Implementing Effective Disaster Recovery in Stock Brokerage
summary
Angel One runs a four-layered stack from Exchange to RMS, Middleware and applications with active‑active, live‑live sync across Mumbai, Chennai and AWS using Redis for hot caches and Kafka for eventual data lakes. DR is one-click DR with DNS flip, manual adapter re-login to the alternate Exchange site, Airflow/GitHub Actions for config and DB failover, and a Recon orchestrator that pulls an exchange dump to reconcile orders because exchange is source of truth. Recovery is 15-minute tested and accepts some TCP-order data loss requiring order replays.
AI-written summary · about
chapters (8)
- 0:00Introduction to disaster recovery
- 2:27System-level disaster recovery challenges
- 5:18Brokerage system integrations overview
- 7:28Multi-data center sharding strategy
- 9:13Disaster recovery playbook principles
- 13:28Executing the disaster recovery process
- 16:56Post-recovery customer communication
- 17:59Observability and monitoring in DR
related talks
44mA Real-World Disaster Recovery Case
05mDisaster Recovery and Contingency Planning for Platforms
25mAutomating Disaster Recovery: The Ultimate Reliability Challenge
14mThe Multi‑Workload Challenge: Predictable Failover
07mLightning Talk: Mission: Impossible - Building Self-Healing Disaster Recovery With...
16mResilient by Design: Architectures for Crisis Response & Recovery
34mBuilding High Resiliency into Digital Solutions
14mDisaster Recovery preparedness