Notes_week6.md

Serverless in Azure and AWS

• There is always a server, but we aren't managing them.
• Before serverless, the only option was to run apps on a virtual machine. Then, you would still have to worry about the infrastructure.
• With serverless, you simply deploy your code.
• No more managing infrastructure.
• Focus on the code.
• Smaller footprint.
• Serverless is not just web frontend apps.
• Backend can also work serverless.
• Scalability
  • Auto-scaling
  • Much better performance

Azure Functions

• Azure Functions are part of the serverless backend family.
• Azure Functions allow you to upload any type of backend code.

Azure Web Apps

• Frontend code.
• Supports many languages.
• Can run pure code or in a container.

AWS Elastic Bean Stalk

• Frontend code.
• Supports multiple languages.
• Here, you can also run containers.

Serverless Monitoring

• In Azure: Azure Monitor and Application Insights.
• In AWS: Cloudwatch.

Serverless Pros and Cons

Pros	Cons
No infrastructure to manage	Data isn't stored
Highly scalable	Vendor lock-in
Pay for what you use	Debugging may be different for sysadmins

Serverless Kubernetes

• GKE Autopilot from GCP.
• No nodes need to be managed.
• Billed per pod.