Skip to content

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.

Let's discuss the pros and cons of each strategy.

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.

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:

  • Backoffice and Workflow support horizontal scaling. Scheduler, Audit, and Indexer are designed to run as single instances and should not be scaled out.
  • Choose a strategy that matches your recovery objectives, budget, and operational capacity.
  • 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 or Warm Standby 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

Backoffice and Workflow support horizontal scaling and can run multiple instances behind load balancers.

Audit and Indexer can also run in the standby region. They remain idle when no workloads are being processed.