-
|
How can people happily use Prometheus to monitor their Raspberry Pi Kubernetes clusters and write about it on the web if it eats up all available RAM within minutes and brings a whole cluster node down. At least that's what happens to me whenever I spin up the prometheus stack, be it kube-prometheus-stack helm chart or this cluster-monitoring repo. As soon as the prometheus pod starts working it fills up the nodes 4 GB RAM within a few minutes. Is there anything I'm missing? Any special settings? How are you guys running this beast? |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment
-
|
No special settings, it should work but depends on the size of the cluster and amount of applications. |
Beta Was this translation helpful? Give feedback.
No special settings, it should work but depends on the size of the cluster and amount of applications.