AWS Explained: The Most Important AWS Services To Know

Summary of Key AWS Services, Features, and Concepts

Core idea / learning approach

• There are 300+ AWS services, but the speaker argues that becoming effective requires knowing ~50 or fewer, and mastering even fewer.
• The video uses an e-commerce-style reference architecture to explain the most important services layer-by-layer.

---

1) Internet Entry, Routing, and Static Web Hosting

Amazon Route 53

• DNS and traffic management for domains.
• Supports health checks so traffic only routes to healthy infrastructure behind a URL.
• Traffic routing controls:
  • Geographic routing (send users to the closest region)
  • Latency-based routing (send to the fastest endpoint)
  • Failover to maintain stability if resources fail

Amazon S3

• General-purpose blob storage for static assets (HTML, images, CSS/JS, text files).
• Used for website hosting/static asset delivery behind the application endpoint.
• Emphasized features:
  • Flexible file storage and integrations (e.g., event triggering)
  • Common use with other services for event-driven workflows
• Regional hosting limitation: objects are stored in a chosen region, which can add latency for global users.

Amazon CloudFront

• Used for global low-latency caching/replication.
• Integrates with S3:
  • Store in S3, then replicate via CloudFront to multiple regions/edges.
  • Routing decisions can be based on geography and latency.

---

2) API Layer and Request Distribution

Elastic Load Balancing (ELB) / Application Load Balancer

• Provides horizontal scaling by forwarding requests to multiple compute instances.
• Load balancer selects which instance handles requests.
• Works with auto scaling groups to add/remove instances with demand.
• Supports sticky sessions (useful for WebSockets).

API Gateway

• Described as the “front door of AWS” for managed APIs.
• Can route to:
  • Compute (backend instances/functions)
  • Or directly to many AWS services via REST endpoints (reducing the need for custom compute)
• API controls/features:
  • Rate limiting/throttling
  • Permissions
  • WebSocket support
• Speaker preference: uses API Gateway over load balancers for many API-style workloads.

---

3) API Security, Encryption, and Authentication/Authorization

AWS WAF (Web Application Firewall)

• Protects against:
  • Bot/scanning attacks
  • Exploitation attempts like SQL injection
• Can ban/prevent IPs/users from hitting endpoints.

AWS Shield

• DDoS protection:
  • Mentions a free tier
  • Shield Advanced is positioned for mission-critical apps (noting it can be costly)

AWS Certificate Manager

• Ensures endpoints have valid TLS certificates.
• Supports encryption so traffic is encrypted end-to-end.
• Certificates can come from AWS or external CAs.

Amazon Cognito

• Adds authentication and authorization to APIs.
• Creates user pools; users authenticate and receive tokens.
• API Gateway uses the token to allow/deny access.
• Also supports authorization via user attributes (e.g., admin vs student permissions).

---

4) Compute Services (Business Logic / Containers / Serverless)

Amazon EC2

• “Rent instances” (virtual machines).
• Highly flexible but requires:
  • Patching
  • Availability management
  • Instance lifecycle/teardown
• Tradeoff: flexibility vs operational overhead.

AWS Lightsail

• “EC2 for dummies” / simplified guided setup.
• Uses templates (example: WordPress).
• Premium vs EC2, but less maintenance.
• Suitable for quick setups; speaker suggests it’s less ideal for larger backend/API workloads.

Elastic Container Service (ECS) + Fargate

• ECS manages container lifecycle/health.
• Containers can run on EC2, or via Fargate (serverless containers).
• Speaker frames Fargate as reducing the need to manage EC2 placement/scaling details.

Elastic Kubernetes Service (EKS)

• Managed Kubernetes platform.
• Mentioned as an option for those in the Kubernetes ecosystem.

AWS Lambda

• Serverless functions: deploy code, AWS runs/scales it.
• Automatically scales “up and down” based on requests/events.
• Integrations and use cases:
  • API backends
  • Event processing (e.g., processing messages)
  • Tight integrations with S3 (upload → trigger Lambda)
  • Cognito triggers (sign-up → trigger Lambda)
  • Integrates with API Gateway and load balancers
• Speaker calls Lambda extremely powerful and low-maintenance.

---

5) Storage Complements for Compute

Amazon EFS

• Elastic File System that can mount to multiple instances.
• Use case: shared static content across instances.

Amazon EBS

• Block storage attached to a single EC2 instance.
• Use case: databases/transactional workloads requiring consistency and high throughput (described like SSD attached storage).

---

6) Secrets and Configuration

AWS Secrets Manager

• Stores and retrieves sensitive credentials safely:
  • Encrypted secrets
  • Controlled access via permissions
  • Apps retrieve secrets by name without hardcoding values

AWS AppConfig

• Manages feature flags externally from compute code.
• Supports agents that pull updates regularly (keeps services current and reduces manual polling).

---

7) Databases and Data Stores (Relational, NoSQL, Specialized, Analytics)

Relational

Amazon RDS

• Managed relational databases (e.g., MySQL, PostgreSQL, Oracle, SQL Server implied).
• Benefits: automation, multi-AZ, scaling, monitoring

Amazon Aurora

• AWS-built relational DB compatible with MySQL/Postgres.
• Includes serverless mode (scale down to zero, scale up on demand)

NoSQL

Amazon DynamoDB

• Managed key-value store; predictable performance for key lookups

Amazon DocumentDB

• MongoDB-compatible managed NoSQL

Amazon Keyspaces

• Managed Cassandra-compatible (described as Cassandra for managed use)

Specialized / Search / Graph

Amazon Neptune

• Graph database; examples include social relationships and relationship queries
• Covers analytical vs transactional graph queries

Amazon OpenSearch

• For analytics/search/log-like workloads
• Described as Elasticsearch-derived (supports dashboards like Kibana)

AWS DMS (Database Migration Service)

• Tool to migrate databases into AWS (e.g., into RDS)

Caching

Amazon ElastiCache

• Managed Redis/Memcached
• Cache state loss when nodes go down; requires rehydration

Amazon MemoryDB

• Redis-compatible caching that preserves state via replication logs so cache can be restored after failures more like a persistent store

• Both cited as microsecond retrieval vs millisecond for other approaches.

---

8) AI Services (Developer-Facing vs Scientist-Facing)

Amazon Bedrock

• Provides access to multiple foundational models (examples named: Llama, Anthropic models).
• Used for:
  • Chatbots
  • Image generation
  • Voice generation
• Positioned for application developers.

Amazon SageMaker

• Positioned for scientists/developers training and deploying models in AWS’s ML ecosystem.

Additional AI niche services

• AWS Rekognition: image/video analysis (faces, emotions, text, scene context)
• Amazon Polly: text-to-speech (voice selection)
• AWS Transcribe: speech-to-text, including speaker/actor identification

---

9) Asynchronous Communication / Event-Driven Architecture

Amazon SNS (Simple Notification Service)

• Pub/sub using topics:
  • publishers send messages to an SNS topic
  • subscribers receive broadcast copies
• Example use cases: fraud detection listening, analytics, fulfillment workflows.

Amazon SQS (Simple Queue Service)

• Queue for holding messages until consumers process them.
• Supports:
  • FIFO (order preserved)
  • Standard queues (best-effort ordering)
• Enables consumer-controlled processing rate to avoid backpressure.

Amazon EventBridge

• Message bus supporting many AWS events (with the note that SNS can’t receive EC2 lifecycle events the same way).
• Supports schemas/patterns for discoverability of message formats.
• Adds:
  • Rules for recurring events (cron-like)
  • EventBridge Scheduler for one-time/specific future execution and expiration

AWS Step Functions

• Workflow orchestration using distributed graphs:
  • sequential and parallel tasks
  • conditionals (if/else)
  • failure handling with alternate paths
• Positioned as a favorite workflow tool, integrated with other AWS services.

MWAA (Managed Apache Airflow)

• Workflow orchestration alternative.
• Open-source based; speaker prefers Step Functions but acknowledges Airflow for ecosystem preferences.

---

10) Data Processing and Analytics

Core storage for analytics: Amazon S3

• Used as raw data storage for processing pipelines.

Processing engines

• Amazon EMR: MapReduce/Hadoop/Iceberg-style processing in clusters
• Amazon Athena: runs SQL directly over S3 without provisioning clusters per query
• AWS Glue: serverless ETL/data preparation; includes Glue Studio for dataset/query readiness and metadata/crawling-like workflows
• Amazon Redshift: columnar data warehouse for large-scale analytics
  • Speaker cautions it’s expensive and best for heavy workloads.

BI / visualization

• Amazon QuickSight
  • Similar to Power BI: dashboards, widgets, pivot tables, BI insights
  • Integrates with Redshift or raw S3 data

Streaming ingestion + analytics

• Amazon Kinesis
  • Ingests massive streaming data via data streams
  • Uses Kinesis Firehose to batch/chunk into S3
  • Supports time-window analytics for telemetry/order streams

---

11) Monitoring, Logging, and Debugging

Amazon CloudWatch

• Main monitoring/logging tool.
• Features:
  • Logs + CloudWatch Logs Insights (SQL-like querying)
  • Dashboards (widgets based on metrics)
  • Alarms on thresholds triggering notifications/actions

AWS CloudTrail

• Audit logs of who/what performed actions on AWS resources (governance/audit)

AWS Config

• Continuous compliance/configuration monitoring:
  • detect scenarios (e.g., public S3 bucket)
  • alarm/alert when policy is violated

AWS X-Ray

• Request tracing/profiling:
  • traces across services to find bottlenecks and debug performance issues

---

12) CI/CD and Infrastructure as Code

CI/CD tools

• AWS CodeBuild: build (including unit tests) on commits
• AWS CodeDeploy: deploy to targets (EC2, Lambda, etc.)
• AWS CodePipeline: orchestrates pipelines (e.g., beta → gamma → prod)

Infrastructure as Code

• AWS CloudFormation
  • templates in YAML/JSON define resources for deployment
• AWS CDK (Cloud Development Kit)
  • more usable than CloudFormation
  • define infrastructure using general-purpose programming languages (Python/Java/Ruby mentioned)
  • uses reusable constructs and control flow

Developer productivity helpers

• AWS Amplify
  • “AWS for dummies” / guided app setup via Amplify CLI (creates auth, hosting, infra pieces)
  • Warning: later may require manual migration when users outgrow the abstraction
• AWS AppSync: managed GraphQL backend/resolvers
• Amazon WorkSpaces: end-user computing/workspace access via VPN
• Amazon Q: IDE integration for autocomplete/chat; can also answer about AWS resources
• AWS MCP (Model Context Protocol; described as a feature): LLM can query AWS resources, with a warning it could create security headaches

---

13) Identity and Access Management + Networking

IAM (Identity and Access Management)

• Controls permissions via users and roles.
• Roles can be assumed by services (example: Lambda execution role).
• Uses policy statements attached to identities.

AWS IAM Identity Center

• Connects external identity providers (e.g., Okta/Google Workspaces) to AWS accounts.
• Maps identities to roles/groups/accounts.

Networking

• Amazon VPC
  • isolate resources in a secure “bubble” inside AWS
  • reduces blast radius and prevents default cross-talk
• VPN
  • secure remote access into VPC via tunnel
• PrivateLink
  • enables secure communication from within VPCs to AWS services/resources despite VPC constraints

---

Main speakers / sources

• Main speaker: the YouTube creator (subtitles reference “my course” and a link to courses.beabetterdev.com; also mentions beabetterdev.com).
• No other distinct speaker is identified in the subtitles.