Azure Functions and Serverless

Azure Functions lets you run event-driven code without managing infrastructure. Write a function, define a trigger, and Azure handles scaling, patching, and availability. It's the core of serverless computing on Azure.

---

What Is Serverless?

Serverless doesn't mean no servers — it means you don't manage them. The cloud provider handles all the infrastructure, and you focus entirely on your code.

Serverless Benefits

• No infrastructure management — no VMs to provision, patch, or scale
• Automatic scaling — from zero to thousands of instances
• Pay-per-execution — billed only for the time your code runs
• Built-in high availability — the platform handles fault tolerance

---

How Azure Functions Work

The Execution Model

1. You write a function — a small, focused piece of code
2. You define a trigger — the event that causes the function to run
3. Optionally, you configure bindings — declarative connections to input/output data sources
4. When the trigger fires, Azure executes your function and scales automatically

Supported Languages

Language	Runtime
C#	.NET (in-process and isolated worker)
JavaScript / TypeScript	Node.js
Python	Python 3.x
Java	Java 8, 11, 17, 21
PowerShell	PowerShell 7.x
Go / Rust / others	Custom handlers

---

Triggers and Bindings

Triggers

A trigger defines what causes a function to run. Each function has exactly one trigger:

Trigger	Event
HTTP	An HTTP request (REST API endpoint)
Timer	A scheduled interval (cron expression)
Blob Storage	A blob is created or updated
Queue Storage	A message is added to a queue
Cosmos DB	A document is created or changed
Event Hub	An event is received
Service Bus	A message is received
Event Grid	An Event Grid event is published

Bindings

Bindings are declarative ways to connect to data sources without writing boilerplate code:

• Input binding — read data when the function runs
• Output binding — write data when the function completes

Example: Queue trigger with Blob output binding (JavaScript)

module.exports = async function (context, queueMessage) {
  // queueMessage is the input (trigger)
  const result = processOrder(queueMessage);

  // context.bindings.outputBlob is the output binding
  context.bindings.outputBlob = JSON.stringify(result);
};

The binding configuration (defined in function.json or via attributes) handles connecting to the queue and blob — no SDK code needed.

---

Hosting Plans