There's a good Bryan Cantrill talk about that.[1] The gist is that eventually, when you throw enough resources at a problem, all that's left are the really unc

There's a good Bryan Cantrill talk about that.[1] The gist is that eventually, when you throw enough resources at a problem, all that's left are the really uncommon problem and bugs, and this is specifically what you get in the data path (including drive firmware) where things get harder and harder to figure out as the code gets more hidden and obscure. As with all his talks, you can expect it to be quite entertaining as well as informative and historical (if from his POV). 1: https://www.youtube.com/watch?v=fE2KDzZaxvE https://www.youtube.com/watch?v=fE2KDzZaxvE