When venturing down the path of investigating and hopefully resolving a production issue, it's important to keep track of what you do so that you can undo it, discard it, or repeat it. It's kind of the promise of infrastructure as code but with a more risky but tighter feedback loop