Disaster Recovery Plan
In this article, we describe the main disaster recovery strategies for the Regula IDV Platform and provide guidance on how to implement them effectively.
Four principal disaster-recovery (DR) strategies are commonly used:
- Backup and Restore – Take regular backups of data and applications and restore them after an incident.
- Pilot Light – Keep a minimal, always-on core of the environment while other components remain offline until needed.
- Warm Standby – Maintain a reduced-capacity but fully functional environment that can be quickly scaled to full production.
- Active-Active – Operate multiple fully functional environments in different locations and route traffic between them in real time.
Backup and Restore
Pros
- Simple to implement and manage.
- Cost-effective for small environments.
- Suitable for both data and application recovery.
Cons
- Long recovery-time and recovery-point objectives (RTO/RPO) – recovery can take hours or even days.
Pilot Light
Pros
- Faster recovery time than Backup and Restore.
- Cost-effective for small to medium environments.
Cons
- Requires more resources than Backup and Restore.
Warm Standby
Pros
- Faster recovery time than Pilot Light.
- Suitable for both data and application recovery.
Cons
- Requires more resources than Pilot Light.
- More complex to implement and manage.
- Higher cost than Pilot Light.
Active-Active
Pros
- Lowest RTO/RPO – near-instant failover.
- Can handle large-scale disasters.
- Provides high availability and load balancing.
- Supports both data and application recovery.
Cons
- Most complex to implement and manage.
- Requires significant resources and infrastructure.
- Highest cost of all strategies.
Implementation Guidance
Regula provides an on-premises solution that can also be deployed in cloud or hybrid environments. Because it is not cloud-native, the DR strategy must be implemented at the infrastructure level and remains the customer’s responsibility.
Recommendations:
- All components of the IDV solution are horizontally scalable; therefore any of the DR strategies above can be implemented.
- Choose a strategy that matches your recovery objectives, budget, and operational capacity.
- If you can leverage cloud-native services, Active-Active offers the best performance and availability.
- When budget or resources are limited, start with Backup and Restore and evolve to Pilot Light or Warm Standby over time.
Stateful-Component Replication
If you select Pilot Light, Warm Standby, or Active-Active, replicate stateful workloads between the primary and secondary sites:
- MongoDB database – Use replica sets to replicate data; see the MongoDB replication documentation.
- Object storage – Replicate file storage (for example, Amazon S3 or an on-premises S3-compatible platform such as MinIO) using the storage-provider’s native replication or a custom mechanism.
- Vector/Text store – IDV uses either MongoDB Atlas or OpenSearch; both provide built-in replication. See the MongoDB Atlas or OpenSearch documentation.
- RabbitMQ message broker – Enable RabbitMQ clustering or federation; refer to the RabbitMQ cluster guide.
Stateless-Component Deployment
All other IDV components are stateless and can be deployed in multiple sites by running additional instances and fronting them with load balancers:
- Web Server
- Workflow Service
- Audit Service
- Indexe