I am using the Memory Request Deviation metric on Kubernetes Pods and would like to set a caution when the memory usage goes over requested by 10% and danger when it goes over by 100%. I am using the "Less Than" comparison Operator. I can set the danger to -100 and when I set the caution to -10, getting the message "The Caution threshold must be higher or equal to the Danger threshold". Playing with the numbers, the error goes away if I set the caution to a number larger than the danger (-110 for example). However, -110 is lower, not higher, than -100. It appears that the check that is being done on the danger and caution are using absolute numbers, not the actual numbers. Can we please get a fix so both the danger and caution can be negative numbers with the caution being the larger number when the "Less Than" operator is used.
Release : SAAS
Thank you very much for the hotfix. I can confirm that the change has resolved the issue on all three tenants. We can now set both the caution and danger thresholds values in alerts to negative numbers as expected.