Alert if scraping jobs fail
In addition to alerting on failing scraping jobs, we need to address our existing failures, which I researched in https://github.com/NixOS/infra/issues/551. Copy pasting analysis from there: - `r13y`: Last successful scrape: 2024-09-21. Website is down. @grahamc didn't respond to my query. - `rfc39`: IMO, this is mis-architected. This scrapes a job that only runs periodically, which means we regularly get scrape failures. I've filed https://github.com/NixOS/rfc39/issues/14 with the upstream project seeking advice. - Since we don't actually alert on these metrics, I propose doing the simplest thing and just disabling this scrape job for now. - `hydra_notify`: Last successful scrape: 2024-08-02. We disabled this service [in 2024](https://github.com/NixOS/infra/commit/66da5cfddb8c67b5e2f0b5bba62899f1f82eec42), there's no reason to keep scraping it.