Simon Pasquier

Are you talking about sharding rules to multiple instances? My initial idea was only to the number of replicas: while it doesn't work for HPA, it would help supporting VPA.

well this is an interesting question: maybe we should pause on this one and think how ThanosRuler could support both HPA and VPA. I suppose that we could add a...

@gameloser distributing rules across multiple instances wouldn't prevent the VPA use case. But being able to scale the ruler horizontally would be a nice feature. @afzal442 nothing concrete yet unfortunately....

> Keep the ready probes' timeouts for Prometheus Agent's DaemonSet mode same as the StatefulSet mode's. DaemonSet would still run, but probably take more time to be ready than needed....

there's no requirement on Prometheus side that alert names are unique within a group. In practice, many alerting rules use the same name with different severity labels.

The issue is either in the way the operator configures the statefulset or with your kubernetes DNS setup.

@A1kmm you can add your project here: https://prometheus.io/docs/operating/integrations/#alertmanager-webhook-receiver

Something's weird with CircleCI reporting `Could not find a usable config.yml, you may have revoked the CircleCI OAuth app.`... Have you enabled CircleCI on your fork repository?

> Often, we use severity labels to denote the criticality of alerts. Changing these labels means that we lose continuity between the fired alerts vs the newly defined alert >...

Our recommendation is to use your own service which offers the greatest flexibility. The `prometheus-operated` service is there for historical reasons. And with the future release of Prometheus operator, it...