I’m not trying to do anything fancy with Nginx Proxy Manager (NPM)… yet. No SSL, no self-referencing HA redirects. Just trying to test it with another service (Plex).
This is one of the simplest use-cases I could imagine, but I can’t make it work. Every service I test gives me 504 Gateway Time-out. I’ve tried setting the log level to trace, it doesn’t provice much info:
2022/01/08 22:00:13 [error] 738#738: *1087 upstream timed out (110: Operation timed out) while connecting to upstream, client: 192.168.4.131, server: plex.local, request: "GET / HTTP/1.1", upstream: "http://192.168.4.20:32400/", host: "plex.local"
[08/Jan/2022:22:00:13 +1100] - 504 504 - GET http plex.local "/" [Client 192.168.4.131] [Length 562] [Gzip -] [Sent-to 192.168.4.20] "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/96.0.4664.110 Safari/537.36" "-"
plex.local is just a placeholder, I have a working DNS server with FQDN
In order to try and diagnose this further, I’ve spend the last 2 days getting root ssh access to the HA OS, and a bash shell in the NPM docker container.
I don’t have much understanding of docker networking, so I may be off the mark, but this looks like a networking issue to me. The docker container (which has a 172.30.x.x address) cannot ping the target machine 192.168.4.20. It can ping the HA OS (docker host) 192.168.4.19.
HA is running in a VM, and AFAIK I haven’t done anything fancy to either the HA networking or addons. So I’m not sure why I can’t get the NPM addon working.
The NPM docker container is connected to the hassio network, which is a bridge type. I would have expected the network type host, can anyone confirm NPM works with bridge for them?
Looking in the NPM container, I see the following:
root@a0d7b954-nginxproxymanager:/$ route
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
default 15ef4d2f-esphom 0.0.0.0 UG 0 0 0 eth0
172.30.32.0 * 255.255.254.0 U 0 0 0 eth0
WHY is the default gateway ESPHome? Technically the ESPHome container name is “addon_15ef4d2f_esphome”. I’m not sure where “15ef4d2f-esphom” is coming from, and I can’t make BusyBox route command show the full name.
Looks like this isn’t an HA or NPM issue. It’s all Synology VMM.
From within the docker container I can ping other devices on my local network, and I can reach internet. The only address I can’t reach is the Synology NAS.
Ditto from HA OS directly. 192.168.4.19 can’t ping 192.168.4.20. Probably because the VM network adapter (.19) is bonded to the NAS network adapter (.20)