Consistency models

How this question breaks down

1. 1Strong vs eventualCore19 minWhat a read is allowed to return once data lives on more than one machine, from "always the latest write" to "the latest write, sooner or later".
2. 2CAPCore16 minWhen the network splits, a replicated store must refuse some requests or answer some of them wrongly. When it doesn't split, every extra guarantee still costs latency.
3. 3Read your writesCore18 minFour cheap promises (read your writes, monotonic reads, monotonic writes, writes follow reads) that make an eventually consistent store feel consistent to each user.
4. 4QuorumsCore19 minWrite to W of N replicas, read from R, and choose R + W > N so every read meets at least one copy of the latest write, then learn where that promise stops.

Work through the parts in order and you’ve covered the full design. Each part is one topic.