go-mmproxy is a standalone application that unwraps HAProxy's PROXY protocol (also adopted by other projects such as NGINX) so that the network connection to the end server comes from client's - instead of proxy server's - IP address and port number.
go-mmproxy is deployed on the virtual machine by default. This project deploys go-mmproxy on Google Cloud GKE, so that the Pod deployed behind the TCP Proxy LB can get the real IP of the client.
kubectl apply -f nginx.yaml
kubectl apply -f nginx-neg.yaml
gcloud compute health-checks create tcp my-tcp-health-check --port 8080
gcloud compute backend-services create my-tcp-lb \
--global-health-checks \
--global \
--protocol TCP \
--health-checks my-tcp-health-check \
--port-name tcp8080
gcloud compute backend-services add-backend my-tcp-lb \
--global \
--network-endpoint-group=my-nginx-neg \
--network-endpoint-group-zone=us-central1-a \
--balancing-mode CONNECTION \
--max-connections 100
gcloud compute target-tcp-proxies create my-tcp-lb-target-proxy \
--backend-service my-tcp-lb \
--proxy-header PROXY_V1
gcloud compute forwarding-rules create my-tcp-lb-ipv4-forwarding-rule \
--global \
--target-tcp-proxy my-tcp-lb-target-proxy \
--ports 80
Note: Wait for 5 minutes after the LB is created before starting the test
curl http://34.160.48.200/
replace 34.160.48.200 to your tcp proxy lb ip
you can see that this ip is the client ip not tcp proxy lb ip.
gcloud compute forwarding-rules delete my-tcp-lb-ipv4-forwarding-rule --global --quiet
gcloud compute target-tcp-proxies delete my-tcp-lb-target-proxy --global --quiet
gcloud compute backend-services delete my-tcp-lb --global --quiet
gcloud compute health-checks delete my-tcp-health-check --quiet
kubectl delete -f nginx.yaml
kubectl delete -f nginx-neg.yaml
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent