AWS Services Quick Reference
The definitive quick-reference for AWS's sprawling service catalog. This pack distills 50+ AWS services into scannable one-page summaries covering what each service does, when to use it, key configuration options, pricing model, and common pitfalls. Stop tab-switching between AWS docs — keep this on your desk and find the right service in seconds.
What's Included
- Compute Services — EC2, Lambda, ECS, EKS, Fargate, Lightsail, Batch, App Runner
- Storage Services — S3, EBS, EFS, FSx, Storage Gateway, Glacier
- Database Services — RDS, DynamoDB, ElastiCache, Neptune, Redshift, Aurora, DocumentDB
- Networking — VPC, Route 53, CloudFront, API Gateway, ELB/ALB/NLB, Transit Gateway
- Security & Identity — IAM, KMS, Secrets Manager, WAF, Shield, GuardDuty, Cognito
- ML & Analytics — SageMaker, Bedrock, Athena, Glue, Kinesis, QuickSight
- DevOps & Management — CloudFormation, CloudWatch, Systems Manager, CodePipeline
- Comparison Charts — Storage tiers, database types, compute pricing models
Preview / Sample Content
Lambda vs EC2 vs Fargate — When to Use What
┌─────────────────────────────────────────────────────────────────┐
│ DECISION: Where should this workload run? │
├──────────────┬──────────────────┬───────────────────────────────┤
│ Duration │ < 15 minutes │ Lambda │
│ │ Long-running │ EC2 or Fargate │
├──────────────┼──────────────────┼───────────────────────────────┤
│ Traffic │ Spiky / Bursty │ Lambda or Fargate │
│ │ Steady-state │ EC2 (Reserved/Savings Plan) │
├──────────────┼──────────────────┼───────────────────────────────┤
│ State │ Stateless │ Lambda or Fargate │
│ │ Stateful │ EC2 with EBS │
├──────────────┼──────────────────┼───────────────────────────────┤
│ Container? │ Yes │ Fargate (serverless) or ECS │
│ │ No │ Lambda or EC2 │
└──────────────┴──────────────────┴───────────────────────────────┘
S3 Storage Classes at a Glance
S3 Standard → Frequently accessed, < 1ms latency
S3 Intelligent-Tier → Unknown/changing access patterns, auto-tiering
S3 Standard-IA → Infrequent access, min 30-day storage, retrieval fee
S3 One Zone-IA → Same as IA but single AZ (non-critical reproducible data)
S3 Glacier Instant → Archive with millisecond retrieval
S3 Glacier Flexible → Archive, 1-5 min (expedited) to 5-12 hr retrieval
S3 Glacier Deep → Cheapest, 12-48 hr retrieval, 180-day minimum
IAM Policy Essentials
{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"s3:GetObject",
"s3:PutObject"
],
"Resource": "arn:aws:s3:::my-bucket/*",
"Condition": {
"IpAddress": {
"aws:SourceIp": "10.0.0.0/8"
}
}
}
]
}
Key IAM rules:
- Default deny — everything is denied unless explicitly allowed
- Explicit deny always wins over any allow
- Use
Conditionkeys to restrict by IP, MFA, time, tags - Prefer managed policies over inline policies
- Never attach policies directly to users — use groups or roles
Quick Reference Table
| Service | Category | Use When... | Pricing Model |
|---|---|---|---|
| EC2 | Compute | Need full OS control, GPU, steady workloads | Per-hour/second |
| Lambda | Compute | Event-driven, < 15 min, spiky traffic | Per-invocation + duration |
| S3 | Storage | Object storage, static hosting, data lake | Per-GB + requests |
| RDS | Database | Relational data, need managed Postgres/MySQL | Per-hour + storage |
| DynamoDB | Database | Key-value, < 10ms latency, auto-scaling | Per-RCU/WCU or on-demand |
| CloudFront | CDN | Global content delivery, edge caching | Per-GB transfer + requests |
| Route 53 | DNS | DNS hosting, health checks, routing policies | Per-hosted-zone + queries |
| SQS | Messaging | Decouple services, async processing | Per-million requests |
| SNS | Messaging | Fan-out notifications, pub/sub | Per-million publishes |
| EKS | Containers | Kubernetes orchestration at scale | Cluster fee + compute |
| Cognito | Auth | User sign-up/sign-in, OAuth, SAML | Per-MAU (free tier: 50k) |
| CloudWatch | Monitoring | Logs, metrics, alarms, dashboards | Per-metric + log ingestion |
Usage Tips
- Start with the decision trees — Don't memorize services. Use the flowcharts to match your workload to the right service.
- Check the pricing gotchas — Each service page lists hidden costs (data transfer, API calls, minimum storage durations).
- Use the comparison charts for architectural decisions — they save hours of doc-reading.
- Pin the IAM reference — Every AWS project starts with getting permissions right.
- Cross-reference with the networking sheet — VPC design affects every other service's performance and cost.
Comparison: Managed Database Services
| Feature | RDS | Aurora | DynamoDB | ElastiCache | Redshift |
|---|---|---|---|---|---|
| Type | Relational | Relational | Key-Value/Doc | In-Memory | Columnar |
| SQL | Full SQL | Full SQL | PartiQL | None | Full SQL |
| Max Storage | 64 TB | 128 TB | Unlimited | 340 GB/node | 2 PB |
| Replication | Multi-AZ | Multi-AZ + Global | Global Tables | Cluster Mode | RA3 nodes |
| Serverless | No | Aurora Serverless v2 | On-Demand mode | No | Serverless |
| Best For | OLTP apps | High-perf OLTP | Web/mobile scale | Caching/sessions | Analytics/OLAP |
This is 1 of 11 resources in the Cheatsheet Reference Pro toolkit. Get the complete [AWS Services Quick Reference] with all files, templates, and documentation for $25.
Or grab the entire Cheatsheet Reference Pro bundle (11 products) for $79 — save 30%.
Top comments (0)