Quality scenarios, the seven design decision categories, and the tactic catalogues for performance, availability, security, modifiability, testability, usability and interoperability.
03chapter
—lists
—total items
Every enumerable list from this chapter, one card each. Search filters the cards live;
the chips above jump to any other chapter in the vault.
03
Quality Attributes
18 lists
Six-Part Quality Scenario 6
Source — who or what triggers it
Stimulus — what arrives
Environment — the system state at the time (normal, peak, degraded)
Artifact — the part of the system that is hit
Response — what the system does
Response measure — how the response is measured
General vs Concrete Scenario 2
General — reusable, system-independent template
Concrete — instantiated for this system with real numbers
Seven Categories of Design Decision 7
Allocation of responsibilities — where behavior and knowledge live
Coordination model — how elements interact
Data model — what is stored, where, partitioning, replication and consistency
Management of resources — threads, memory, connections, bandwidth
Mapping among architectural elements — what scales, fails or releases independently
Binding time — compile time, build, startup or run time
Choice of technology — flexibility traded against complexity and cost
Performance — Concerns 5
Latency
Throughput
Capacity
Deadlines
Resource use
Performance — Tactics 2 groups
Control demand
Sampling / manage sampling rate
Admission control and rate limiting
Prioritize events
Reduce overhead
Bound execution times
Manage resources
Increase resources
Introduce concurrency
Replicate computation or data
Bound queue sizes
Schedule resources
Performance — Measurement Rules 3
Measure percentile latency (p95, p99), not averages
Measure under representative load, not a quiet machine
Keep headroom — latency rises sharply as utilization nears capacity