HA scrape target

[sayf eddine Hammemi]

Hello

I have two scrape configs that provide the same metrics +- timestamps ofc for HA reasons. is there any simple way to configure Prometheus to ignore one and take another ? so if one of them is down we switch to the second. I am trying to avoid using HA DNS apparatus

Thanks

[Jake Utley]

Are you saying that you have one Prometheus instance that has redundant scrape configurations? That's an unusual pattern. Usually you want to have two Prometheus instances with identical configurations for high availability. Generally you can scrape either one and you'll get a reasonable result, or you can use Thanos or Cortex to deduplicate the HA data.