Nginx can't resolve hostname because of hardcoded dns resolver in nginx.conf

[nyarau]

Currently a hard coded resolver address 127.0.0.11 is set in nginx.conf
It works in typical Docker setup, however it fails in setups using Podman with an implicit IPAM

piped-nginx     | 2023/10/05 09:33:20 [error] 31#31: recv() failed (111: Connection refused) while resolving, resolver: 127.0.0.11:53
piped-nginx     | 2023/10/05 09:33:24 [error] 30#30: *1 piped could not be resolved (110: Operation timed out), client: 10.11.8.1, server: piped-api.example.com, request: "GET /channel/UCBR8-60-B28hp2BmDPdntcQ HTTP/1.1", host: "piped-api.example.com"

---

Referring to https://stackoverflow.com/a/31864024 , I removed resolver line in nginx.conf and used the following config as pipedapi.conf, which indeed fixed the issue.

proxy_cache_path /tmp/pipedapi_cache levels=1:2 keys_zone=pipedapi:4m max_size=2g inactive=60m use_temp_path=off;

upstream pipedapi {
    server piped:8080;
}
server {
    listen 80;
    server_name piped-api.example.com;

    location / {
        proxy_cache pipedapi;
        proxy_pass http://pipedapi;
        proxy_http_version 1.1;
        proxy_set_header Connection "keep-alive";
    }
}

[Bnyro]

I think that's a reasonable change, can you please create a pull request?