Our monitoring solution fires a number of alerts notifying of the application and cluster’s health.
We have added extra alerts to warn when containers use more than 200% of it’s resource requests for 30 minutes.
This has already been rolled out everywhere.