I never found Prometheus that difficult. I would say I've spent a moderate amount of time over the past few years reading and using it (almost exclusively just

I never found Prometheus that difficult. I would say I've spent a moderate amount of time over the past few years reading and using it (almost exclusively just reading the docs and blog). I haven't used Datadog. The places I've used Prometheus tended not to use Saas, which meant trying to get the wallet opened or approval for sending data offsite were basically non-starters. I know that's not an issue for most here. I set it up as a side project because we were basically blind and reactive to issues. The other options I was looking at was Nagios and Cacti. Machine metrics are basically plug-and-play. The only real downside with Prometheus is that you have to install the pieces yourself (and know enough to do it). After you understand how the pieces fit together and get basic knowledge of each piece the time suck is writing and grokking new metrics and noodling with a dashboard--which I figure any solution would have. This is probably home-hobby geekery, but I'm running Prometheus at home. My Synology has Prometheus and Grafana running in docker containers. The only gnarly part was getting dd-wrt's SNMP data and creating a dashboard of router metrics.