[chart] Don't set prometheus.io pod annotations when ServiceMonitor is enabled

[z0rc]

Describe the bug
https://github.com/kubernetes-sigs/aws-load-balancer-controller/blob/v2.4.1/helm/aws-load-balancer-controller/templates/deployment.yaml#L25-L26 are set always. But with ServiceMonitor enabled, these labels still present. This results in double scraping, as prometheus discovers both pod annotations and ServiceMonitor as independent scraping targets.

Steps to reproduce
Enable ServiceMonitor via helm values, observe, that prometheus.io annotations still set on pods.

Expected outcome
When ServiceMonitor is enabled, prometheus.io pod annotations should be omitted.

Environment

• AWS Load Balancer controller version: 2.4.1
• Kubernetes version: 1.22
• Using EKS (yes/no), if so version? Yes

Additional Context: None

[tamboliasir1]

Hi @z0rc.. Hi I would like to work on this. Just wanted to confirm so I wanted to check serviceMonitor is enabled or not from values if it is enabled then omit Prometheus annotations right?

[z0rc]

@tamboliasir1 you are correct, wrap prometheus.io annotations in if block and enable annotations when .Values.serviceMonitor.enabled is false.

[tamboliasir1]

Got it.. Thank you @z0rc

[tamboliasir1]

/assign

[tamboliasir1]

Hi @z0rc .. Just wanted to confirm condition will look something like this right?
{{- if not .Values.serviceMonitor.enabled }} prometheus.io/scrape: "true" prometheus.io/port: "{{ (split ":" .Values.metricsBindAddr)._1 | default 8080 }}" {{- end}}

[z0rc]

@tamboliasir1 aside from github formatting, this looks okay. But it's better to create a PR and do a code review there.

[tamboliasir1]

Hi @z0rc I have created a PR please review. Thanks