How to add a static_config to the promethes deployed by helm?

42 views

Skip to first unread message

kazakaze

unread,

Sep 17, 2021, 6:42:08 AM9/17/21

to Prometheus Users

My prometheus is deployed via helm-charts

# kubectl get po -n monitor

NAME READY STATUS RESTARTS AGE

alertmanager-prometheus-stack-kube-prom-alertmanager-0 2/2 Running 8 (19d ago) 74d

prometheus-prometheus-stack-kube-prom-prometheus-0 2/2 Running 9 (19d ago) 74d

prometheus-stack-grafana-56b797cc96-sgspf 2/2 Running 8 (19d ago) 74d

prometheus-stack-kube-prom-operator-cfbf6c47f-mjr6m 1/1 Running 4 (19d ago) 74d

prometheus-stack-kube-state-metrics-67889f75bd-tl6k7 1/1 Running 4 (19d ago) 74d

prometheus-stack-prometheus-node-exporter-r77qj 1/1 Running 4 (19d ago) 74d

I started the example through go run, and I know that scrape_configs needs to be modified. My question is:

1. It seems that I need to encapsulate the program into a mirror image and add a new crawl through PodMonitor/ServiceMonitor? Must it be packaged as a mirror image?

2. How to automatically modify scrape_configs, if I want to manually, which pod should I exec to?

Harald Koch

unread,

Sep 17, 2021, 4:55:35 PM9/17/21

to Prometheus Users

On Fri, Sep 17, 2021, at 06:42, kazakaze wrote:

My prometheus is deployed via helm-charts

[...]

2. How to automatically modify scrape_configs

Since I just pieced this together myself last week, I'll share. :) If you're using kube-prometheus-stack (which you should be):

helm show values prometheus-community/kube-prometheus-stack

Will give you a list of all configuration values for the helm chart. The one you're looking for is prometheus.prometheusSpec.additionalScrapeConfigs.