Reliability engineering for distributed transactions: consistent hashing, partitioning/replication, two-phase commit protocol, atomic commits, Dynamo architecture.
Covers all-or-nothing and before-or-after atomicity, write-ahead logging, undo/redo logging, crash-consistent file systems, distributed transactions.
Mechanisms for fault-tolerant systems: replicated state machines, consistency models (atomic, eventual, causal), Raft consensus protocol, CAP theorem.
Principles of fault tolerance: hardware/software/operational faults, error detection/correction, replication models (active-active, active-passive), checkpointing.
Covers multicore scalability, Amdahl’s law, cache coherence overhead, vertical and horizontal scaling, load balancing, and application state management.