Distributed Systems
Eventual Consistency in Practice
How distributed systems converge to a consistent state over time, and the patterns that make it workable.
Eventual consistency guarantees that, if no new updates are made, all replicas will eventually return the same value. It sacrifices immediate global agreement in favour of availability and low latency. The system converges through background processes like read-repair, hinted handoff, and anti-entropy reconciliation.
Continue reading
- →
Dependency Injection and IoC Containers
Invert control of dependency creation to keep classes decoupled from their collaborators and trivially testable.
- →
The Observer Pattern and Event-Driven Design
Decouple publishers from subscribers by letting objects register interest in events without the producer knowing who is listening.
- →
The Strategy Pattern: Encapsulating Algorithms
Define a family of algorithms, encapsulate each one, and make them interchangeable — open for extension, closed for modification.