Hey everyone! I'm diving into a hobby project to learn more about logging and observability—something many of us developers often overlook. I'm interested in what you think are the most important metrics to keep an eye on when it comes to monitoring node or server health and resource utilization. I'm setting up my Prometheus stack and want to avoid cluttering my dashboards with unnecessary data that isn't really useful in practice. Any insights would be appreciated!
1 Answer
Honestly, I don't really bother with dashboards too much. I rely on alarms instead. The metrics I monitor mostly come from the kube-prometheus-stack, and so far, I haven't run into any issues that weren't already flagged by my alarms. It’s a pretty effective way to manage things without getting overwhelmed.
That's interesting! But what are your alarms based on? Are they set for things like disk pressure, CPU usage, or memory utilization? I haven't fully explored the kube-prometheus stuff yet, so just trying to get a handle on it!