AWS Knowledge Hub

English

Tiếng Việt

English

Tiếng Việt

Appearance

Amazon EC2 ​

Summary ​

  • Amazon EC2 provides virtual servers (EC2 instances) in the AWS Cloud with elastic scale up/down, helping you reduce hardware costs and deploy applications faster.
  • The instance type you choose defines CPU, memory, storage, and networking, so choosing the right type for your workload is critical.
  • The EC2 ecosystem includes services like Amazon EBS, VPC, Security Groups, IAM, Auto Scaling, Elastic Load Balancing, and CloudWatch to secure, scale, and observe your workloads.
  • Pricing models (On‑Demand, Reserved Instances, Spot, Savings Plans) have a major impact on cost; combining them with right‑sizing is key to optimization.

Architecture diagram (EC2 in a typical web stack) ​

mermaid
flowchart LR
  User[User / Client] --> ALB[Application Load Balancer]
  ALB --> ASG[Auto Scaling Group]
  ASG --> EC2A[EC2 Instance A]
  ASG --> EC2B[EC2 Instance B]
  EC2A & EC2B --> EBS[(Amazon EBS Volumes)]
  EC2A & EC2B --> CW[Amazon CloudWatch]
  VPC[VPC + Subnets + Security Groups] --- EC2A
  VPC --- EC2B
  EC2A & EC2B --> RDS[(Amazon RDS / Aurora)]

Best Practices ​

  • Right-size your instances: use CloudWatch metrics (CPU, memory, network, I/O) to choose the smallest instance type and size that comfortably meets your workload needs.
  • Use Auto Scaling with load balancers: place EC2 instances in an Auto Scaling Group behind an Elastic Load Balancer to handle traffic spikes and improve high availability.
  • Harden networking with VPC and Security Groups: keep application/database tiers in private subnets, restrict inbound rules to only what is necessary, and use security groups as sources instead of wide CIDR ranges.
  • Standardize AMIs and patching: rely on golden AMIs and automated patching (AWS Systems Manager) to keep OS and agents up to date and reduce configuration drift.
  • Enable monitoring and logging: send system and application logs to CloudWatch Logs, configure dashboards and alarms for CPU, status checks, disk, and network metrics.
  • Protect and back up data: encrypt EBS volumes and snapshots by default, use scheduled snapshots (or AWS Backup), and regularly test restore procedures for disaster recovery.
  • Optimize cost proactively: combine Savings Plans/Reserved Instances for steady usage, Spot Instances for fault‑tolerant/batch workloads, and regularly clean up idle EC2, EBS volumes, Elastic IPs, and old snapshots.

Exam Notes ​

  • Understand the EC2 instance lifecycle (pending → running → stopping/stopped → shutting‑down → terminated) and how it affects billing, EBS, and instance store data.
  • Be able to distinguish pricing models (On‑Demand, Reserved Instances, Spot, Savings Plans) and choose the right one in scenarios about long‑running, bursty, or batch workloads.
  • Know the difference between EBS volumes (persistent, network‑attached) and instance store (ephemeral, local) and common volume types (gp3, io1/io2, st1, sc1).
  • Understand the role of Security Groups (stateful) and Network ACLs (stateless), and how they work with VPC, subnets, and Internet/NAT Gateways.
  • Recognize standard patterns for high availability on EC2: Multi‑AZ, Auto Scaling Groups, and load balancers for web/app tiers, and Multi‑AZ databases when needed.

AWS documentation ​