Hébergement et mise à l’échelle des applications fédérées

Welcome! In this lesson, we'll explore how to host and scale your Micro Frontends (MFEs). Proper hosting ensures your apps are available and performant.

Scaling means making sure your application can handle varying amounts of user traffic smoothly. It's key for a great user experience and reliable service.

Modern MFEs often live in the cloud. Popular providers include:

• AWS (Amazon Web Services): Offers S3 for static assets, CloudFront for CDN, EC2 for servers, and EKS for Kubernetes.
• Azure (Microsoft Azure): Provides Blob Storage, Azure CDN, App Service, and AKS for Kubernetes.
• GCP (Google Cloud Platform): Features Cloud Storage, Cloud CDN, Compute Engine, and GKE for Kubernetes.

These platforms offer robust infrastructure for your federated applications.

Many remote Micro Frontends are built as static assets (HTML, CSS, JavaScript). These can be hosted very efficiently.

Services like AWS S3, Azure Blob Storage, or Google Cloud Storage are perfect for this. They offer high availability and low cost.

You can then point a CDN to these storage buckets for global distribution.

A Content Delivery Network (CDN) caches your MFE assets at locations worldwide. When a user requests your app, content is served from the closest cache.

This drastically reduces latency and improves loading speeds, especially for users far from your main servers. CDNs are a must-have for global MFE deployments.

Not all MFEs are purely static. If a Micro Frontend involves server-side rendering (SSR), API calls, or complex backend logic, it needs dynamic hosting.

Services like AWS EC2, Azure App Service, or Kubernetes clusters (EKS, AKS, GKE) are suitable for running server-side MFE components or their associated backends.

Scaling is about ensuring your application can handle varying workloads. As user traffic grows, your MFEs must be able to keep up without slowing down or crashing.

Effective scaling means your application remains responsive and available, providing a consistent user experience regardless of demand spikes.

Horizontal scaling means adding more instances (copies) of your application or service. Imagine having multiple identical servers running your MFE.

A load balancer then distributes incoming user requests across these instances. This is a common and highly effective way to handle increased traffic.

Vertical scaling means increasing the resources (CPU, RAM) of a single instance. Think of upgrading a server with a more powerful processor or more memory.

While simpler to implement initially, vertical scaling has limits. Eventually, a single server can't be made any more powerful, making horizontal scaling necessary for significant growth.

Cloud platforms offer auto-scaling features. These automatically adjust the number of instances based on predefined metrics like CPU utilization or request queue length.

For example, Kubernetes' Horizontal Pod Autoscaler (HPA) can scale your MFE pods up or down automatically.

apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: my-mfe-hpa
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: my-mfe-deployment
  minReplicas: 2
  maxReplicas: 10
  metrics:
  - type: Resource
    resource:
      name: cpu
      target:
        type: Utilization
        averageUtilization: 70

This ensures optimal resource usage and cost efficiency.

A load balancer is crucial for horizontally scaled applications. It sits in front of your MFE instances and intelligently distributes incoming user requests among them.

This prevents any single instance from becoming overwhelmed, improving overall performance and reliability. Common types include Application Load Balancers (ALB) and Network Load Balancers (NLB).

Great job! We covered essential concepts for hosting and scaling Micro Frontends:

• Cloud platforms like AWS, Azure, and GCP offer various hosting services.
• Static hosting (S3, Blob Storage) is great for remote MFE assets.
• CDNs significantly improve global MFE performance.
• Dynamic hosting handles server-side MFE components.
• Scaling (horizontal and vertical) ensures responsiveness under load.
• Automated scaling and load balancers are vital for efficient, resilient MFEs.