there is a corollary to "developers should own their code", and it is this: *On call must not be hell*.
ops has always had a problem with masochism, and we have to cut that shit out. don't be okay with abusing yourself, no matter WHAT kind of engineer you are.
ops has always had a problem with masochism, and we have to cut that shit out. don't be okay with abusing yourself, no matter WHAT kind of engineer you are.
i mean i'm over 30, i want to sleep thru the night (if i managed to get to sleep that night ☺️) doesn't matter what my title is. so i get you. took me a while to start valuing my own sleep that much tho.
management is super key here. if you're a manager of engineers, you need to give them cover to pay down the technical debt, clean up alerts, fix things that are nagging problems and flappy stressballs.
another big headwind is just the fact that so many people have never seen a better world. they don't know that on call doesn't have to be hell, doesn't have to be life-impacting. you can't be what you don't see, to coin a phrase.
feel like the people who are doing this well are hesitant to talk about it because they don't want to sound like they're bragging (or jinx themselves). this definitely holds me back from talking about e.g. why on call isn't horrible for my teams.
but like: we have a fast-growing platform with unpredictable traffic **and a homegrown storage engine**, yet we have a customer-impacting outage approx once every 6 months. and we work hard to protect our engineers' time and sleep if customers aren't hurting.
managers who push their engineers to just ship features and don't give enough time to clean house ... they are key villains in this narrative.
this is why managers should always be on call, too. (in some creative fashion. i like pinch hitting and substitute-oncalling)
this is why managers should always be on call, too. (in some creative fashion. i like pinch hitting and substitute-oncalling)
