Disclaimers:
First thing first, I'm new to the whole Fediverse, and Lemmy thing, so please don't hesitate to point out any problems you're foreseeing.
Secondly, I'm by no means saying this is the ideal implementation, something something see above. Please don't hesitate to make recommendations for improvements.
Lastly, I'm not sure if it is completely working. I'm still noticing a few issues that I will document and monitor towards the end of the post. If you know of the cause or how to debug further, please do let me know!
Notes and Assumptions:
- I am using an ARM server. So I'm using ARM images, you will need to make sure you're using the correct architecture image.
- I assume you have Traefik up and running in a separate network. I used docker compose to bring traefik up, minimal configurations, and I'm just hijacking the
default network there (project folder was gateway so the complete network name is gateway_default)... there's probably better ways to do this.
- On note of networks, I really don't like the fact that the default postgres was left wide open on the
lemmyexternalproxy network. I think I've locked my down, but you may wish to double check my work.
- I'm not sure if what I am doing with the hostnames are correct, but it seems to work for the most part, so I'm not complaining. If there is a better way, please do advise!
- I used an override file for docker compose to apply extra settings. This allows me to keep the original
docker-compose.yml untouched, and I can just pull in new changes (theoretically).
- Since I'm using traefik, I don't need nginx running doing nothing. I replaced it with a light weight alpine image that just shuts down successfully, so it doesn't use resources.
Without further delays, here's my files:
docker-compose.override.yml:
version: "3.3"
networks:
lemmyexternalproxy:
internal: true
lemmygateway:
name: gateway_default
external: true
services:
lemmy:
image: dessalines/lemmy:0.17-linux-arm64
labels:
- "traefik.enable=true"
- "traefik.http.routers.lemmy.entrypoints=websecure"
- "traefik.http.routers.lemmy.rule=Host(`lemmy.chiisana.net`) && HeadersRegexp(`Accept`, `^application/`) || Host(`lemmy.chiisana.net`) && Method(`POST`) || Host(`lemmy.chiisana.net`) && PathPrefix(`/{path:(api|pictrs|feeds|nodeinfo|.well-known)}`)"
- "traefik.http.routers.lemmy.tls=true"
- "traefik.http.services.lemmy-svc.loadbalancer.server.port=8536"
- "traefik.docker.network=gateway_default"
networks:
- lemmygateway
lemmy-ui:
image: dessalines/lemmy-ui:0.17-linux-arm64
environment:
- LEMMY_UI_HOST=0.0.0.0:1234
- LEMMY_UI_LEMMY_INTERNAL_HOST=lemmy:8536
- LEMMY_UI_LEMMY_EXTERNAL_HOST=lemmy.chiisana.net
- LEMMY_UI_HTTPS=true
- LEMMY_UI_DEBUG=false
labels:
- "traefik.enable=true"
- "traefik.http.routers.lemmy-ui.entrypoints=websecure"
- "traefik.http.routers.lemmy-ui.rule=Host(`lemmy.chiisana.net`)"
- "traefik.http.routers.lemmy-ui.tls=true"
- "traefik.http.services.lemmy-ui-svc.loadbalancer.server.port=1234"
- "traefik.docker.network=gateway_default"
networks:
- lemmygateway
proxy:
image: alpine:latest
command: "true"
entrypoint: "true"
restart: "no"
pictrs:
image: asonix/pictrs:0.4.0-rc.3
lemmy.hjson:
setup: {
admin_username: "chiisana"
admin_password: "password-redacted-duh"
site_name: "chiisana lemmy site"
}
database: {
host: "postgres"
user: "lemmy"
password: "password-redacted-duh"
database: "lemmy"
}
email: {
smtp_server: "smtp.mailgun.org:587"
smtp_login: "lemmy@chiisana.net"
smtp_password: "password-redacted-duh"
smtp_from_address: "lemmy@chiisana.net"
tls_type: "tls"
}
pictrs: {
url: "http://pictrs:8080/"
api_key: "API_KEY"
}
hostname: "lemmy.chiisana.net"
bind: "0.0.0.0"
port: 8536
tls_enabled: true
}
Known issue(s)?
- ~~I have my registration disabled as the instance is supposed to be just for my own auth not be depended on other instances. In my
/admin section, I'm seeing a ton of users from endlesstalk.org pop up as banned users. I have no idea what that is about, as endlesstalk.org seems to also be used only by one user. I'll be monitoring this and see what's to come of it.~~ Edit: Looks like this is just the way the system is designed, and not a configuration error on my part! All good here. Thanks for clarifying it @lemmy@endlesstalk.org !
- I'm not sure if I'm getting all the messages federated. In this community, for example, I can see most if not all recent threads. However, most threads have no comments in it. Some newer threads, I see comments, but it seems to be incomplete. I'm not sure if I'm only supposed to receive new messages, or if something else is happening. I'll be monitoring this, and hoping the federation will just catch up over time.
- Edit: It would appear this post itself is not federating to !selfhosted@lemmy.world for some reason... I'm partially hoping it is just caught in some kind of moderation queue, but seeing other posts made after this appear on the list leads me to believe there's still something amiss.
If you encounter any other issue, please do post back so we can try to debug it together. Hope this helps someone!
The amount of confidently incorrect responses is exactly what one could expect from Lemmy.
First: TCP and UDP can listen on the same port, DNS is a great example of such. You’d generally need it to be part of the same process as ports are generally bound to the same process, but more on this later.
Second: Minecraft and website are both using TCP. TCP is part of layer 4, transport; whereas HTTP(S) / Minecraft are part of layer 7, application. If you really want to, you could cram HTTP(S) over UDP (technically, QUIC/HTTP3 does this), and if you absolutely want to, with updates to the protocol itself, and some server client edits you can cram Minecraft over UDP, too. People need to brush up on their OSI layers before making bold claims.
Third: The web server and the Minecraft server are not running on the same machine. For something that scale, both services are served from a cluster focused only on what they’re serving.
Finally: Hypixel use reverse proxy to sit between the user and their actual server. Specifically, they are most likely using Cloudflare Spectrum to proxy their traffic. User request reaches a point of presence, a reverse proxy service is listening on the applicable ports (443/25565) + protocol (HTTPS/Minecraft), and then depending on traffic type, and rules, the request gets routed to the actual server behind the scenes. There are speculations of them no longer using Cloudflare, but I don’t believe this is the case. If you dig their mc.hypixel.net domain, you get a bunch of direct assigned IP addresses, but if you tried to trace it from multiple locations, you’d all end up going through Cloudflare infrastructure. It is highly likely that they’re still leaning on Cloudflare for this service, with a BYOIP arrangement to reduce risk of DDOS addressed towards them overflow to other customers.
In no uncertain terms:
mc.hypixel.net, but also have a SRV record for_minecraft._tcp.hypixel.netset for 25565 onmc.hypixel.netmc.hypixel.netdomain has CNAME record formt.mc.production.hypixel.io.which is flattened to a bunch of their own direct assigned IP addresses.