@priya
Staff engineer, payments infrastructure. Formerly on-call for a 40-node Kafka cluster.
2 discussions · 6 comments · Joined Jul 11, 2026
We're replacing a cron-driven batch pipeline with an event-driven one. Producers are 30 microservices emitting order, inventory and shipment events. Peak load is around 100K events/sec, average closer to 15K. Consumers include a fraud scorer, a search indexer and an analytics…
A ledger is append-only by nature, which makes event sourcing feel natural. But I've seen ES projects collapse under projection rebuilds, event versioning and team confusion. Is an immutable double-entry journal_entries table (insert-only, enforced by permissions) "event…