A team wants to verify that CloudWatch alarms correctly trigger before deploying new auto-scaling policies to production. Which approach achieves this non-destructively?
-
A
Manually scale the fleet down to zero and observe alarms
-
B
Use the CloudWatch set-alarm-state API to force alarm states and verify downstream actions
-
C
Delete the existing alarms and recreate them with new thresholds
-
D
Disable the auto-scaling group and test alarm math manually