The alarm service allows you to monitor metrics and events to inform other services when certain conditions are met. The two most common uses for the alarm service are auto-scaling compute instances according to CPU load and auto-healing them when they fail a health check.
Auto scaling groups enable right sizing compute resources according to demand and improving application availability. Compute instances can be scaled out/in according to CPU utilisation, dynamically optimising the performance and cost of applications running on the Catalyst Cloud. You define the auto scaling policy and the cloud adjusts the number of compute instances based on the measured load.
Auto healing automatically repairs compute instances that fail to meet the health checks defined, increasing application availability.