Auto Scaling

Overview

Auto-Scaling allows you define thresholds for starting and stopping extra instances. Scalarium monitors basic statistics like CPU usage, memory usage or load average on every instance. Apart from graphing this data, Scalarium can use those metrics to automatically add or remove instances.

Scalarium also makes it possible to start and stop instances in your cloud for only specific hours during the day. If you don't need some of your instances during the night, Scalarium will shut them down automatically.

Easy scaling of Rails and PHP applications in the cloud

Scaling Example

You could for example define a threshold of 40% - 80% CPU usage for Rails application servers. If the average CPU usage for all Rails application servers is above 80% for a number of measure intervals, Scalarium will boot additional Rails application servers.

Thanks to Auto Configuration the additional instances will be used as soon as they are online. Scalarium will continue to boot additional instances as long as your threshold value is exceeded. As soon as the load settles, no further instances are started.

Once the average CPU usage is below 40% for a number of measure intervals, Scalarium will shutdown one instance a time (down to a configured minimum).

Manual Intervention

You can of course start or stop instances at any time. If you know there is a traffic spike coming, start some instances in advance. This will ensure that there are enough instances once the traffic spike hits your application. When you don't need the additional instance anymore, just shut them down. This is exactly the total cost flexibility everybody is talking about!