- Time Series+1 more
- 16. Sep
Prometheus-to-sd is a simple component that can scrape metrics stored in prometheus text format from one or multiple components and push them to the Stackdriver. Main requirement: k8s cluster should run on GCE or GKE.
For scraping metrics from the component it's name, host, port and metrics should passed through the flag source in the next format: component-name:http://host:port?whitelisted=a,b,c. If whitelisted part is omitted, then all metrics that are scraped from the component will be pushed to the Stackdriver.
If metric descriptors already exist on the Stackdriver (created manually or by different component) then autodiscovery feature could be used. In such case prometheus-to-sd will push metrics for which metric descriptors are available on the Stackdriver. To use this feature a flag auto-whitelist-metrics=true has to be passed.
Scrape interval vs. export interval
There are two flags: scrape-interval and export-interval that allow specifying how often metrics are read from the sources and how often they are exported to Stackdriver, respectively. By default both are set to 1m. The scrapes can be more frequent than exports, however, to achieve grater precision for metrics being exported. For example, if metrics are exported once every minute and a container dies between scrapes, up to 1 minutes of metrics can be lost. Frequent scrapes mitigate that, at the cost of elevated resource usage.