Change data capture, dual-write patterns, eventual consistency, and keeping distributed financial data in sync across systems — the patterns that survive partial failures.
The dual-write problem and how to solve it: change data capture with Debezium, the transactional outbox pattern, idempotent downstreams, schema drift, and the reconciliation that catches the edge cases.
Active/active vs active/passive multi-region data: conflict resolution strategies, Kafka MirrorMaker 2 with offset translation, Postgres streaming and logical replication, IBM Db2 HADR, and the SAMA data-residency constraints that shape all of it.
Daily reconciliation that catches CDC drift the application can’t see: fingerprinting & hash comparison at scale, automated break investigation and classification, SAMA-aligned audit trails, and the tamper-evident reporting that survives an examination.
Kafka ACLs for domain data ownership, Avro schema with PDPL classification metadata, Kafka Streams GlobalKTable joins across domain boundaries, and the governance model that turns Kafka from a shared bus into a set of domain-owned data products.
CockroachDB geo-partition survival goals, PostgreSQL conflict resolution trigger with last-write-wins and merge strategies, Redis 7 PN-Counter Lua for distributed counters, and the SAMA BCP requirements that make multi-master necessary.