Stackbook Logo
← Back to Syllabus
Day 7Level 1 — Core Mental Models

Consistency Models: Strong, Eventual, Causal, Read-Your-Writes

Learn: 45 minMap: 30 minDesign: 60 min

Core Concepts (5)

  • 1.Strong consistency: linearizability — single copy illusion
  • 2.Eventual: all replicas converge, but when? conflicts possible
  • 3.Causal: if A causes B, all nodes see A before B
  • 4.Read-your-writes: session guarantee — your writes visible to you
  • 5.The consistency spectrum is a cost/latency/availability trade-off

Map to Your Experience

Your system has a user profile service. What consistency model does it need? What happens if you relax from strong to causal? What breaks?

Design Brief: Consistency Model Selection for a Feature

Requirements

  • Identify feature and its consistency needs
  • Specify exact consistency model required
  • Analyze impact of weakening by one level
  • Document conflict resolution if eventual

Constraints

  • Use real feature from your system
  • Be specific about anomaly scenarios

Must Cover

model-selectionjustificationweakening-analysisconflict-resolution

Trade-off Axis

consistency strength vs latency/availability/cost

Competencies Advanced

data state → L3reliability ops → L2

Ready to Design?

Open this brief in the Design Studio. Your constraints, requirements, and patterns will be pre-filled. Create a Design Record and get AI critique.

Open in Studio →