glindsell / grpc-bidi-streaming

Routeguide-app

Description

This app demonstrates a GRPC server and client communicating using bidirectional streaming in a single node k8s cluster, with load balancing attempted between three pods using Linkerd - LOAD BALANCING FAILS.

Run the app

Follow the instructions to install minikube:
https://kubernetes.io/docs/tasks/tools/install-minikube/

Start minikube:
minikube start

Set docker env:
eval $(minikube docker-env)

Build docker image:
docker build -t routeguide-app .

Run local image as client in minikube:
kubectl run routeguide-app-client --image=routeguide-app:latest --port=10000 --image-pull-policy=Never

Check that it's running:
kubectl get pods

Install Linkerd:
curl -sL https://run.linkerd.io/install | sh

Add linkerd to your path:
export PATH=$PATH:$HOME/.linkerd2/bin

Verify the CLI is installed and running correctly:
linkerd version

To check that your cluster is configured correctly and ready to install the control plane, you can run:
linkerd check --pre

Install the lightweight control plane into its own namespace (linkerd):
linkerd install | kubectl apply -f -

Validate that everything’s happening correctly:
(This command will patiently wait until Linkerd has been installed and is running.)
linkerd check

In a new terminal view the Linkerd dashboard by running:
linkerd dashboard

Back in terminal view deployments:
kubectl -n linkerd get deploy

Install the app as a new deployment (using docker hub image at glindsell/helloworld-app):
linkerd inject routeguide-app-grpc.yml | kubectl apply -f -

View new deployment in the Linkerd dashboard.

Expose the deployment as a service:
kubectl expose deployment routeguide-app-server --type=NodePort

Get IP Address of services:
kubectl get services

Exec into client:
kubectl exec -it routeguide-app-client-<id-goes-here> -- /bin/bash

cd into app dir:
cd src/routeguide-app/

Change to CLUSTER-IP of helloworld-app-server (line 32 of helloworld-app/greeter_client):
vim client/client.go
change:
address = "<IP-ADDR>:10000"

Run the client to send requests:
go run client/client.go

Observe that no load balancing occurs over server pods.

Click on Grafana icon to view individual pod’s stats, and view pod logs to confirm.

To bring everything down:
minikube stop
minikube delete

Note... TLS is available via:

go run server/server.go -tls=true

and

go run client/client.go -tls=true