我具有以下设置,并且终生无法弄清为什么我无法连接到api。
nginx.conf
worker_processes auto;
worker_rlimit_nofile 200000;
events {
use epoll;
accept_mutex on;
multi_accept on;
worker_connections 1024;
}
http {
error_log /etc/nginx/error_log.log warn;
client_max_body_size 20m;
server {
listen 80;
listen [::]:80;
location / {
proxy_pass http://api:8080/;
}
location /health {
return 200;
access_log off;
}
}
}
docker-compose.yml
version: "3.7"
services:
nginx:
container_name: nginx
image: "nginx:latest"
ports:
- "8000:80"
networks:
- internal_net
volumes:
- ./container/nginx.conf:/etc/nginx/nginx.conf
api:
container_name: api
build:
context: .
dockerfile: container/Dockerfile
expose:
- "8080"
networks:
- internal_net
depends_on:
- postgres
command: ["./wait-for-it.sh", "postgres:5432", "--timeout=60", "--", "./52-server-go"]
postgres:
container_name: postgres
image: "postgres:9.5-alpine"
expose:
- "5432"
networks:
- internal_net
networks:
internal_net:
driver: bridge
volumes:
container:
我的api设置为在端口8080上运行,当我进入容器并对该地址运行curl请求时,它可以工作。根据撰写文件,应该将该地址公开给包括nginx在内的所有服务共享的本地撰写网络。
根据nginx配置,它应该将每个请求(有效的/health
除外)传递给api
服务。相反,返回的是nginx的502。
我在哪里混在一起?我做错了什么?
答案 0 :(得分:0)
问题实际上出在我的go应用中。使用golang gorilla/mux
,必须更改地址:
原始(损坏)
// Start server
address := "127.0.0.1:8080"
srv := &http.Server{
Handler: r,
Addr: address,
WriteTimeout: 15 * time.Second,
ReadTimeout: 15 * time.Second,
}
修复
// Start server
address := "0.0.0.0:8080"
srv := &http.Server{
Handler: r,
Addr: address,
WriteTimeout: 15 * time.Second,
ReadTimeout: 15 * time.Second,
}
我不知道为什么 ,但这已解决。