I have, but it doesn't seem to matter. Skip level manager seems to believe that this sort of work is more important than the actual product, and my direct manager sees it as important because it impresses his boss. I haven't actually asked to transfer, but I doubt either of them would take it well since the rest of the team has had the good sense not to show any aptitude for this sort of work.

this is what i wish i had done in the beginning

We don't really have those requirements though, and we already have an issue with everyone ignoring alerts when they're on call because we alert on too many trivial things. 99% of the time it just means there was an error somewhere in a service my team doesn't control. Adding more noise just for the sake of adding noise doesn't help.

I mean, you can go look at the latency, but it's generally not a good criteria for automatic alerts that wake people up at 3 am. If the latency is too long, you'd get a timeout and then the error alerts would kick in. We've not really had any issues regarding latency spikes, so it's not like it's even a reaction to something.

But yeah, maybe they can hire a SRE and have them do this mind numbing shit.

That was my plan, a month ago. I assumed that she would eventually run out of shit to monitor, but the requests just keep getting more and more ridiculous.

For instance, right now I'm tasked with finding a reasonable latency threshold for each of our endpoints and creating an alert when a request exceeds that threshold. We have 30 services, each with approximately 10 endpoints. So we're talking 300 individual alerts to maintain, and whoever is on call getting bombarded if even one or two thresholds are exceeded on a regular basis. I've explained why this is a problem, but it falls on deaf ears.