Tbh like 99% of web apps aren’t critical - most of them are for buying something or providing infrastructure to make it easier to buy something anyway. It’s fine if your online shop is down for a few minutes (of course the business won’t see it like that but it’s true)

Steam, +130 million users, biggest gamming store, multi billion dolar company, is offline every week for a few minutes.

99.999999 is so overated.

It could probably be done with a simple if statement, but I'm guessing that dang enjoys the tradition of doing it manually every year.

Everyone went right for “downtime” - no, that’s not an issue. I would have expected a configuration change, which wouldn’t require a restart but might indeed result in downtime.

Or even a configuration change that some control system notices and does restart the service.

It’s the manual, hands-on connotation (maybe only in my mind) of telling someone a restart is involved. Automate this stuff - don’t want the code in the server all year? Fine, have a process rebuild and relaunch on a schedule that makes sense. Might have downtime, but definitely have less hands-on.