⚖️ Scalability & Auto Scaling

📌 Overview

Scalability is the ability of a system to handle increased load. In Azure, this is a core benefit of the cloud. You can add resources manually or automatically to meet demand.

1. Types of Scaling

⬆️ Vertical Scaling (Scale Up/Down)

"Making the machine bigger."

• Action: Increasing the size (CPU, RAM) of a Virtual Machine.
• Example: Moving from a Standard_D2s_v3 (2 vCPU, 8GB RAM) to a Standard_D4s_v3 (4 vCPU, 16GB RAM).
• Cons: Requires downtime (reboot usually needed). Limited by maximum hardware size.
• Use Case: Monolithic apps that are hard to distribute.

↔️ Horizontal Scaling (Scale Out/In)

"Adding more machines."

• Action: Adding more instances of the same VM to a pool.
• Example: Increasing your web server count from 2 VMs to 5 VMs.
• Pros: Zero downtime (add/remove nodes while others run). Theoretically infinite scale.
• Use Case: Web applications, Microservices.

2. Azure Scaling Solutions

🖥️ Azure Virtual Machine Scale Sets (VMSS)

Allows you to create and manage a group of load-balanced VMs. The number of VM instances can automatically increase or decrease in response to demand or a defined schedule.

• Feature: Supports up to 1,000 VM instances.
• Use Case: Running large compute workloads, Big Data clusters, or container orchestration nodes (AKS nodes).

🌐 Azure App Service Autoscale

Built-in feature for PaaS Web Apps.

• Action: Increases the number of instances running your Web App code.
• Triggers: Scale based on CPU usage > 70%, Memory usage, or Queue depth.

3. Architecture Diagram

Vertical vs Horizontal Scaling

       VERTICAL SCALING (Scale Up)             HORIZONTAL SCALING (Scale Out)
       ===========================             ==============================

      +-------------------+                       +-----+    +-----+    +-----+
      |   Big Server      |                       | VM1 |    | VM2 |    | VM3 |
      |  (8 vCPU, 32GB)   |                       +-----+    +-----+    +-----+
      |                   |                          ^          ^          ^
      +-------------------+                          |          |          |
               ^                               +--------------------------+
               |                               |      Load Balancer       |
               |                               +--------------------------+
      +-------------------+                                 ^
      |   Small Server    |                                 |
      |  (2 vCPU, 8GB)    |                              (Traffic)
      +-------------------+

3. Elasticity

The ability to automatically scale out and scale in based on demand.

• Scenario:
  • 9:00 AM: Users log in. Azure Autoscale adds 5 VMs (Scale Out).
  • 6:00 PM: Users log off. Azure Autoscale removes 4 VMs (Scale In).
• Benefit: Cost Optimization. You only pay for the 5 VMs during the day, and only 1 VM at night.

💡 Hinglish Explanation (Scaling)

1. Vertical Scaling (Scale Up/Down)

• Concept: Machine ko powerful banana.
• Hinglish: Aapke paas ek chota laptop hai (4GB RAM). Aapne usme aur RAM lagakar 16GB kar diya. Machine wahi hai, bas powerful ho gayi. Par iske liye laptop kholna padega (Shutdown/Restart required).

2. Horizontal Scaling (Scale Out/In)

• Concept: Machines ki ginti badhana.
• Hinglish: Aapke paas ek delivery boy hai. Kaam badha toh aapne 4 aur delivery boys hire kar liye. Kaam fast ho gaya aur kisi ek ko rokne ki zaroorat nahi padi.

3. Elasticity (Auto-Scaling)

• Hinglish: Yeh system ka "Magic" hai. Jab users badhte hain, toh Azure khud naye servers add kar deta hai (Scale Out). Jab users chale jate hain, toh servers delete kar deta hai (Scale In) taaki bill kam aaye.

💡 Exam Tips for AZ-900

• Scalability: Manual or planned capacity increase.
• Elasticity: Dynamic, automatic scaling based on demand (spikes).
• Vertical: Resize the VM (Restart required).
• Horizontal: Add more VMs (No restart required).