Consistency is a Spectrum in Practice
What you'll learn: Real distributed systems don't just choose "consistent" or "not consistent"—they offer tunable consistency levels you can dial up or down.
The Reality of Consistency Levels
When we learned about CAP theorem, it might have seemed like systems are either consistent (CP) or available (AP). But in practice, consistency isn't binary—it's a spectrum with multiple levels you can choose from based on your needs.
Think of consistency like a thermostat, not a light switch. You don't just pick "hot" or "cold"—you can set it to 68°, 72°, or 75° depending on what you need right now.
Common Consistency Levels
Strong consistency guarantees that after a write completes, all subsequent reads see that value immediately. Every reader sees the same data at the same time. This is the strictest level—like everyone reading from the same single source of truth.
Eventual consistency promises that if no new writes occur, all replicas will eventually converge to the same value. Reads might temporarily see stale data, but things will sync up "eventually." Think of it like gossip spreading through a group—everyone learns the news, but not all at the same moment.
Causal consistency sits in the middle: if operation A causally influences operation B, everyone sees A before B. Related events stay in order, but unrelated ones can appear differently to different observers.
Systems like Cassandra let you tune these levels per-query: use strong consistency for financial transactions, eventual for social media likes.
Key Takeaway: Modern distributed systems offer tunable consistency levels—strong, eventual, and causal—allowing you to trade consistency strength for availability and performance based on each operation's requirements.