dev-ops-challenges

Deploy Guestbook Application on Kubernetes

Technical Overview

Deploying a multi-tier web application on Kubernetes requires orchestrating frontend and backend services to ensure seamless communication, scalability, and resource limits. A classic example is the Redis Guestbook Application:

1. Web Frontend Tier: A PHP application that serves the web UI to users, accepts guestbook entries, and retrieves records. It is exposed to external users via a NodePort Service.
2. Redis Database Tier (Backend): A replicated storage system consisting of:
   • A Redis Master instance for handling write operations.
   • A Redis Slave cluster (replicated follower instances) for handling read operations, ensuring high read availability.

The communication between the frontend tier and backend tier is dynamically resolved via Kubernetes DNS services.

graph TD
    subgraph FrontendTier ["Frontend Tier"]
        ServiceFE["Service (NodePort): frontend <br> Port: 80 / NodePort: 30009"] -->|Routes HTTP Traffic| PodFE1["Pod: frontend-1"]
        ServiceFE -->|Routes HTTP Traffic| PodFE2["Pod: frontend-2"]
        ServiceFE -->|Routes HTTP Traffic| PodFE3["Pod: frontend-3"]
    end

    subgraph BackendStorageTier ["Backend Storage Tier"]
        PodFE1 & PodFE2 & PodFE3 -->|Writes to port 6379| ServiceMaster["Service (ClusterIP): redis-master <br> Port: 6379"]
        PodFE1 & PodFE2 & PodFE3 -->|Reads from port 6379| ServiceSlave["Service (ClusterIP): redis-slave <br> Port: 6379"]
        
        ServiceMaster -->|Routes Writes| PodMaster["Pod: redis-master <br> 1 Replica"]
        ServiceSlave -->|Routes Reads| PodSlave1["Pod: redis-slave-1"]
        ServiceSlave -->|Routes Reads| PodSlave2["Pod: redis-slave-2"]
        
        ServiceFollower["Service (ClusterIP): redis-follower <br> Port: 6379"] -->|Selector: app=redis-slave| PodSlave1
        ServiceFollower -->|Selector: app=redis-slave| PodSlave2
        
        PodMaster -.->|Replication| PodSlave1 & PodSlave2
    end

    User["Web Client"] -->|HTTP Request to NodeIP:30009| ServiceFE

---

Service Discovery & Resource Allocation Deep Dive

1. DNS-based Service Discovery

In the Guestbook application, the frontend needs to connect to the Redis database without knowing the static IP addresses of the backend pods.

• DNS Resolution: Kubernetes provides a built-in CoreDNS service. When a Kubernetes Service is created (e.g., redis-master), a DNS record is registered as redis-master.default.svc.cluster.local.
• Environment Variable Configuration: The environment variable GET_HOSTS_FROM is set to dns. This directs the frontend container’s connection client to lookup hostnames redis-master and redis-slave via the cluster DNS resolver, rather than relying on legacy environment variables.

2. Resource Allocation (Requests vs Limits)

Defining CPU and memory requirements ensures that containers have guaranteed resources to run and prevents rogue containers from exhausting host memory.

• Resource Requests (resources.requests): Specifies the minimum resources a container requires to start. The Kubernetes scheduler uses this value to determine which node has enough capacity to host the Pod.
• Units:
  • CPU: Measured in millicores (e.g., 100m is $0.1$ of a CPU core).
  • Memory: Measured in Megabytes/Mebibytes (e.g., 100Mi is $100 \times 2^{20}$ bytes).

---

Infrastructure & Configuration Requirements

• Target Cluster: Nautilus Kubernetes Cluster
• Jump Host User: thor
• Namespace: default

1. Back-End Tier Requirements

Redis Master:

• Deployment Name: redis-master
• Replicas: 1
• Container Name: master-redis-xfusion
• Image: redis
• Container Port: 6379
• Resource Requests: CPU 100m, Memory 100Mi
• Service Name: redis-master (Port/TargetPort: 6379)

Redis Slave:

• Deployment Name: redis-slave
• Replicas: 2
• Container Name: slave-redis-xfusion
• Image: gcr.io/google_samples/gb-redisslave:v3
• Container Port: 6379
• Environment Variables: GET_HOSTS_FROM = dns
• Resource Requests: CPU 100m, Memory 100Mi
• Service Name: redis-slave (Port: 6379)
• Follower Service Name: redis-follower (Port/TargetPort: 6379, Selector: app = redis-slave)

2. Front-End Tier Requirements

Frontend PHP:

• Deployment Name: frontend
• Replicas: 3
• Container Name: php-redis-xfusion
• Image: gcr.io/google-samples/gb-frontend@sha256:a908df8486ff66f2c4daa0d3d8a2fa09846a1fc8efd65649c0109695c7c5cbff
• Container Port: 80
• Environment Variables: GET_HOSTS_FROM = dns
• Resource Requests: CPU 100m, Memory 100Mi
• Service Name: frontend (Type: NodePort, Port: 80, NodePort: 30009)

---

Step-by-Step Implementation

Step 1: Connect to the Kubernetes Jump Host

Establish an SSH connection to the cluster control host:

ssh thor@jump_host_ip

---

Step 2: Create the Redis Master Manifest

Create a file named redis-master.yaml to define the Redis Master Deployment and Service:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: redis-master
  labels:
    app: redis
    role: master
    tier: backend
spec:
  replicas: 1
  selector:
    matchLabels:
      app: redis
      role: master
      tier: backend
  template:
    metadata:
      labels:
        app: redis
        role: master
        tier: backend
    spec:
      containers:
      - name: master-redis-xfusion
        image: redis
        resources:
          requests:
            cpu: 100m
            memory: 100Mi
        ports:
        - containerPort: 6379
---
apiVersion: v1
kind: Service
metadata:
  name: redis-master
  labels:
    app: redis
    role: master
    tier: backend
spec:
  ports:
  - port: 6379
    targetPort: 6379
  selector:
    app: redis
    role: master
    tier: backend

Apply the manifest file:

kubectl apply -f redis-master.yaml

---

Step 3: Create the Redis Slave & Follower Manifest

Create a file named redis-slave.yaml defining the Redis Slave Deployment, redis-slave Service, and redis-follower Service:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: redis-slave
  labels:
    app: redis-slave
    role: slave
    tier: backend
spec:
  replicas: 2
  selector:
    matchLabels:
      app: redis-slave
      role: slave
      tier: backend
  template:
    metadata:
      labels:
        app: redis-slave
        role: slave
        tier: backend
    spec:
      containers:
      - name: slave-redis-xfusion
        image: gcr.io/google_samples/gb-redisslave:v3
        resources:
          requests:
            cpu: 100m
            memory: 100Mi
        env:
        - name: GET_HOSTS_FROM
          value: dns
        ports:
        - containerPort: 6379
---
apiVersion: v1
kind: Service
metadata:
  name: redis-slave
  labels:
    app: redis-slave
    role: slave
    tier: backend
spec:
  ports:
  - port: 6379
  selector:
    app: redis-slave
    role: slave
    tier: backend
---
apiVersion: v1
kind: Service
metadata:
  name: redis-follower
  labels:
    app: redis-slave
    role: slave
    tier: backend
spec:
  ports:
  - port: 6379
    targetPort: 6379
  selector:
    app: redis-slave

Apply the slave manifest file:

kubectl apply -f redis-slave.yaml

---

Step 4: Create the Frontend Tier Manifest

Create a file named frontend.yaml to define the Frontend Deployment and its NodePort Service:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: frontend
  labels:
    app: guestbook
    tier: frontend
spec:
  replicas: 3
  selector:
    matchLabels:
      app: guestbook
      tier: frontend
  template:
    metadata:
      labels:
        app: guestbook
        tier: frontend
    spec:
      containers:
      - name: php-redis-xfusion
        image: gcr.io/google-samples/gb-frontend@sha256:a908df8486ff66f2c4daa0d3d8a2fa09846a1fc8efd65649c0109695c7c5cbff
        resources:
          requests:
            cpu: 100m
            memory: 100Mi
        env:
        - name: GET_HOSTS_FROM
          value: dns
        ports:
        - containerPort: 80
---
apiVersion: v1
kind: Service
metadata:
  name: frontend
  labels:
    app: guestbook
    tier: frontend
spec:
  type: NodePort
  ports:
  - port: 80
    targetPort: 80
    nodePort: 30009
    protocol: TCP
  selector:
    app: guestbook
    tier: frontend

Apply the frontend manifest file:

kubectl apply -f frontend.yaml

---

Post-Deployment Verification

1. Verify Pod Execution and Scaling

Ensure that all pods have reached the Running state and the replica counts are matching:

kubectl get pods

Expected Output:

NAME                            READY   STATUS    RESTARTS   AGE
frontend-6bdf9c7fdb-abcde      1/1     Running   0          45s
frontend-6bdf9c7fdb-fghij      1/1     Running   0          45s
frontend-6bdf9c7fdb-klmno      1/1     Running   0          45s
redis-master-5b8d4c9d74-xyz12   1/1     Running   0          2m
redis-slave-7c5b6b8d4e-123ab    1/1     Running   0          1m
redis-slave-7c5b6b8d4e-cdef5    1/1     Running   0          1m

---

2. Verify Services Port Bindings

Check that the services are registered with the correct IP addresses, selector mappings, and ports:

kubectl get svc

Expected Output:

NAME             TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)        AGE
kubernetes       ClusterIP   10.96.0.1       <none>        443/TCP        2d
frontend         NodePort    10.96.155.80    <none>        80:30009/TCP   45s
redis-master     ClusterIP   10.96.221.14    <none>        6379/TCP       2m
redis-slave      ClusterIP   10.96.102.73    <none>        6379/TCP       1m
redis-follower   ClusterIP   10.96.90.118    <none>        6379/TCP       1m

---

3. Verify Redis Follower Selector Mapping

Verify that redis-follower endpoint references the correct targets:

kubectl get endpoints redis-follower

Expected Output showing two endpoints corresponding to the redis-slave pods:

NAME             ENDPOINTS                         AGE
redis-follower   10.244.1.4:6379,10.244.2.6:6379   1m

---

4. Verify Frontend Access and Storage Synchronization

Verify connectivity to the guestbook frontend and verify that data written to the frontend persists correctly in Redis:

# Perform a curl request to check the frontend landing page
curl -I http://<NODE_IP>:30009

Expected Output:

HTTP/1.1 200 OK
Date: ...
Server: Apache/...
Content-Type: text/html; charset=UTF-8
...

The multi-tier Guestbook Application has been successfully deployed and verified on Kubernetes!