Topics for later

The following deep dives are still pending and would naturally extend this guide. They are distributed-systems theory that complements the operational content already here:

• Consistency models — linearizable, sequential, causal, eventual. The "periodic table" of distributed-system guarantees. Aphyr's chart is the canonical reference: jepsen.io/consistency. Touched on in the Replication chapter but deserves its own treatment.
• Consensus protocols (Raft, Paxos) — the "five nodes vote on a log" mental model; when you need consensus vs when you can skip it. The Raft paper is approachable: raft.github.io/raft.pdf. MIT 6.824 labs build it from scratch.
• Stream processing — event-time vs processing-time, watermarks, exactly-once semantics, stateful processing. The "Streaming 101" series by Tyler Akidau on O'Reilly is the best entry point. CDC and event sourcing are covered; stream-processing engines (Flink, Kafka Streams, Beam) deserve their own chapter.
• Time and ordering — Lamport timestamps, vector clocks, hybrid logical clocks, Google's TrueTime. Foundational for understanding why distributed-systems primitives behave the way they do. Lamport's original paper is 8 pages.
• Database-specific internals — Postgres (MVCC, autovacuum, WAL, replication slots), Redis (single-thread model, persistence, cluster mode), Kafka (segments, replication, exactly-once semantics).