You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
we are currently facing issues with some alerts with changing label.values (e.g. high memory).
I did some research in the web and learnt for Prometheus every alert is identified by its combination of metric_name and labels. That means if the 'label.value' changes this is handled as a new alert. This on the one hand, could cause issues while alerting, because if the value changes in the 'for' time range it might starts over again as it is handled as a new alert and on the other hand alerts get resolved and fired over and over again while the 'label.value' changes.
Therefore I would like to remove the 'label.value' for some of the 'threshold' alerts.
Hi,
we are currently facing issues with some alerts with changing label.values (e.g. high memory).
I did some research in the web and learnt for Prometheus every alert is identified by its combination of metric_name and labels. That means if the 'label.value' changes this is handled as a new alert. This on the one hand, could cause issues while alerting, because if the value changes in the 'for' time range it might starts over again as it is handled as a new alert and on the other hand alerts get resolved and fired over and over again while the 'label.value' changes.
Therefore I would like to remove the 'label.value' for some of the 'threshold' alerts.
Any thoughts on this @frodenas
The text was updated successfully, but these errors were encountered: