When alertdiff restarts, the timeseries it generates can get different pod name/instance/node labels resulting in duplicated details in the alerts. We should exclude these labels from the alert definitions:
IF max(alertdiff_spurious_alert) without (kubernetes_pod_name,instance,node) > 0
When alertdiff restarts, the timeseries it generates can get different pod name/instance/node labels resulting in duplicated details in the alerts. We should exclude these labels from the alert definitions:
IF max(alertdiff_spurious_alert) without (kubernetes_pod_name,instance,node) > 0