AWS Classroom Series – 28/Oct/2021

High Availability(HA) on Amazon RDS

  • It is important to have HA architecture in place for database as it is heart of every thing. (If database goes down, everything goes down).

  • Lets evaluate the various architectures of RDS

  • Simplest Architecture: Single AZ Deployment:

    • If we need to do some sort of proof of concept or deploy development environments or deploy non critical non production environments, we may not need a HA architecture, as we can live with downtimes.
    • Therefore in the previously mentioned scenarios, we can launch the RDS instance in the single AZ.
    • With this, you get a single RDS instance inside the vpc with necessary storage as shown below
  • High Availability: Multiple Azs

    • If we are trying to run a mission critical database, want to have an architecture where we cannot afford to loose data, have a tight recovery point object or can’t afford much downtime, we must deploy the database in multi-AZ architecture
    • When we deploy a database in multi-az architecture, we can choose the AZ’s where we want our primary db instance. RDS will choose the standby instnace & storage in other AZ of the same Region.
    • The instance running in the standby will be of same type as your master and storage will be of same configuration & size as primary
    • In case of multi-az, primary db is known as master database, handles all the traffic. The stand by database is always kept ready and is in a state that whenever the master goes down, it takes the role of master or primary and supports its application.
    • Failover Scenario:
      • For example, the host instance can go down & underlying storage can fail, the network connectivity to the primary instance is lost, the AZ itself goes down
      • When the failover happens, the standby is automatically propagated to the master and all the application traffic fails over to the new master
      • In multi AZ architecture of RDS, the application connects to the Database server using a DNS endpoint that is mapped to the master and standby instances.
      • As a result we don’t have to repoin the application to new master or change anything from application side
      • RDS automatically does the DNS failover which typically takes 30 to 60 seconds
    • Lets create a Multi AZ mysql rds db instance:
    • Lets try to connect to the database from mysql workbench as it is publicy accesible (Not recommended for production databases)

Leave a ReplyCancel reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Please turn AdBlock off
Animated Social Media Icons by Acurax Responsive Web Designing Company

Discover more from Direct DevOps from Quality Thought

Subscribe now to keep reading and get access to the full archive.

Continue reading

Exit mobile version
%%footer%%