Observability · Technology in use
Prometheus
A time-series database that collects metrics by scraping and queries them with PromQL.
The problem it solves
A time-series database that collects metrics by scraping and queries them with PromQL.
Technical rationale
Prometheus stores the whole series rather than the last value, which is what makes it possible to reconstruct what was happening in the minutes before a failure. It collects by scraping, so a target that stops answering is itself a signal; I run it on a separate guest so it does not share fate with what it measures.
How I have used it
It collects homelab metrics from its own guest, outside the cluster it watches.