Latency

Kind: metric

Layer: Performance Core

Record: architecture:latency

Severity: contextual

Scope: API, service, user flow

Canonical: Ontology

A measure of the time between a request and its response, usually reported at percentiles.

Listed in Architecture principles, after Throughput and before Performance Engineering.

Requires

• Time Budget

Reinforces

• User Experience
• Performance Engineering

Enables

• Responsiveness

Conflicts with

• Long Blocking Work

In tension with

• Throughput/Batching
• Total-Order Broadcast
• CAP Theorem
• Consensus
• Message Queue
• Throughput
• Consistency
• Pessimistic Locking
• Total-Order Broadcast / Latency
• CAP Theorem / Latency
• Consensus / Latency
• Message Queue / Latency
• Throughput / Latency
• Consistency / Latency
• Pessimistic Locking / Latency

Tensions

• Latency / Throughput/Batching

Violated by

• Long Blocking Work

Refactored by

• Caching
• Introduce Async Event
• Add Index

Severity

• contextual

Category

• Scalability / Performance / Optimization

Reinforced by

• PACELC Theorem
• CDN / Edge Caching
• Queuing Theory

Linked from

• Causality / Ordering / Distributed Time
• Control / Coordination / Centralization
• Event / Messaging / Asynchronous Architecture
• Model Architecture
• Scalability / Performance / Optimization
• Transactions / State / Concurrency
• Core Vocabulary
• Event Messaging Async
• Quality Attributes
• Scalability / Performance / Optimization
• Severity levels
• The resolutions