Trying to set this up with newly built 3.11 cluster. Had some trouble getting it going initially but added the following to charts/vertical-pod-autoscaler/templates/recommender-deployment:
Allowed me to override the prometheus address. I now have an issue where the recommender pod does not like speaking to ssl enabled prometheus endpoint, showing the following error:
Trying to set this up with newly built 3.11 cluster. Had some trouble getting it going initially but added the following to charts/vertical-pod-autoscaler/templates/recommender-deployment:
command: ["/recommender"] args:
Allowed me to override the prometheus address. I now have an issue where the recommender pod does not like speaking to ssl enabled prometheus endpoint, showing the following error:
1 cluster_feeder.go:187] Cannot get cluster history: cannot get usage history: cannot get timeseries for cpu: Retrying GetTimeseries unsuccessful: tried 10 times, last error: error getting data from Prometheus: Get https://prometheus-k8s.openshift-monitoring.svc.cluster.local:9091/api/v1/query?query=container_cpu_usage_seconds_total%7Bjob%3D%22kubernetes-cadvisor%22%2C+pod_name%3D~%22.%2B%22%7D%5B8d%5D: x509: certificate signed by unknown authority
If I rsh to the recommender pod there is no /etc/ssl/certs directory so I'm not sure how any ssl connection would be trusted.
Do I need to make prometheus available via plain http?
Cheers