AWS Disaster Recovery Options
A useful comparison chart of DR options from a CloudAcademy AWS Architect (Associate) Certification course:
Traditional DR vs DR on Public Cloud
click on image for enlarged view |
Options | Backup & Restore | Pilot Light | Warm Standby | Multi Site |
Description | Like using AWS as a Virtual Tape Library | Minimal version of environment running on AWS | Scaled down version of fully functional environment always running | Fully operational version of fully functional environment always running off site or in another region |
Services Used | AWS Storage Gateway, Import/Export, Glacier, S3 | AMIs, ELBs, CloudFormation, RDS replication | AMIs, ELBs, CloudFormation, RDS replication | All |
RTO |
High (8-24 hours) |
Moderate (4-8 hours) |
Minimal < 4 hours |
Lowest < 60 minutes |
RPO | Since the last backup; up to 24 hours | Since the last snapshot. While core pieces of system are in place, some installation and preparation may be required. | Since the last data write if a master / slave multi-AZ DB. May be asynchronous only which would increase the RPO | Choice of data replication influences RPO |
Cost considerations | Low Recovery time may involve getting tapes/media delivered to site Disk/tape management |
Low Keeping all services/ libraries / patches up to date adds an administrative overhead |
Medium Environment can be used for dev/test off setting cost |
High The ongoing cost of maintenance / operation needs to be factored in. |
Traditional DR vs DR on Public Cloud
- Update Status Page
- Restore Datastore(s) in prodY from latest prodX
- DB
- Authentication
- Authorization
- Cache
- Blob Storage
- Restore backend microservices
- Bootstrap services with particular focus on upstream and downstream dependencies
- Swap CloudFront distribution(s)
- Swap API endpoint(s) via DNS
- Update DNS records to point to prodY API endpoints
- Verify recovery is complete
- Redeploy stack from user account to verify service level
- Update Status Page
Comments
Post a Comment