/nginx_failover_poc

Proof of concept to see how Nginx can handle failing upstream servers

Primary LanguageNginx

Experiment with Nginx to handle failing backend servers

Create a Docker network

docker create network nut

Build images for both backend server and nginx:

cd server && docker build -t server .
cd nginx && docker build -t nginx .

Start services:

docker run --rm --name c1 --network nut -e name=c1 server rackup -E production
docker run --rm --name c2 --network nut -e name=c2 server rackup -E production
docker run --rm --name nginx --network nut -p 80:80 nginx

If you kill c1 (the main backend) with:

docker stop c1

You'll notice requests go yo c2.