Partitioning

Kind: technique

Layer: Performance Core

Record: architecture:partitioning

Severity: contextual

Scope: data, workload, service

Canonical: Ontology

A technique for dividing data or work into independent partitions by a key, so each can be processed in parallel.

Listed in Architecture principles, after Sharding and before Caching.

Requires

• Partition Strategy

Reinforces

• Scalability
• Isolation

Enables

• Parallelism

Conflicts with

• Global Shared State

In tension with

• Rebalancing Complexity

Tensions

• Partitioning / Rebalancing Complexity

Violated by

• Single Monolithic Store

Severity

• contextual

Category

• Scalability / Performance / Optimization

Refactors

• Event Ordering
• Sharding

Linked from

• A concern is a component
• Causality / Ordering / Distributed Time
• Scalability / Performance / Optimization
• Transactions / State / Concurrency
• Scalability / Performance / Optimization
• Severity levels
• The resolutions