Hi,
In the process of evaluating grafana-agent (flow) as a replacement for node_exporter for metrics, promtail for most logs, and Loki docker driver (for Docker logs), i see that it seems the recommendation is to not serve metrics on the system, but instead perform remote writes to prometheus.
I got it working without much difficulty, but i can’t seem to figure out how to handle system up/down alerting.
Using the old way (prom + node_exporter) - When Prometheus can’t reach a target it alerts that the system is down, but as far as i know, there’s no such target configuration for remote write systems.
So my question is, how are people setting up alerts (in Alertmanager or Grafana) to alert when an expected system hasn’t sent any date in a while?
