Describe the bug
The prometheus.io/port annotation for the proxy statefulset uses the port value for the service instead of the value for the container. If those two things are not the same (which they aren't by default), prometheus scrapes will fail.
To Reproduce
Steps to reproduce the behavior:
deploy pulsar proxy using the helm chart and default values in a kubernetes cluster where prometheus is configured to scrape pods based on annotations
Notice prometheus is trying to scrape port 80, when the pods are actually listening on port 8080
Expected behavior
Prometheus scrape annotations should contain the port the pods are listening on, allowing them to be scraped.
This is a simple fix. I'll be submitting a PR to correct it.
Describe the bug The prometheus.io/port annotation for the proxy statefulset uses the port value for the service instead of the value for the container. If those two things are not the same (which they aren't by default), prometheus scrapes will fail.
To Reproduce Steps to reproduce the behavior:
Expected behavior Prometheus scrape annotations should contain the port the pods are listening on, allowing them to be scraped.
This is a simple fix. I'll be submitting a PR to correct it.