Cloud Technology and Services Study Guide — CLF-C02

Cloud Technology and Services – Study Guide (CLF-C02)

This domain makes up 34% of the exam and focuses on your ability to identify the right AWS service for the right use case. You won't be asked to configure services in depth, but you must know what each service does, when to choose it over alternatives, and how pricing and architecture decisions affect cost and availability.

Compute

Amazon EC2 – Know the purchasing models cold:

• On-Demand: pay per second/hour, no commitment, highest cost. Use for unpredictable workloads.
• Reserved Instances (RI): 1- or 3-year commitment, up to 72% savings. Use for steady-state workloads.
• Spot Instances: up to 90% savings, but AWS can reclaim with 2-minute notice. Use for fault-tolerant, interruptible work (batch jobs, rendering).
• Savings Plans: flexible discount (Compute or EC2), commitment is $/hour spend, not specific instance type. More flexible than RIs.
• Dedicated Hosts: physical server for compliance/licensing requirements. Most expensive option.

Trap: Spot ≠ Reserved. Spot is cheap but interruptible. Reserved is a billing commitment, not a capacity reservation (unless you use Capacity Reservations separately).

AWS Lambda – Serverless, event-driven. Runs code in response to triggers (S3 events, API Gateway, DynamoDB streams). Pricing is per request + duration (GB-seconds). Max execution timeout is 15 minutes. No server management. Best for short-lived, event-driven tasks.

ECS vs. EKS – Both run containers. ECS is AWS-native and simpler; EKS is managed Kubernetes for teams already using K8s. Fargate is the serverless compute option for both (no EC2 management).

Elastic Beanstalk – PaaS. You upload code; AWS handles provisioning, load balancing, scaling, and monitoring. Still gives you access to underlying EC2 instances. Good for developers who want simplicity without full abstraction.

Storage

Amazon S3 – Know the storage classes by use case:

• Standard: frequent access, low latency
• Standard-IA / One Zone-IA: infrequent access, lower cost, retrieval fee applies
• Glacier Instant Retrieval / Flexible / Deep Archive: archival tiers with increasing retrieval times (milliseconds → hours) and decreasing cost
• Intelligent-Tiering: auto-moves objects between tiers based on access patterns; no retrieval fees

Lifecycle policies automate transitioning objects between classes. Versioning protects against accidental deletion. S3 can host static websites (HTML/CSS/JS only — no server-side code). S3 is object storage, not a file system or block device.

Amazon EBS – Block storage attached to a single EC2 instance (mostly). Persists independently of instance life. Use for databases, OS volumes, anything needing low-latency block I/O.

Amazon EFS – Managed NFS file system. Shared access across multiple EC2 instances simultaneously. Use for shared workloads (web farms, content management).

Trap: EBS = one instance, EFS = many instances. Neither is a replacement for S3.

Databases

Amazon RDS – Managed relational DB (MySQL, PostgreSQL, SQL Server, Oracle, MariaDB). Handles patching, backups, failover. Multi-AZ = synchronous standby replica for high availability (automatic failover). Read Replicas = asynchronous copies for read scaling (not HA failover).

Amazon Aurora – AWS-proprietary, MySQL/PostgreSQL-compatible. Up to 5× faster than standard MySQL. Storage auto-scales, up to 15 read replicas. More performant and resilient than standard RDS.