Hello! I have been struggling through a few tutorials on getting a lemmy instance to work correctly when setup with Docker. I have it mostly done, but there are various issues each time that I do not have the knowledge to properly correct. I am familiar with Docker, and already have an Oracle VPS set up on ARM64 Ubuntu. I already have portainer and an NGINX proxy set up and working okay. I have an existing lemmy instance “running” but not quite working. My best guess here would be to have someone assist with setting up the docker-compose to work with current updates/settings, as well as the config.hjson.
TIA, and I cant wait to have my own entry into the fediverse working right!
Hey, if you still feel like helping out :D
Ive been through a boatload of changes today since earlier. Ive rebuilt using mostly the provided yml in the official guide, and after some tweaking, almost everything is working. The internal proxy is now working, and the containers are working amongst themselves fully as far as I can tell. I do not know how to setup a web facing reverse proxy in a way that works around the internal proxy already running (other than the already in place NPM). I turned the NPM back on, and was able to get it working to reach the site, however I cannot reach any other communities from within my site. I believe the reverse proxy NPM is just not set up right. Error message in lemmy:
spoiler
ERROR HTTP request{http.method=GET http.scheme="https" http.host=lemmy.bulwarkob.com http.target=/api/v3/ws otel.kind="server" request_id=69004ca6-7967-48c3-a4d2-583e961e34d3 http.status_code=101 otel.status_code="OK"}: lemmy_server::api_routes_websocket: couldnt_find_object: Request error: error sending request for url (https://midwest.social/.well-known/webfinger?resource=acct:projectzomboid@midwest.social): operation timed out 0: lemmy_apub::fetcher::search::search_query_to_object_id at crates/apub/src/fetcher/search.rs:17 1: lemmy_apub::api::resolve_object::perform with self=ResolveObject { q: "[email protected]", auth: Some(Sensitive) } at crates/apub/src/api/resolve_object.rs:21 2: lemmy_server::root_span_builder::HTTP request with http.method=GET http.scheme="https" http.host=lemmy.bulwarkob.com http.target=/api/v3/ws otel.kind="server" request_id=69004ca6-7967-48c3-a4d2-583e961e34d3 http.status_code=101 otel.status_code="OK" at src/root_span_builder.rs:16
I would be happy to remove NPM from this stack if its not too difficult to get a correctly working reverse proxy set up. The documentation doesnt give much to work with in it.
from the log it seems that lemmy cannot reach https://midwest.social/ - if you have more such operation timed outs - probably there is some networking issue with outgoing requests - maybe you have some kind of firewall? i can reach your instance from other direction: https://group.lt/c/[email protected]
probably the easiest way to setup lemmy and another front facing reverse proxy is to use nginx that comes with lemmy on another port and setup simple reverse proxying with NPM to it. i myself using caddy for reverse proxying, using this config: https://join-lemmy.org/docs/en/administration/caddy.html
I see that the instance can be reached, and posts are shown, however comments are not. I have found in the official docs that there is a config snippet for a web facing reverse proxy. https://join-lemmy.org/docs/en/administration/troubleshooting.html https://github.com/LemmyNet/lemmy-ansible/blob/main/templates/nginx.conf
And this config appears quite different from the “Install with Docker” config instructions: https://join-lemmy.org/docs/en/administration/install_docker.html
spoiler
worker_processes 1; events { worker_connections 1024; } http { upstream lemmy { # this needs to map to the lemmy (server) docker service hostname server "lemmy:8536"; } upstream lemmy-ui { # this needs to map to the lemmy-ui docker service hostname server "lemmy-ui:1234"; } server { # this is the port inside docker, not the public one yet listen 80; # change if needed, this is facing the public web server_name localhost; server_tokens off; gzip on; gzip_types text/css application/javascript image/svg+xml; gzip_vary on; # Upload limit, relevant for pictrs client_max_body_size 20M; add_header X-Frame-Options SAMEORIGIN; add_header X-Content-Type-Options nosniff; add_header X-XSS-Protection "1; mode=block"; # frontend general requests location / { # distinguish between ui requests and backend # don't change lemmy-ui or lemmy here, they refer to the upstream definitions on top set $proxpass "http://lemmy-ui"; if ($http_accept = "application/activity+json") { set $proxpass "http://lemmy"; } if ($http_accept = "application/ld+json; profile=\"https://www.w3.org/ns/activitystreams\"") { set $proxpass "http://lemmy"; } if ($request_method = POST) { set $proxpass "http://lemmy"; } proxy_pass $proxpass; rewrite ^(.+)/+$ $1 permanent; # Send actual client IP upstream proxy_set_header X-Real-IP $remote_addr; proxy_set_header Host $host; proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; } # backend location ~ ^/(api|pictrs|feeds|nodeinfo|.well-known) { proxy_pass "http://lemmy"; # proxy common stuff proxy_http_version 1.1; proxy_set_header Upgrade $http_upgrade; proxy_set_header Connection "upgrade"; # Send actual client IP upstream proxy_set_header X-Real-IP $remote_addr; proxy_set_header Host $host; proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; } } }
Do you know if I should expect to have TWO unique NGINX proxy instances (assuming I use NGINX)? One in-stack, and one separate for web facing reverse proxy? Or do I need a combination of the two configs into one instance?
I am going to see if I can get a caddy reverse proxy setup in the meantime and see how it performs given your configuration there.
you can have two nginx proxy instances, one as a front (serving other sites besides lemmy instance) and another - coupled with lemmy instance. in such case the first one can be configured minimally with basic proxy stuff to internal lemmy one, no need for this fancy lemmy and lemmy-ui proxying.
I believe I have the simple set up for the NPM reverse proxy. Just as you say, it points to the docker address of the lemmyInstance NGINX. I can get to my instance with HTTPS secured and good. Just getting errors when communicating to other instances. I can only imagine it is web socket related, but I am not familiar enough to look at the proxy configs and determine what might be wrong with it unfortunately. I might need to try and find someone to essentially look through it with me in real time.
spoiler
ERROR HTTP request{http.method=GET http.scheme="https" http.host=lemmy.bulwarkob.com http.target=/api/v3/ws otel.kind="server" request_id=67d75886-bf48-4444-a435-d98d8fc1e303 http.status_code=101 otel.status_code="OK"}: lemmy_server::api_routes_websocket: couldnt_find_object: Request error: error sending request for url (https://lemmy.ml/.well-known/webfinger?resource=acct:asklemmy@lemmy.ml): operation timed out 0: lemmy_apub::fetcher::search::search_query_to_object_id at crates/apub/src/fetcher/search.rs:17 1: lemmy_apub::api::resolve_object::perform with self=ResolveObject { q: "[email protected]", auth: Some(Sensitive) } at crates/apub/src/api/resolve_object.rs:21 2: lemmy_server::root_span_builder::HTTP request with http.method=GET http.scheme="https" http.host=lemmy.bulwarkob.com http.target=/api/v3/ws otel.kind="server" request_id=67d75886-bf48-4444-a435-d98d8fc1e303 http.status_code=101 otel.status_code="OK" at src/root_span_builder.rs:16
spoiler
WARN Error encountered while processing the incoming HTTP request: lemmy_server::root_span_builder: Request error: error sending request for url (https://beehaw.org/u/Jattatak): operation timed out 0: lemmy_server::root_span_builder::HTTP request with http.method=POST http.scheme="https" http.host=lemmy.bulwarkob.com http.target=/inbox otel.kind="server" request_id=f413d3e5-262a-4dac-bc2e-700b9a053954 http.status_code=400 otel.status_code="OK" at src/root_span_builder.rs:16 LemmyError { message: None, inner: Request error: error sending request for url (https://beehaw.org/u/Jattatak): operation timed out Caused by: 0: error sending request for url (https://beehaw.org/u/Jattatak): operation timed out 1: operation timed out, context: "SpanTrace" }
from the logs it seems that lemmy docker does not communicate with outside servers.
also i have a bit different config for lemmy.hjson
{ # for more info about the config, check out the documentation # https://join-lemmy.org/docs/en/administration/configuration.html setup: { # username for the admin user admin_username: "adminuser" # password for the admin user admin_password: "adminpassword" # name of the site (can be changed later) site_name: "group.lt" } opentelemetry_url: "http://otel:4317" # the domain name of your instance (eg "lemmy.ml") hostname: "group.lt" # address where lemmy should listen for incoming requests bind: "0.0.0.0" # port where lemmy should listen for incoming requests port: 8536 # settings related to the postgresql database # address where pictrs is available pictrs: { url: "http://pictrs:8080/" # api_key: "API_KEY" } database: { # name of the postgres database for lemmy database: "lemmy" # username to connect to postgres user: "lemmy" # password to connect to postgres password: "lemmy" # host where postgres is running host: "postgres" # port where postgres can be accessed port: 5432 # maximum number of active sql connections pool_size: 5 } # # optional: email sending configuration email: { # # hostname and port of the smtp server smtp_server: "postfix:25" smtp_from_address: "[email protected]" tls_type: false } }
also check in admin interface if federation is enabled and you do not blacklist instances
(https://lemmy.bulwarkob.com/admin) and maybe you can try to enable federation debug mode for awhile
The differences I see are the otel link, and the TLS setting:
# Whether the site is available over TLS. Needs to be true for federation to work. tls_enabled: true
I see you dont have it on there, which I would assume means you cant be federated? I have added the otel link and enabled the debug mode. Federation is already enabled and the instance is set to “ALL”. Still no luck on this end. Same status, except now im not getting any log errors in the container logs (Viewed from Portainer).
Including this in case it is a possible issue: federation enabled, host is lemmy.bulwarkob.com
Starting http server at 0.0.0.0:8536
also pictrs: { url: “http://pictrs:8080/” # api_key: “API_KEY” }
about tls setting - don’t remember why i have removed it, but group.lt federates fine. not sure about what you mean instance set to ALL.
what about network isolation in portainer? maybe it is on?
I see my Pictrs appears to be the same as what you had sent over. Protainer network isolation does not appear to be in place. All are bridged networks, and I would assume access issues would be more encompassing if that were a direct correlation to the issue. Im still betting on User Error for configuration so far. Being myself, of course.