AWS Disaster Recovery part 1
Introduce
This post is a study for the following articles.
Objectives of DR(Disaster Recovery)
RTO(Recovery Time Object)
The maximum allowable delay between service interruption and service restoration.
This determines the acceptable duration of service downtime.RPO(Recovery Point Object)
The maximum amount of time you can tolerate data loss in the event of a failure after the last time data was recoverable.
Scope of impact for a disaster event
Multi-AZ Strategy
Multi-Region Strategy
Type of DB Strategies
Active/Active Pattern
Active/Active DR strategies are active at the same time with all systems and services recovered.
Active/Passive Pattern
An Active/Passive DR strategies primarily consists of a primary site (Active) and a backup site (Passive).
DB Strategies
Backup & Restore : RPO/RTO Hours
Pilot Light : RPO/RTO 10s of minutes
Warm Standby : RPO/RTO Minutes
Multi-site active/active : RPO/RTO Real-time
Case 1 : Backup & Restore
Use AWS CloudFormation, Cloud Development Kit Tools
The backup and recovery strategy has the highest RTO compared to other strategies, which means it takes a lot of times to recover.
To compensate for this, you can use AWS resources like Amazon EventBridge to improve detection and recovery scenarios for disaster situations, reducing and automating recovery time.
Case 2 : Pilot Light
With Pilot Light strategy, data is replicated to the region to other region or AZ(availability zone) to be recovered in real time.
But the region or AZ to be recovered is ideal, with the data store and database up to data.
Prepared compute resource is cold standby state.
Case 3 : Warm Standby
Prepared compute resource is warm standby state.
Case 4 : Multi-site Active/Active
All compute resources is activated.