This bosh release is deprecated by RFC 0022. This will be removed from CF Deployment after Nov 23, 2023 and will be archived after Jan 23, 2024.
The Metrics Discovery Release is intended to give release authors an easy way to discover Prometheus Exposition formatted metrics endpoints in their Cloudfoundry Deployments. It consists of two components: The Metrics Discovery Registrar and the Scrape Config Generator.
If you have any questions, or want to get attention for a PR or issue please reach out on the #logging-and-metrics channel in the cloudfoundry slack
The Metrics Discovery Registrar publishes scrape configs to CF NATS to be consumed by a Scrape Config Generator.
Interested jobs should provide a metric_targets.yml file and place it in the location described by the targets_glob
property -- by default var/vcap/data/*/metric_targets.yml.
The metric_targets.yml should provide information
about the endpoint to be scraped using the Prometheus format.
The Scrape Config Generator subscribes to CF NATS and consumes published metric targets. It aggregates the metric targets
and saves a Prometheus-formatted scrape config
to the path defined by the scrape_config_file_path property -- by default /var/vcap/data/scrape-config-generator/scrape_targets.json.
The scrape config will be modified as metric targets come and go. Interested metric scrapers should watch the scrape config file for changes.
An agent that proxies to components with a prom_scraper_config.yml and
receives metrics from the Forwarder Agent and exposes them on a prometheus-scrapable endpoint.
More information can be found in the docs