Helion Stackato can automatically scale the number application instances within a specified range based on the average CPU utilization across all instances.
The auto-scaling behavior can be configured in the Management Console, in the manifest.yml config file, or using the stackato client by setting CPU thresholds for scaling events, and a minimum and maximum number of application instances.
The minimum and maximum CPU Threshold values specify the upper and lower limits where scaling operations will occur.
CPU utilization is measured as a moving average across all currently running instances of the application. Because new application instances may take some time to start, they are added or removed at a maximum rate of one per minute to allow for more accurate measurement of the new average as load is redistributed.
With the CPU threshold values unset, no application auto-scaling will occur. The Management Console will display minimum and maximum values of 0%.
The Instances values set the minimum and maximum number of instances in the application pool.
When scaling values are set, they override the
instances setting for
the application, scaling the application up or down as required into the