We're struggling a bit with custom job metrics being pushed to stackdriver, here's what we do:
Install prometheus from helm
Install a sidecar by patching deployment
Installing a custom exporter for uwsgi and celery (in cluster services)
Modifying prometheus' configmap to add a new scraping job that is utilizing exporters above
Sending our exporter metrics to stackdriver (this doesn't work yet)
So, we're seeing all other kubernetes based prometheus metrics being sent to stackdriver, but not thje ones that come from our exporters. We've also confirmed they're present in prometheus and are being scraped correctly. Is there some secret sauce when you define a new scraping job in order for it to be sent correctly to stackdriver ?
Hi All,
We're struggling a bit with custom job metrics being pushed to stackdriver, here's what we do:
So, we're seeing all other kubernetes based prometheus metrics being sent to stackdriver, but not thje ones that come from our exporters. We've also confirmed they're present in prometheus and are being scraped correctly. Is there some secret sauce when you define a new scraping job in order for it to be sent correctly to stackdriver ?