share those with the user. But it can be brute-forced

Something more secure but also a bit more demanding would be some kind of email otp
https://docs.goauthentik.io/add-secure-apps/flows-stages/stages/authenticator_email/
Set up authentik to send time-limited link to their email that well let them through. But they would have to authenticate every time they access

Reverse proxy with fail2ban or crowdsec. It's possible to set up things like Pangolin which ultimately use a VPN between external and internal access points but not at the client, though it takes more setup if you want to use apps over pangolin instead of just the browser.

Why no VPN?

I toyed with the idea of exposing ports and decided against it. I don't understand networking well enough yet. For me specifically, VPN access has been perfectly workable in the US with both speed and ease of access.

Can you use fail2ban on Jellyfin? That might be a wise step.

I use a Wireguard tunnel to a VPS and fail2ban with geoblocking: https://codeberg.org/skjalli/jellyfin-vps-setup