principle-distributed-systems

What problem does it solve?

Distributed systems fail in partial, non-deterministic ways, so teams need principled guidance to make correct trade-offs across consistency, availability, latency, and failure modes.

Core Features & Use Cases

• CAP and PACELC decision-making: explicitly choose consistency vs availability under partitions, and name the latency vs consistency trade-off in normal operation (PACELC).
• Correct consistency model selection: compare linearizable, causal, read-your-writes, monotonic reads, and eventual consistency to match actual correctness needs.
• Ordering, clocks, and delivery semantics: replace unsafe wall-clock ordering with Lamport/vector/HLC concepts, and distinguish exactly-once delivery myths from exactly-once effects via idempotency and atomicity.

Quick Start

Use principle-distributed-systems to decide whether your system should be CP or AP under network partitions and what consistency model best matches your application’s correctness requirements.