SKILL: Designing Data-Intensive Systems (Kleppmann)

Source: Designing Data-Intensive Applications by Martin Kleppmann Domain: Distributed systems, data architecture, reliability engineering Applies to: Building systems where data complexity (not computation) is the bottleneck

DECISION POINTS

1. Consistency Model Selection

IF: Strong consistency required (banking, inventory)
  AND: Can tolerate higher latency + coordination overhead
THEN: Use linearizability with synchronous replication

IF: Operations need ordering but not global agreement
  AND: User experience matters more than strict consistency
THEN: Use causal consistency (preserve cause-effect, allow concurrent ops)