
a thorough grasp of system models—timing, network communication and node failure mode——is very necessary if you're going to understand distributed systems papers.
i'd argue all three models should be committed to memory. it's the foundation every other idea in distributed systems build upon.
English
























