This issue has impact on availability monitoring. There's an observable 50s lag between updates of the up metric when the scrape target is down, and interval between samples increases from 10s to 30s.
This issue was introduced in VictoriaMetrics v1.66.0 and affects PMM v2.27.0 or later.
How to test
Run the feature build and make sure that everything works as usual.
How to document
Before VictoriaMetrics v1.93.2, vmagent didn't honor the scrape timeout in stream parsing mode, causing unexpectedly long delays when scrape target is down.
Activity
Show:
Roma Novikov September 11, 2023 at 10:26 AM
We prepared for 2.40, but we can try to get this version in the upcoming release.
Before VictoriaMetrics v1.93.2, vmagent didn't honor the scrape timeout in stream parsing mode, causing unexpectedly long delays when scrape target is down, see https://github.com/VictoriaMetrics/VictoriaMetrics/issues/4847.
This issue has impact on availability monitoring. There's an observable 50s lag between updates of the up metric when the scrape target is down, and interval between samples increases from 10s to 30s.
This issue was introduced in VictoriaMetrics v1.66.0 and affects PMM v2.27.0 or later.