Skip to content

Rapid Battle Cards 21–30

21 — NoSQL: Document vs Key-Value vs Column vs Graph

  • Document: self-contained document structure.
  • Key-value: retrieve opaque/value structure by a known key.
  • Column-oriented: organize sparse/large data by columns/families.
  • Graph: nodes + variable connections/traversal.
  • Clue: match storage/access shape to the requirement.

22 — Conceptual vs Logical Model

  • CDM: high-level concepts, relationships, scope, vocabulary.
  • LDM: detailed technology-independent entities, attributes, domains, keys.
  • Clue: high-level business scope vs detailed requirements.

23 — Logical vs Physical Model

  • LDM: technology-independent.
  • PDM: DBMS/storage-specific implementation.
  • Clue: vendor datatypes, indexes, partitions, physical naming → PDM.

24 — Canonical Model vs Database PDM

  • Canonical: shared message/payload structure for data in motion.
  • PDM: stored structure in a chosen database technology.
  • Clue: interface/message vs stored database.

25 — View vs Materialized View

  • View: virtual, computed/retrieved on demand.
  • Materialized: instantiated/stored result refreshed at chosen times.
  • Clue: on-demand vs precomputed/stored.

26 — Vertical vs Horizontal Partition

  • Vertical: split columns.
  • Horizontal: split rows by values/ranges.
  • Clue: columns vs rows.

27 — Normalization vs Denormalization

  • Normalize: remove redundancy/dependency defects and stabilize logic.
  • Denormalize: intentionally add/combine redundancy in physical design for justified performance/security need.
  • Clue: logical cleanup vs controlled physical redundancy.

28 — 1NF vs 2NF vs 3NF

  • 1NF: atomic/no repeating groups/valid key.
  • 2NF: full dependency on complete minimal key.
  • 3NF: no non-key dependency on non-key.
  • Clue: atomicity → full-key dependency → key-only dependency.

29 — Generalization vs Specialization

  • Generalization: move common features up to supertype.
  • Specialization: move distinguishing features into subtypes.
  • Clue: common upward vs differences downward.

30 — Forward vs Reverse Engineering

  • Forward: requirements → CDM → LDM → PDM.
  • Reverse: existing DB → PDM → LDM → CDM.
  • Clue: starting evidence — requirements or implementation.

Source anchors: current Battle Cards, cards 21–30.

← Cards 11–20 · Cards 31–38 →