Tutorial: First Look at Dapr for Microservices Running in Kubernetes
In a previous article, I introduced the architecture and building blocks of Dapr, a portable, event-driven runtime for distributed systems originally developed by Microsoft. To appreciate the platform, let’s zoom into the state management building block of Dapr. This hands-on guide will walk you through all the steps involved in dealing with state management in Dapr.
Background
We are going to deploy two microservices written in Node.js and Python in Kubernetes. The services will use Redis as the persistence layer to store the state. Since we use Dapr, we will swap out Redis with etcd while continuing to run the microservices.
To complete this tutorial, you need a Kubernetes cluster running within Minikube or a managed service such as the Azure Kubernetes Service (AKS).
I am running Minikube on my development machine.
Installing Dapr
Download the Dapr CLI for your OS from the releases page of GitHub repository, rename and add the binary to the path.
Run the below command to install Dapr in your Kubernetes environment.
|
1 |
dapr init --kubernetes |
The installer deploys a few Pods in the default Namespace that are a part of Dapr control plane. Like the service mesh, Dapr has a control plane that integrates with Kubernetes and a data plane that runs as a sidecar inside each Pod.
The dapr-operator Pod watches for Pods that have the Dapr Annotations. The dapr-sidecar-injector Pod is responsible for adding the sidecar container to each Pod annotated as “dapr.io/enabled: true”. Finally, the dapr-placement Pod manages the communication across all the sidecar containers injected into the Pods.
Configuring Redis as the Persistence Layer
Since we are dealing with the state store, the next step is to deploy Redis and configuring it as the default state store for the microservices.
Deploy Redis by submitting the below YAML file to Kubernetes. This results in the creation of a Pod and a ClusterIP Service.
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 |
apiVersion: v1 kind: Service metadata: name: redis labels: app: redis spec: ports: - port: 6379 name: redis targetPort: 6379 selector: app: redis --- apiVersion: apps/v1 kind: Deployment metadata: name: redis spec: selector: matchLabels: app: redis replicas: 1 template: metadata: labels: app: redis spec: containers: - name: redis-server image: redis:3.2-alpine |
|
1 |
kubectl apply -f redis.yaml |
With the Redis Pod up and running, let’s configure it as a Dapr state store. Create a YAML file with the below specification and apply it.
|
1 2 3 4 5 6 7 8 9 10 11 |
apiVersion: dapr.io/v1alpha1 kind: Component metadata: name: statestore spec: type: state.redis metadata: - name: redisHost value: redis:6379 - name: redisPassword value: "" |
|
1 |
kubectl apply -f redis-state.yaml |
This creates a Rudr component for Redis. Rudr is a reference implementation of the Open Application Model (OAM) specification jointly created by Microsoft and Alibaba. For more details on OAM and Rudr, refer to this article and the tutorial.
Deploying Microservices That Use Dapr State Store
Let’s start by creating a Pod and Service spec that runs the first microservice written in Node.js. You can take a look at the code for this microservice at Dapr samples repo.
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 |
kind: Service apiVersion: v1 metadata: name: nodeapp labels: app: node spec: selector: app: node ports: - protocol: TCP port: 80 targetPort: 3000 nodePort: 32000 type: NodePort --- apiVersion: apps/v1 kind: Deployment metadata: name: nodeapp labels: app: node spec: replicas: 1 selector: matchLabels: app: node template: metadata: labels: app: node annotations: dapr.io/enabled: "true" dapr.io/id: "nodeapp" dapr.io/port: "3000" spec: containers: - name: node image: dapriosamples/hello-k8s-node ports: - containerPort: 3000 imagePullPolicy: Always |
Note that the Pod has Annotations for Dapr which act as a hint to inject the sidecar container.
|
1 2 3 4 |
annotations: dapr.io/enabled: "true" dapr.io/id: "nodeapp" dapr.io/port: "3000" |
|
1 |
kubectl apply -f node.yaml |
If you analyze the code, you realize that the microservice is not directly referring to the persistent store in the microservice. Instead, it makes a call to the REST endpoint exposed by Dapr runtime. The sidecar container is responsible for enabling this communication between the microservice and the Dapr runtime.
|
1 2 3 |
const daprPort = process.env.DAPR_HTTP_PORT || 3500; const stateStoreName = `statestore`; const stateUrl = `http://localhost:${daprPort}/v1.0/state/${stateStoreName}`; |
Let’s deploy the second microservice written in Python that continuously invokes the API exposed by the first service deployed in the previous step.
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 |
apiVersion: apps/v1 kind: Deployment metadata: name: pythonapp labels: app: python spec: replicas: 1 selector: matchLabels: app: python template: metadata: labels: app: python annotations: dapr.io/enabled: "true" dapr.io/id: "pythonapp" spec: containers: - name: python image: dapriosamples/hello-k8s-python |
|
1 |
kubectl apply -f python.yaml |
Since both the Pods are annotated for Dapr, the control plane has injected the sidecar into the Pods.
Checking the logs of the Node Pod shows that the state is being persisted.
|
1 2 |
NODE_POD=$(kubectl get pods -l app=node -o jsonpath='{.items[0].metadata.name}') kubectl logs -f $NODE_POD -c node |
You can also access the same from the NodePort of Minikube.
|
1 2 |
export NODE_APP=`minikube ip`:32000 curl $NODE_APP/order |
You can check the key/value pair by accessing the CLI in the Redis Pod.
|
1 2 |
REDIS_POD=$(kubectl get pods -l app=redis -o jsonpath='{.items[0].metadata.name}') kubectl exec -it $REDIS_POD -- redis-cli HGETALL "nodeapp-order" |
Replacing Redis State Store with etcd
Dapr also supports etcd as one of the Components for the state store building block service. Let’s now replace the Redis state store with etcd
First, create a etcd cluster. You can use a Helm Chart or the below YAML spec to deploy a 3-node etcd cluster in Kubernetes.
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 |
apiVersion: v1 kind: Service metadata: name: etcd-client spec: ports: - name: etcd-client-port port: 2379 protocol: TCP targetPort: 2379 selector: app: etcd --- apiVersion: v1 kind: Pod metadata: labels: app: etcd etcd_node: etcd0 name: etcd0 spec: containers: - command: - /usr/local/bin/etcd - --name - etcd0 - --initial-advertise-peer-urls - http://etcd0:2380 - --listen-peer-urls - http://0.0.0.0:2380 - --listen-client-urls - http://0.0.0.0:2379 - --advertise-client-urls - http://etcd0:2379 - --initial-cluster - etcd0=http://etcd0:2380,etcd1=http://etcd1:2380,etcd2=http://etcd2:2380 - --initial-cluster-state - new image: quay.io/coreos/etcd:latest name: etcd0 ports: - containerPort: 2379 name: client protocol: TCP - containerPort: 2380 name: server protocol: TCP restartPolicy: Always --- apiVersion: v1 kind: Service metadata: labels: etcd_node: etcd0 name: etcd0 spec: ports: - name: client port: 2379 protocol: TCP targetPort: 2379 - name: server port: 2380 protocol: TCP targetPort: 2380 selector: etcd_node: etcd0 --- apiVersion: v1 kind: Pod metadata: labels: app: etcd etcd_node: etcd1 name: etcd1 spec: containers: - command: - /usr/local/bin/etcd - --name - etcd1 - --initial-advertise-peer-urls - http://etcd1:2380 - --listen-peer-urls - http://0.0.0.0:2380 - --listen-client-urls - http://0.0.0.0:2379 - --advertise-client-urls - http://etcd1:2379 - --initial-cluster - etcd0=http://etcd0:2380,etcd1=http://etcd1:2380,etcd2=http://etcd2:2380 - --initial-cluster-state - new image: quay.io/coreos/etcd:latest name: etcd1 ports: - containerPort: 2379 name: client protocol: TCP - containerPort: 2380 name: server protocol: TCP restartPolicy: Always --- apiVersion: v1 kind: Service metadata: labels: etcd_node: etcd1 name: etcd1 spec: ports: - name: client port: 2379 protocol: TCP targetPort: 2379 - name: server port: 2380 protocol: TCP targetPort: 2380 selector: etcd_node: etcd1 --- apiVersion: v1 kind: Pod metadata: labels: app: etcd etcd_node: etcd2 name: etcd2 spec: containers: - command: - /usr/local/bin/etcd - --name - etcd2 - --initial-advertise-peer-urls - http://etcd2:2380 - --listen-peer-urls - http://0.0.0.0:2380 - --listen-client-urls - http://0.0.0.0:2379 - --advertise-client-urls - http://etcd2:2379 - --initial-cluster - etcd0=http://etcd0:2380,etcd1=http://etcd1:2380,etcd2=http://etcd2:2380 - --initial-cluster-state - new image: quay.io/coreos/etcd:latest name: etcd2 ports: - containerPort: 2379 name: client protocol: TCP - containerPort: 2380 name: server protocol: TCP restartPolicy: Always --- apiVersion: v1 kind: Service metadata: labels: etcd_node: etcd2 name: etcd2 spec: ports: - name: client port: 2379 protocol: TCP targetPort: 2379 - name: server port: 2380 protocol: TCP targetPort: 2380 selector: etcd_node: etcd2 |
|
1 |
kubectl apply -f etcd.yaml |
The etcd endpoint is exposed as a ClusterIP Service.
Now, we can delete the existing state store Component and recreate it with a pointer to etcd.
The below spec creates an etcd State Store:
|
1 2 3 4 5 6 7 8 9 10 11 |
apiVersion: dapr.io/v1alpha1 kind: Component metadata: name: statestore spec: type: state.etcd metadata: - name: endpoints value: "etcd-client:2379" - name: dialTimeout value: "5s" |
|
1 2 |
kubectl delete -f redis-state.yaml kubectl apply -f etcd-state.yaml |
Delete the Node.js Pod to force Kubernetes to launch a new Pod.
|
1 2 |
NODE_POD=$(kubectl get pods -l app=node -o jsonpath='{.items[0].metadata.name}') kubectl delete pod/$NODE_POD |
Wait for the Node.js Pod to become ready before checking if everything is intact by invoking the NodePort endpoint of the first microservice.
|
1 2 |
export NODE_APP=`minikube ip`:32000 curl $NODE_APP/order |
You can also use etcdctl, the client, from one of the etcd Pods to check the key/value pair maintaining the state.
|
1 2 3 |
kubectl exec -it etcd0 /bin/sh export ETCDCTL_API=3 etcdctl get nodeapp-order |
This tutorial demonstrated how to use Dapr state service with Kubernetes. In one of the future tutorials, we will explore the resource binding building block of Dapr.
Janakiram MSV’s Webinar series, “Machine Intelligence and Modern Infrastructure (MI2)” offers informative and insightful sessions covering cutting-edge technologies. Sign up for the upcoming MI2 webinar at http://mi2.live.