🗓️ 29042026 1600
📎 #distributed_systems #moc
Map of Content for the distributed-systems concept cluster. Walk top-to-bottom; if a concept resists self-explanation in two minutes, re-read the zettel.
Concept Order
Foundations
- cap_theorem — the trilemma framing for trade-offs under partition.
- consistency_models — what "C" actually means at finer granularity.
- quorum_reads_writes — N/W/R as the lever between CP and AP.
- consistent_hashing — sharding scheme that minimises movement on N changes.
- raft — consensus protocol; concrete instance of a CP system.
Extensions (planned)
pacelc_theorem— adds latency-vs-consistency to the steady-state picture.paxos_basics— peer to Raft, harder to implement, foundational to ZAB.linearizability_vs_serializability— single-object real-time vs multi-object schedule equivalence.vector_clocks_lamport_clocks— causal tracking and monotonic IDs.
Edge cases (planned)
gossip_protocol— eventually-consistent membership / state dissemination.crdt_basics— conflict-free replicated data types for AP-leaning merges.
How to Use This MOC
- First pass: read each zettel back-to-back, take no notes, just absorb.
- Second pass: walk this index, write a 3-bullet recall for each entry from memory, then compare to the zettel. Gaps point to where the concept hasn't landed.
- Application drill: take a system design prompt (e.g. "design a globally-replicated KV store") and try to weave 5+ of these zettels into the answer naturally. The connections matter as much as the individual concepts.
- Tag hygiene: zettels here are tagged by topic (
#distributed_systems,#consistency,#sharding). Drop#wiponce a zettel is revised and stable.
Bridges to Other Domains
- → DB internals: ACID, concurrency_control, mvcc_innodb_read_view, transaction_isolation_levels.
- → Caching: redis_cluster,
redis_distributed_lock(planned). - → Messaging: exactly_once_semantics,
kafka_architecture(planned). - → System design templates:
system_design_framework(planned),back_of_envelope_estimation(planned).