EC2, S3, RDS, Lambda, CloudFront, IAM, VPC and when to use each — one page, 50+ services mapped with decision guides and real-world use cases for beginners to cloud practitioners.
Run applications in the cloud — from virtual machines to containers to serverless functions.
| Instance Family | Best For |
|---|---|
| t3 / t4g | General purpose, low-cost web apps |
| c6i / c7g | CPU-intensive (compute optimized) |
| r6i / x2idn | Memory-heavy (SAP, in-mem DB) |
| p4d / g5 | GPU / Machine learning training |
| i3 / i4i | NVMe storage, high I/O |
| Service | When to Use |
|---|---|
| ECS (Fargate) | Containers without managing servers |
| ECS (EC2) | More control, custom AMIs needed |
| EKS | Already using Kubernetes; complex workloads |
| ECR | Private Docker image registry on AWS |
| App Runner | Simplest way to deploy containers |
| Need | Pick | Why |
|---|---|---|
| Full OS control, custom software | EC2 | IaaS, full root access, any OS/runtime |
| Run existing app, don't manage infra | Elastic Beanstalk | PaaS, handles scaling/patching automatically |
| Containerised microservices (AWS-native) | ECS + Fargate | No cluster management, per-task billing |
| Kubernetes workloads | EKS | Managed K8s control plane |
| Event-driven, short tasks < 15 min | Lambda | Serverless, pay per invocation |
| Simple web app from container image | App Runner | Fully managed, simplest deployment path |
| High performance computing / batch jobs | AWS Batch | Managed batch scheduling on EC2/Spot |
Object, block, file, and archive storage — pick the right one for your access patterns.
| Storage Class | Use Case |
|---|---|
| S3 Standard | Frequently accessed data |
| Standard-IA | Infrequent access, rapid retrieval |
| One Zone-IA | Infrequent, non-critical, single AZ |
| Intelligent-Tiering | Unknown/changing access patterns |
| Glacier Instant | Archive, millisecond retrieval |
| Glacier Deep | Long-term archive, hours retrieval |
| Service | Type | Key Trait |
|---|---|---|
| EBS gp3 | Block | General SSD, single instance |
| EBS io2 | Block | High IOPS, databases |
| EBS st1 | Block | Throughput HDD, big data |
| EFS | File | Shared NFS, multi-EC2 |
| FSx Lustre | File | HPC, ML training fast FS |
| FSx Windows | File | Windows file share (SMB) |
Relational, NoSQL, in-memory, time-series, graph — AWS has a managed database for every data model.
| Engine | Best For |
|---|---|
| Aurora MySQL | Up to 5× faster MySQL, cloud-native |
| Aurora PostgreSQL | Up to 3× faster Postgres, cloud-native |
| MySQL | Standard MySQL, lower cost |
| PostgreSQL | Advanced SQL, JSON, extensions |
| Oracle | License included or BYOL |
| SQL Server | Microsoft workloads |
| MariaDB | MySQL fork, open-source |
| Database | Type | When to Use |
|---|---|---|
| RDS / Aurora | Relational | Structured data, ACID transactions, reporting, ERP/CRM systems |
| DynamoDB | Key-Value / Doc | High-scale web apps, gaming, IoT, user sessions, shopping carts |
| ElastiCache (Redis) | In-Memory | Caching, leaderboards, pub/sub, real-time analytics, session store |
| ElastiCache (Memcached) | In-Memory | Simple caching, multi-threaded, stateless |
| Neptune | Graph | Social networks, fraud detection, knowledge graphs, recommendations |
| Timestream | Time-Series | IoT telemetry, DevOps metrics, industrial sensor data |
| DocumentDB | Document | MongoDB-compatible workloads, content management, catalogs |
| Keyspaces | Wide Column | Cassandra-compatible, IoT at scale, time-series-like |
| Redshift | Data Warehouse | Petabyte-scale analytics, BI tools, OLAP queries |
| OpenSearch | Search | Full-text search, log analytics, Elasticsearch-compatible |
VPC, subnets, load balancers, DNS, CDN — control how traffic flows in and out of AWS.
| Service | Layer | Use For |
|---|---|---|
| ALB | L7 HTTP/S | Web apps, path-based routing, gRPC |
| NLB | L4 TCP/UDP | Ultra-low latency, static IP, gaming |
| GLB | L3/L4 | Deploy network appliances (firewalls) |
| CLB | L4/L7 | Legacy (prefer ALB) |
Control who can access what in AWS. Identity, encryption, compliance, and threat detection.
Build event-driven apps without provisioning servers. Pay only for what you use.
Build, test, deploy, and manage infrastructure as code with AWS developer services.
From pre-built AI APIs to full ML training pipelines — AWS has services for every ML skill level.
Know what's happening in your AWS environment — metrics, logs, traces, and alarms.
Every major AWS service, category, and one-line description for fast lookup.