It's a blog post by an Alice Shrugged team member about keeping their team-puzzle-state server up and running. We need jargon for this. HintOps? Solutions Reliability Engineering?