Significantly, AWS – like Meta in the more recent outage – blamed human error for what turned out to be a 13-hour systems ...