🧠Exam Guide: Cloud Practitioner
Domain 3: Cloud Technology & Services
📘Task Statement 3.3
🎯What Is This Task Testing?
You need to recognize AWS compute options and match them to scenarios:
- Amazon EC2 and common Instance Families (compute optimized, storage optimized, etc.)
- Container services (Amazon ECS, Amazon EKS)
- Serverless Compute (AWS Lambda) and Serverless Container Compute (AWS Fargate)
- Elasticity via Auto Scaling
- The purpose of Load Balancers
1) 🖥️ Amazon EC2
Virtual Servers
Amazon EC2 provides resizable virtual servers. Can you choose the right instance type for a workload?
Compute Optimized
Compute Optimized Instances Are Best For: CPU-heavy workloads
Examples: high-performance web servers, batch processing, gaming servers, some HPC tasks
"Words" Associated With Compute Optimized Instances: “CPU intensive,” “high compute,” “needs lots of processing power.”
Storage Optimized
Storage Optimized Instances Are Best For: high, fast local storage access (high IOPS/throughput)
Examples: data warehousing, log processing, distributed file systems, workloads needing very fast local disk
"Words" Associated With Storage Optimized Instances: “high I/O,” “low-latency storage,” “lots of reads/writes,” “high throughput.”
(You might also see other families on exams, like memory optimized or general purpose, but compute-optimized and storage-optimized are common callouts, and the wording is self-explanatory, it shouldn't be hard to figure out.)
2) 🧩 Containers
Containers package code and dependencies so applications run consistently.
Amazon ECS (Elastic Container Service)
A managed container orchestration service from AWS.
Amazon ECS is Best For:
- running containers without needing Kubernetes
- teams that want AWS-native simplicity
- microservices and batch/container workloads
“Run containers with a simpler AWS-managed orchestrator” → ECS.
Amazon EKS (Elastic Kubernetes Service)
A managed service to run Kubernetes on AWS.
Amazon EKS is Best For:
- organizations already using Kubernetes
- portability across Kubernetes environments
- needing the Kubernetes ecosystem and tooling
“Need Kubernetes” → EKS.
3) ⚡ Serverless Compute Options
AWS Lambda
Runs code in response to events without managing servers.
AWS Lambda is Best For:
- event-driven workloads (file uploads, API calls, scheduled jobs)
- short-running tasks and automation
- rapid development where you don’t want server management
“Run code without provisioning servers” / “event-driven” → Lambda.
AWS Fargate
A serverless compute engine for containers (works with ECS and EKS).
AWS Fargate is Best For:
- running containers without managing EC2 worker nodes
- reducing operations overhead while still using containers
“Containers but no server/cluster management” → Fargate.
4) 📈 Elasticity with Auto Scaling
Auto Scaling automatically adjusts capacity to match demand.
Why Auto Scaling Matters:
- provides elasticity (scale out/in based on usage)
- improves availability during traffic spikes
- helps control cost by scaling in when demand drops
“Traffic is unpredictable” or “need to automatically add/remove instances” → Auto Scaling.
5) ⚖️ Load Balancers
A Load Balancer distributes incoming traffic across multiple targets like EC2 instances, containers, or IPs.
Core Purpose of Load Balancers:
- High Availability: if one target fails, traffic can route to healthy targets
- Scalability: supports adding/removing targets without changing the client experience
- Improved Performance: spreads traffic to avoid overload
“Distribute traffic across multiple servers” / “single entry point” / “improve availability” → Load Balancer.
✅ Quick Exam-Style Summary
-
EC2: virtual machines, pick instance families based on workload:
- compute optimized: CPU-heavy
- storage optimized: high I/O, fast local storage
- ECS: AWS-managed container orchestration, EKS: Kubernetes on AWS.
- Lambda: event-driven serverless functions, Fargate: serverless containers.
- Auto Scaling: delivers elasticity by scaling capacity up/down automatically.
- Load balancers: distribute traffic to improve availability and scale.
Top comments (0)