Created : March 30, 2020

This note was taken while learning AWS on Udemy.

RDS (Relational Database Service)

  1. SQL
  2. MySQL
  3. PostgreSQL
  4. Oracle
  5. Aurora
  6. MariaDB

Amazon's NoSQL = DynamoDB

  • RDS runs on virtual machines
  • cannot log in to these OS
  • patching of RDS OS & DB is Amazon's responsibility
  • RDS : NOT serverless

    • only Aurora Serverless is serverless


  • OLTP (Online Transaction Processing)

    • pull up a single row of data
    • like name, date, address, status
    • RDS is suitable for OLTP
  • OLAP (Online Analytics Processing)

    • pull in large numbers of records
    • run queries against your database
    • like net profit for product, sum of products sold in Pacific
    • Redshift is suitable for OLAP

Data Warehousing

  • used for business intelligence
  • used to pull in very large and complex datasets
  • usually used by management to do queries on data
  • use different type of architecture both from a database perspective & infrastructure layer

    • Amazon's data warehouse solution = Redshift


  • easily deploy/operate/scale in-memory cache in cloud.
  • to speed up performance of existing databases (frequent identical queries)
  • a way of caching your most common web queries

    • so when your web servers get a request,
    • they'll go to ElastiCache first, and get information
    • it takes hugh load off of your database
  • improve performance of web applications by allowing you to retrieve information from fast, managed, in-memory caches.
  • when your databases overloaded, what to do to perform better

    • choose ElastiCache or Read replicas
  • support two in-memory caching engines

    • Memcached

      • for something very simple
      • when you need to scale horizontally
    • Redis

      • for more advanced data types
      • Multi-AZ
      • can do backups and restores

Backups for RDS

Automated Backups

  • recover database to any point in time within a retention period

    • retention period = 1~35days
  • take a full daily snapshot & store transaction logs throughout day
  • enabled by default
  • backup data is stored in S3

    • free storage space same size of database
    • if you have 10GB RDS, you'll get 10GB free storage in S3
  • during backup, storage I/O may be suspended

Database snapshots

  • done manually
  • stored even after you delete original RDS instance
  • Restoring Backups : restored version of database has

    • new RDS instance
    • with new DNS endpoint

Multi-AZ vs Read Replicas


  • for disaster recovery only
  • have exact copy of your production database in another AZ
  • if one AZ fails - update DNS settings and RDS endpoint automatically

Read replicas

  • for performance
  • must have automatic backup turned on
  • up to 5 read replica copies of any database
  • can have read replicas of read replicas.
  • each read replica have its own DNS endpoint
  • can be Multi-AZ
  • can be in different regions


  • a fast & flexible NoSQL database service
  • stored on SSD storage
  • spread across 3 geographically distinct data centres
  • Eventual Consistent Reads (default)

    • consistency across all copies of data is usually reached within a second.
    • repeating a read after a short time should return the updated data.
  • Strongly Consistent Reads

    • returns a result that reflects all writes that received a successful response prior to the read
    • read data within or less than a second


  • fast data warehousing service in the cloud
  • used for business intelligence
  • only available in 1 AZ
  • columnar data storage

    • advanced compression : easily compressed much more than row-based system
    • because similar data is stored sequentially
  • MPP (Massively Parallel Processing)

    • Redshift automatically distributes data and query load across all nodes
    • makes it easy to add nodes to your data warehouse
    • enables you to maintain fast query performance as your data warehouse grows.
  • Backups

    • retention period : 1day(default) ~ 35days
    • always attempts to maintain at least 3 copies of data

      • original & replica : on the compute nodes
      • backup : in S3
    • can replicate snapshots to S3 in another region for disaster recovery
  • charged by

    • compute node hours

      • not be charged for leader node hours
    • backup
    • data transfer - only within a VPC, not outside it


  • Single node : 160GB
  • Multi node

    • Leader node : manage client connections & receive queries
    • Compute node : store data & perform queries, computations (up to 128 compute nodes behind a leader node)


  • a MySQL and PostgreSQL-compatible relational database built for the cloud
  • start with 10GB, scale in 10GB increments to 64TB (storage autoscaling)
  • compute resources can scale up to 32vCPUs & 244GB memory.
  • 2 copies of data in each AZ, with minimum of 3 AZ. = total 6 copies

    • loss in a couple of AZ - no affect on performance
  • 3 types of replicas available

    • Aurora replicas

      • can have up to 15 replicas
      • automated failover is only available with Aurora replicas
    • MySQL replicas - can have up to 5 replicas
    • PostgreSQL replicas
  • backups

    • automated backups are always enabled
    • can take snapshots - can share with other AWS accounts
    • backups don't impact on database performance

Aurora Serverless

  • an on-demand, autoscaling configuration for MySQL, PostgreSQL-compatible editions of Aurora.
  • DB cluster automatically starts up, shuts down, and scales capacity up or down based on your application's needs.
  • use if you want simple, cost-effective option for infrequent, intermittent, or unpredictable workloads.

    • when you don't know when people will access your website.
  • pay on a per invocation basis

    • pay only when someone's actually accessing your website (not by hours)