Stale Reads and Cascading Failures in Distributed Systems

TLDR: Stale reads return superseded data from replicas that haven't yet applied the latest write. Cascading failures turn one overloaded node into a cluster-wide collapse through retry storms and redi