Collecting metrics for only latest workflow runs

[arinzeakubue]

I find that I cannot setup alerting based on the status of the latest pipeline runs. Is it possible to setup tracking for run numbers so that only the metrics for runs since the last Prometheus scrape are exposed? Persisting a pointer or using a key-value store may be helpful in this regard.

[deebsman]

I've hit this too when trying to setup alerts. Did you find a workaround?