Hi Prometheus Users, I am working on writing a (my first) custom exporter that collects metrics from an application's REST API endpoint that I have no control over. Now that I am half way through it, I am questioning myself whether I am taking the right approach to get the collected metrics published to my /metrics endpoint for Prometheus to scrape, and I could use some guidance. Let me attempt to explain what I'm doing.
My approach: Metrics are published on a /metrics endpoint that Prometheus will scrape every 15 minutes. But, how do metrics get to the /metrics endpoint? My idea was to have my exporter collect data from an external REST API on it's own schedule and publish the metrics to the /metrics endpoint so they would be ready and waiting for Prometheus to scrape them. This way, Prometheus would be scraping /metrics on it's own scrape_interval = 15m, and my exporter would be collecting data and publishing to /metrics on it's own export_interval = 15m. I might even lag ahead the export_interval by a minute or so, because each of the ~10 API calls it makes could take up ~20 seconds to complete. Is it necessary to have the 2 different schedules? Or would the Prometheus scrape_interval schedule be expected to trigger the API calls in the exporter? Basically I don't want to bog down that external applications performance or look like a DOS attack on the external application by continuously barraging it with requests when I only need the data once every 15 minutes or so. Hopefully I explained this alright, but let me know if you need any clarification. Thanks, Cory -- You received this message because you are subscribed to the Google Groups "Prometheus Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/prometheus-users/a119b024-37a0-4f5a-a239-990fc35c0cbc%40googlegroups.com.
