Reputation: 11
I'm using the kube-prometheus-stack which uses the prometheus-operator and I am seeing a lot of logging messages because prometheus is trying to access the /metrics endpoint of my application, which is not allowed (because not existing). When I open Prometheus-WebUI/Targets is see there a "k8s"-target which lists all containers with "/metrics"-endpoints. Most of them are down, including the one which produces the annoying logs.
How can I disable this pool completely or for some containers/pods?
I couldn't find any configuration of this pool (no serviceMonitor, no configMap, ...), also the documentation is not telling anything about a k8s-scrape-pool (or at least I couldn't find).
Upvotes: 0
Views: 525
Reputation: 11
I found it: It was created by my own additionalScrapeConfig
additionalScrapeConfigs:
- job_name: k8s
kubernetes_sd_configs:
- role: pod
relabel_configs:
- action: labelmap
regex: "__meta_kubernetes_pod_label_(.+)"
replacement: "label_$1"
I didn't know, that this would produce a new scrape pool. I only additionally wanted the pod-labels.
Upvotes: 1