Database

RDS (Default)

• Auto scale the size of storage when the free storage is almost exceeded

• Create read replicas and the replication is async

• To sync the replication to other instance of other AZ for failover and increase availability

• Can enable daily up and restore at any point of time or create DB snapshot manually to restore data onto new DB instance

RDS (Custom)

• To deactivate automation mode

• Gain full access for underlying database and EC2

Aurora

• Support PostgresSQL and MYSQL

• Better performance

• 6 copies of data across 3 AZ

• Up to 15 Read Replicas serve read

• Automatically Failover

• Restore data at any point in time without backups

• On Aurora serverless, no capacity needed to be planned, for unpredicted workload

• Multiple master can be defined

• On Aurora global, up to 16 read replicas on each region

• There is reader endpoint for load balancing when perform read to read replica

• it can define custom endpoint for specific type of read replicas

• it is faster to use aurora cloning compared with snapshot & restoring

RDS Proxy

• Reducing the stress and minimize the open connection in order to handle lots of request

Snow Family

• Size: SnowMobile (PB) > Snowball edge > Snowcone

AWS Storage Gateway

• A bridge between on premise data and cloud data

Dynamo DB

Introduction

• NoSQL db

• Low cost and auto scaling

• 2 modes for managing table capacity (read/write throughput)

• Provision mode: specify number of read/write per second

• On-demand mode: auto scale up and down based on the workload, more expensive

Dynamo DB Accelerator (DAX)

• in-memory cache for dynamo DB

Stream

• Order stream of item-level modification (CRUD) in the table

• Use cases: react to the db changes, real time analysis ...

DocumentDB

• MongoDB

Neptune

• Graph database respresenting relationship