Disaster Recovery of Stateful Applications in a Multi-Cluster Environment

Explore disaster recovery strategies for stateful applications in multi-cluster environments through this conference talk. Learn about leveraging replication-capable storage systems like Ceph/Rook for cross-cluster disaster recovery, and discover the essential features needed in a multi-cluster control plane for seamless disaster recovery solutions. Gain insights into building robust disaster recovery solutions for cloud environments, understanding concepts such as RPO and RTO, and exploring tools like Open Cluster Management and DROrchestrator Ramen. Delve into topics including storage replication management APIs, dynamic provisioning challenges, and cross-cluster actions. Discover practical approaches to deploying, recovering, and relocating applications in multi-cluster scenarios, and explore future adaptations in the field of disaster recovery for stateful applications.

Intro
What is Ceph?
What is Rook?
Why Disaster Recovery?
Disaster Recovery and High Availability
RPO and RTO
An Ideal DR Solution
Example: Rook+Ceph Regional DR
Regional DR Goal - Relocate
Backups as Replication
Storage System Based Replication
Storage Replication Management APIs
Recovery Management
Dynamic Provisioning Conundrum
Cross Cluster Actions
Open Cluster Management - 101
Ensuing User Complexity: Deploy
Ensuing User Complexity: Recover
Ensuing User Complexity: Relocate
DROrchestrator: Ramen
Ramen: DRPolicy API
Ramen: DRPlacementControl API
Ramen: Operator Deployments
Future work and adaptations
Links and References