Can't reach other containers via hostname

[maurxce]

Hey, I've been wanting to cut back on exposed ports and thus wanted to use hostnames in my proxy hosts.

I have encountered one problem with hostnames so far:

$ docker exec <pingvin id> ping nginx
PING nginx (172.21.0.2): 56 data bytes
64 bytes from 172.21.0.2: seq=0 ttl=64 time=0.102 ms

$ docker exec <nginx id> ping pingvin
OCI runtime exec failed: exec failed: unable to start container process: exec: "ping": executable file not found in $PATH: unknown
$ docker exec <nginx id> curl pingvin:3000
curl: (7) Failed to connect to pingvin port 3000 after 1 ms: Couldn't connect to server

Edit: with ip's:

# vv nginx ip
root@dev:~# docker exec <pingvin id> ping 172.21.0.2
PING 172.21.0.2 (172.21.0.2): 56 data bytes
64 bytes from 172.21.0.2: seq=0 ttl=64 time=0.222 ms

# vv pingvin ip
root@dev:~# docker exec <nginx id> curl 172.21.0.3:3000
curl: (7) Failed to connect to 172.21.0.3 port 3000 after 0 ms: Couldn't connect to server

Same issue appears when trying to curl :80 instead. I find it pretty weird that one container can reach the other, but not the other way around.

My compose.yml:

services:
  nginx:
    image: jc21/nginx-proxy-manager:2.11.0
    restart: unless-stopped
    ports:
      - 80:80
      - 81:81
      - 443:443
    volumes:
      - ./data:/data
      - ./letsencrypt:/etc/letsencrypt
    networks:
      - proxy

networks:
  proxy:
    driver: bridge

services:
  pingvin:
    image: stonith404/pingvin-share:v0.21.5
    restart: unless-stopped
    volumes:
      - ./data:/opt/app/backend/data
      - ./public:/opt/app/frontend/public
    networks:
      - nginx_proxy   

networks:
  nginx_proxy:
    external: true

Can someone help me figure this out?

[lastsamurai26]

First issue what is see is that your npm network is proxy not nginx_proxy so pleae put your pingvin container to the proxy network ;)

[maurxce]

it would make sense to be the config, but it's almost exactly the same
I ruled out the specific npm version too, using :latest didn't do anything

Using an identical network setup didn't work, and I doubt it's caused by the different order of ports or the volumes

[lastsamurai26]

strange :-( i'm at the end of my Latin

[maurxce]

It seems like my pingvin:v0.21.5 image was faulty (on my pc at least, haven't tried server).
Did a quick docker image rm and then redownloaded it, seems to work so far

That would also explain why your config worked and mine didn't. You didn't add a specific version, so it ended up using a different image

[maurxce]

I did actually try every version from 0.20.0 to 0.21.5, so the actual issue turns out to be that the files generated by 0.21.5 are corrupted or something. Upgrading from 0.21.0 to 0.21.5 works fine.

I'll close this discussion now

[gremo]

I have the exact same problem with any other container connected to NPM using an internal network: works with IP, not working with the container name (i.e. portainer).