WhiteBooks at Scale

WhiteBooks processes 30+ Crore IRNs per year, peaks at 1.6 million IRNs in 24 hours, and sustains p50 IRN latency of 87 milliseconds. The architecture is multi-tenant, horizontally scalable, idempotent on the request envelope, and built for ERP-grade concurrency. Zero-downtime deployments ship multiple times per week without an SLA-impacting outage in the last 18 months.

Headline numbers (rolling 90 days)

MetricValue
30-day uptime (production gateway)99.994%
90-day uptime (production gateway)99.991%
p50 IRN latency (24h)87 ms
p95 IRN latency (24h)312 ms
p99 IRN latency (24h)892 ms
Peak IRNs / 24h1.6 M
Annual IRN volume30+ Crore
Active customers5,000+
Cities reached8,000+

Scale patterns we engineer for

Zero-downtime deployments

Blue-green deploys via Cloudfront-managed traffic shifting. New version receives 1% of traffic for 15 minutes, then 25%, then 100%. Each step gated by error-rate + latency checks. If any check trips, traffic shifts back to the previous version automatically. The customer never sees a deployment.

Customer-side load patterns we support

Part of the WhiteBooks Enterprise cluster. Last updated 27 May 2026.