Disaster Recovery (DR) is paramount in financial systems, where operational continuity is non-negotiable. Regulatory bodies and exchanges mandate stringent guidelines to ensure resilience and reliability. At Angel One (A1), we've developed a sophisticated DR strategy that ensures our trading platform remains robust against disruptions. Our system architecture comprises a four-layered stack: from the Exchange at the base to RMS, Middleware, and applications at the top. This structure is designed for live-live state synchronization across DR sites and the main site, maintaining order, trade, margin, and other critical data in real-time with eventual consistency.

The activation of DR involves meticulous reconciliation at each layer, ensuring data consistency with the Exchange as the ultimate source of truth. A1's proactive approach includes a detailed playbook and regular DR drills, independent of or in alignment with Exchange DR schedules. Our automated DR process encompasses service recovery, load balancers, DNS adjustments, post-DR checks, and user communication, all executable with a single click. This talk will delve into the flow of order placement and execution, the complexities of our DR setup, compliance adherence, network topology, critical components, and the trade-offs involved in state rebuilding during DR scenarios. Attendees will gain insights into key metrics, typical DR workflows, and real-world DR performance metrics, including time frames and data loss figures.