AWS Services

IAM

• Setup user and roles
• User can login to AWS
• Roles have permissions to different AWS services or even different rows in a database
• Attach user to roles

Identity Center

• Links identity providers (Google workspace, Okta, etc) to AWS
• And link those user to different roles in IAM

Elastic Load Balancer

• Redirect traffic to EC2 instances

API Gateway

• Can act like a load balancer forwarding reqest to a server
• Can forward request to other AWS service directly without needing a server
• Can setup throttling, rate limiting, and user authentication and authorization (through Cognito)

Elastic Compute Cloud (EC2)

• Choose memory, OS, CPU and you have a virtual computer

Lambda

• EC2 runs a server and you pay for every second
• In Lambda you upload code and it runs on an event (like when file upload in S3, when user profile is created in Cognito)
• Scaling is handelled automatically
• You only pay for number of requests and run time

Elastic Container Registry

Elastic Container Service (ECS)

• starting, stoping and allocating virtual machines to run docker containers

Elastic Kubernetes Service (EKS)

• Gives you more control over your containers

Simple Notification Service (SNS)

• Can be used to spread events do application / different servers

Simple Queue Service (SQS)

• Queues
• No backpressure. The clients poll and process what they can

Step Function

• Graph of tasks/steps you want to do.

Simple Storage Service (S3)

• General purpose file storage
• To serve static files
• We need to choose the S3 server location

Glacier

• To store files for archival purposes

CloudFront

• To do regional caching of files hosted on S3
• Because directly serving from S3 is costly

Elastic Block Storage (EBS)

• Is like hard disks
• Can attach to one EC2 instance at once
• High throughput IO and consistency

Elastic File System (EFS)

• You can mount EFS to multiple servers at once
• Is performant like Block Storage
• Higher cost

Secrets Manager

• Store and provide secrets to EC2

App Config

• To store feature flags. EC2 instance have to poll, Lambda uses new feature flags automatically

Certificate Manager

• Ensure that the endpoint server/service on Route 53 has a valid certificate
• Issued by AWS or other services

Relational Database Service (RDS)

• Supports Postgressql, oracle, MySQL, …
• Better than using EC2 is it handles scaling, storage, … stuff

Aurora

• Similar to RDS but managed and faster
• Also supports a serverless variant

Dynamo DB

• key value database

Document DB (MongoDB)

• not mongoDB but is 1-1 mapping to mongoDB

Keyspaces (Cassandra)

• Managed version of Cassandra

Neptune

• Graph database

ElasticSearch (OpenSearch)

• based on apache lucene
• for document storage, and full text search capabilities

Timestream

• for timeseries datasets

Elastic Cache

• Supports Redis or MemCaches

MemoryDB

• Cache but data persists when the node is down

Redshift

• Data Warehouse (where we dump structured dataset from multiple sources)

Kinesis

• Real time data stream
• to use on stuff like Apache Spark using Amazon Map Reduce

MSK (Managed Streaming for Apache Kafka)

SageMaker

• For training ML models
• Provies a managed Jupyter notebook

Simple Email Service (SES)

Route 53

• It is a DNS, Domain Management and Traffic management service
• Can do health checks
• Can route DNS based on geographic location, latency

Cloud Watch

• Collect logs and events

Auto Scaling

• Uses logs from Cloud Watch to auto scale EC2 instances

Cloud Formation

• Infrastructure as Code for AWS