Last week I mentioned the HBR article, “What Every CEO Need to Know About the Cloud”. Today I saw Brandon Williams posted about a piece of the article I did not focus on and should have: the Chaos Monkey.
Brandon mentioned that the Chaos Monkey is probably revenue impacting. Yup! That makes people scared! But if you have a real event and can’t handle it properly, that will likely turn into an even more revenue impacting incident than it could have been. And when it comes to Operations: things *will* fail someday.
I agree there should be controlled incidents in operations and especially in security. But I would caution that there is a very fine line between doing these controlled incidents (or tests) in a positive way or in a way that people hate and will react badly to.
Learning from tests is paramount. Pointing fingers and slapping hands and smacking ears because something is missed is called Doing It Wrong. You’ll just result in scared, sensitive, and not-happy people. No one likes to be tested and to fail and then have it shoved into their face or performance review. Yes, that gets results, but the point of so much testing and education is improvement and positive conditioning, not negative conditioning and fear.