Currently, I use dockerproxy + swag and Cloudflare for externally-facing services. I really like that I don’t have to open any ports on my router for this to work, and I don’t need to create any routes for new services. When a new service is started, I simply include a label to call swag and the subdomain & TLS cert are registered with Cloudflare. About the only complaint I have is Cloudflare’s 100MG upload limit, but I can easily work around that, and it’s not a limit I see myself hitting too often.
What’s not clear to me is what I’m missing by not using Traefik or Caddy. Currently, the only thing I don’t have in my setup is central authentication. I’m leaning towards Authentik for that, and I might look at putting it on a VPS, but that’s the only thing I have planned. Other than that, almost everything’s running on a single Beelink S12. If I had to, I could probably stand up a failover pretty quickly, though.
I switched from SWAG to Caddy. Its config file is much simpler, with many best practice settings being default resulting in each sites being like 3 lines of code. Implementing something like mTLS requires one line per site, just super nice to configure, and you’re not left without a template config for more obscure services.
That being said, SWAG does more than enough and Nginx is a powerful software so you really aren’t missing out on anything but more streamlined config.
Traefik is kind of just like, a nightmare that tries to sell you on it being “self configuring” but it takes some work to get to that point and the “self configuring” requires the same amount of time in a text editor as manually configuring Caddy does. I can see Traefik being powerful if you’re using it with actually clustered k8s and distributed workloads. If that’s not your use case it’s kinda just more work than it’s worth.