Which reverse proxy do you use/recommend?
Which reverse proxy do you use/recommend?
I use both, Traefik on my docker host that’s also used for trying out new stacks, and NPM at work for a config that won’t change (ever, probably).
Yes, the NPM web ui is somewhat easier in regard to proxying targets outside Docker.
I use NPM in a docker container. It could not be easier in my opinion but then again, I did not use any of the alternatives so I might be missing out on something, who knows. I did manage a couple of proxy servers in the past based on Apache and I can tell you that NPM is much easier and logical to me than that.
Just create a compose file and start it. Create DNS records pointing to your NPM IP address/exposed IP and make a host in NPM sending traffic to the right container IP:port. The compose file is super simple, could not be easier. Here’s mine for example:
services: nginx-proxy-manager: container_name: nginx-proxy-manager image: 'jc21/nginx-proxy-manager:latest' restart: always ports: - '80:80' - '443:443' volumes: - ./data:/data - ./letsencrypt:/etc/letsencryptI just make sure ports 443 and 80 are exposed on my router so DNS records can point to that IP adrdess. All traffic on port 80 gets re-routed to 443.
I’m probably stating all the obvious things here 😀
I mean yes, that seems obvious now that I’ve learned this.
But I wish I read this comment 3 years ago when I was starting to dive into self hosting. Would have saved me a bunch of time. So always assume some piece of knowledge is not obvious for someone out there and share ᕕ( ᐛ )ᕗ
So always assume some piece of knowledge is not obvious for someone out there and share
You just described a thing of mine I cannot help but do; explain the ever loving crap out of things
I need to be careful with that though as relatives start to complain and push back on me telling things over and over.
Thing is, until I see a full comprehension on the other side on what I try to convey I just keep explaining in variations, keep finding metaphors and keep pestering you until you ‘get it’. Some say it is a virtue, some say it is a hindrance.
I have had therapy on this… 😂
Nginx for my intranet because configuration is fully manual and I have complete control over it.
Caddy for the public services on my vps because it handles cert renewal automatically and most of its configuration is magic which just works.
It is unbelievable how shorter caddy configuration is, but on my intranet:
I switched to caddy just for the certs. I get trusted certs on all my internal subdomains without maintenance.
I use haproxy, nginx and caddy at work including a caddy instance with internal CA. 4 lines in config and its signed by our normal CA, so its trusted by all our devices.
it seems easier to manage stuff not in docker
Read into Traefik’s dynamic configuration. Adding something outside of Docker is as easy as adding a new config file in the dynamic configuration folder. E.g. jellyfin.yml:
The moment you save that file it will be active and working in Traefik.
I mean, the basic config file for Caddy is 1 line, and gives you Let’s Encrypt by default. The entire config file for a reverse proxy can be as few as 3 lines:
my.servername.net { reverse_proxy 127.0.0.1:1234 }It’s a single executable, and a single 3-line file. Caddy is an incredible piece of software.
When I was researching reverse proxies I first stumbled upon nginx and traefik and especially nginx seemed a bit intimidating. As someone who hadn’t done it before I was worried if I’d do it right. Then I found caddy and yeah just used a threeliner like that in config and that was that. Simple and easy to get it right.
I’ve since switched to having my stuff behind wireguard instead of reverse proxy, but I keep caddy around so I can just spin it back up if I want to access Jellyfin on someone’s tv or something.
but I’d like to give Nginx Proxy Manager a try, it seems easier to manage stuff not in docker.
NPM is pretty agnostic. If it receives a request for a specific address and port combination it just forwards the traffic to another specific address and port combination. This can be a docker container, but also can be a physical machine or any random URL.
It also has Let’s Encrypt included (but that should be a no-brainer).
I use and love nginx.
Maybe a bit more old fashioned than more modern solutions, but steady solid and versatile. I use it as reverse proxy ad well as proxy for php stuff and more.
I use Traefik at home. The initial setup was more complex than others but now it’s set up it’s by far the easiest to add new routes than any other I’ve tried, just by virtue of being right there in the compose/k8s files I’m already writing. Static routes are manual annoyingly, but so are every other proxy so that’s no different. The config files are the same markup language as your Compose/k8s files so you’re not learning a whole new syntax and having to switch languages mentally as you switch between them.
Caddy is super easy, but the fact that the Docker labels thing was a plugin is a con to me, I’d prefer it being first party. It also isn’t as performant as Traefik, higher CPU usage while also having higher latency.
As far as I’m aware, Nginx and Nginx Proxy Manager support no such thing, you have to manually write those routes every time you create a new service.
Nginx Proxy Manager is a lot like Portainer. It’s useful for people who don’t want to learn Nginx and just want to click a few buttons. But anything complex you’re suddenly going to be thrown into the deep end.
You’ve already set up Traefik, you’ve already done the complex bit. IMO there’s no reason to change, from this point everything else is more complicated.
Caddy is the only reverse proxy I have ever managed to successfully make use of. I failed miserably with Nginix and Traefik.
Caddy has worked very well for me for several years now. It gets the SSL certificate from my domain name provider and all.
I’ve been mostly using Nginx Proxy Manager, but I recently set up Bunkerweb as a WAF for a couple of public services I’m hosting and I kind of like it. It does reverse proxy along with a bunch of other things (bad behavior blocking, geographic blocking, SSL cert handling, it does a lot).
Mentioning it because I didn’t see any other mention of it yet.
NPM is easy to use. Caddy sounds like something I’d like to try too now.
I use nginx for static websites and TLS passthrough servers.
I use traefik as a reverse proxy for sites with many services and SSO.
Nginx is definitely easier to configure for simple things. But I prefer traefik for more complex setups.
Simple question but can be a complex answer. Basically it depends where your phone gets DNS from: if it’s using the ISP DNS (or some other public DNS server) it will resolve the public internet IP of your server and the data will route out to the ISP WAN before being routed back in.
On the other hand you can configure a split DNS system, so say you are using your modem/gateway as your DNS server and it forwards DNS queries up to your ISP (or other) DNS server - a common setup, 1. you can add in a static host entry for your local server. Eg ‘yourservice.yourserverdomain.com = 192.168.1.20 (your server’s LAN IP)’
Now when your phone is on the WiFi and it looks up your server’s address it gets the local IP and routes locally, which will be faster.
If you need more info, search for terms like ‘reverse proxy split DNS best practice’.
I’ve looked at it but never actually given the Synology proxy a go despite using their DNS server. Does it do auto certificate renewal?
Have you considered using a Cloudflare tunnel to bypass the CGNAT? You can do that into a proxy or straight into the service.
Does it do auto certificate renewal?
Yes.
Have you considered using a Cloudflare tunnel to bypass the CGNAT?
I did before when I had some free domain over there, but I don’t think there are any worthy free domains out there anymore, and even when they are cheap, I really don’t need it and don’t feel comfortable to pay for something that I can’t use in its fullest (due to CGNAT).
For example, I am aware cloudflare tunnels can’t be used for a Plex/Video streaming and that is the number 1 service that I want to be exposed to the Internet.
For now I am living with my IPv6 address and the Synology DDNS with the reverse proxy features… My personal fallback are Tailscale and Zerotier.
Traefik is a PITA.
Caddy all the way. If you build it with Docker support (or grab the prebuilt), you can use docker container names to reverse proxy using names instead of any IP addresses or ports. It’s nice because if the IP updates, so does caddy. All automatically.
Here’s what my caddyfile looks like;
{ acme_dns cloudflare {key} } domain.dev { encode zstd gzip root * /var/www/html/domain.dev/ php_fastcgi unix//run/php/php8.1-fpm.sock tls { dns cloudflare {key} } } *.domain.dev { encode zstd gzip tls { dns cloudflare {key} } @docker host docker.domain.dev handle @docker { encode zstd gzip reverse_proxy {portainer} } @test host test.domain.dev handle @test { encode zstd gzip reverse_proxy 127.0.0.1:10000 } @images host i.domain.dev handle @images { encode zstd gzip reverse_proxy 127.0.0.1:9002 } @proxy host proxy.domain.dev handle @proxy { encode zstd gzip reverse_proxy proxy } @portal host portal.domain.dev handle @portal { encode zstd gzip reverse_proxy portal } @ping host ping.domain.dev handle @ping { encode zstd gzip respond "pong!" } }DNS hosted by cloudflare but because caddy handles ACME certs, all the subdomains automatically get SSL.
Actually I found traefik rather easy, I just had to make the proper docker labels and config.
PITA
Unrelated, I’m going to sound like a grammar nazi here, but holy shit there are so many acronmys, how am I supposed to know every one of them without googling? Please just say “traefik is a pain in the ass”. Also please don’t take this as a snarky reply.
PITA = pain in the ass.
I never said it was hard. Just a real pain in the ass. Like iptables vs UFW. They’re the same thing, but one is easy and a pain in the ass and the other is just easy… So I opt to make my life easier. lol