Deployment

Kipchak is designed for modern cloud environments. It is fully containerised and ready for deployment on Kubernetes.

Docker

The Kipchak starter project includes a Dockerfile that is optimized for production. It uses FrankenPHP as the base image, ensuring high performance out of the box.

Production Readiness

Please change the FROM line in your Dockerfile to the production PHP Image by removing -dev:

Change:

FROM 1x.ax/mamluk/php/8.4:franken-dev

to:

FROM 1x.ax/mamluk/php/8.4:franken

Building the Image

To build your production image:

docker build . -t my-kipchak-api .

Running the Container

docker run -p 80:80 my-kipchak-api

Kubernetes

Kipchak is Kubernetes-ready. Here's an example manifest of how to deploy the starter project to Kubernetes. You'll just need to change the container image name to your own and container registry and the correct name for your registry secret and the correct environtment variables.

apiVersion: apps/v1
kind: Deployment
metadata:
  name: kipchak-api
  namespace: kipchak-namespace
spec:
  replicas: 2
  strategy:
    type: RollingUpdate
    rollingUpdate:
      maxSurge: 2
      maxUnavailable: 0
  selector:
    matchLabels:
      app: kipchak-api
  template:
    metadata:
      labels:
        app: kipchak-api
      annotations:
        prometheus.io/scrape: "true"
        prometheus.io/endpoint: "/metrics"
        prometheus.io/port: "2019"
    spec:
      imagePullSecrets:
        - name: your-registry-secret
      topologySpreadConstraints:
        - maxSkew: 1
          topologyKey: kubernetes.io/hostname
          whenUnsatisfiable: ScheduleAnyway
          labelSelector:
            matchLabels:
              app: aladhan-com-api
      containers:
        - name: aladhan-com-api
          image: 1x.ax/islamic-network/api.aladhan.com:$COMMIT_TAG
          env:
            - name:  MEMCACHED_HOST
              value: "memcached"
            - name:  MEMCACHED_PORT
              value: "11211"
          livenessProbe:
            httpGet:
              path: /status
              port: 80
            initialDelaySeconds: 3
            periodSeconds: 7
            timeoutSeconds: 5
            failureThreshold: 7
          startupProbe:
            httpGet:
              path: /status
              port: 80
            periodSeconds: 3
            failureThreshold: 3
          resources:
            requests:
              cpu: "250m"
              memory: 350Mi
            limits:
              cpu: "750m"
              memory: 750Mi
          ports:
            - containerPort: 80
              name: api
              protocol: TCP
            - containerPort: 2019
              name: metrics
              protocol: TCP

Horizontal Pod Autoscaling (HPA)

Because Kipchak is lightweight and starts quickly, it works exceptionally well with Kubernetes HPA, allowing your API to scale based on CPU or memory usage.

Notes

Kipchak follows best practices for containerised applications:

• Logs: Logs are sent to stdout and stderr for easy collection by logging drivers.
• Environment Variables: Configuration is primarily handled through environment variables.
• Health Checks: Built-in support for readiness and liveness probes for the starter API.