Exploring AWS !!

Day 83:

Disaster Recovery — Part 3

If a particular region goes down, we have active/passive strategies in AWS implemented using Multi-region Disaster Recovery, where active is the actual active region which is up and running, and passive is the standby.
We also have active-active, where both regions are up and running, in case of failure the

Business Continuity plan:
Before going for any Disaster Recovery plans, you need to perform a Business Impact Analysis (BIA), what are the consequences for your organisation if this workload fails, you ned to identify how quickly you need to recover the data, how much data loss in time (not in GB or TB), how much data loss can be tolerated. You also need to do risk assessment, like what is the risk of a natural disaster and also geographical impact of these disasters. So, we need tot hink about this before planning the DR strategy and even before planning to build application.

Recovery Objectives:
How much data can you afford to recreate or lose? — Recovery Point Objective (RPO)
How quickly must you recover, and what is the cost of downtime ? — Recovery Time Objective (RTO)

RTO is maximum amount of acceptable delay between the interruption of service and restoration of service.
RPO is maximum amount of data that can be afforded to be lost or recreated.

Disaster Recovery is different in cloud:
DR strategies evolve with technology
— Single AWS Region
Risk of disruption or loss of one data centre
Implement High Availability workload
Do not forget backups
— Multiple AWS Regions
Risk of disruption or loss of multiple data centres
Implement cross region availability
Do not forget backups

Strategies for Disaster Recovery:
Backup & Restore
Pilot Light
Warm Standby
Multi-Site active/active