Data is being collected using node exporter in EKS Cluster(1).
I am planning to use Prometheus on another EKS Cluster(2) for monitoring. Currently, the node exporter and Prometheus are connected in EKS Cluster (1), and data is transferred from Prometheus to Prometheus in EKS Cluster (2) through remote_write.
However, my leader says he does not like this method because running Prometheus on EKS Cluster(1) will use a lot of resources. So, let’s find a configuration that pulls the Node exporter data of EKS Cluster(1) directly from EKS Cluster(2) Prometheus.
I am curious about whether this is the method my leader pursues, and I am also curious about whether the leader’s method is best practice.
The best structure in my opinion is the Prometheus federation structure.