A method for auto-scaling the infrastructure capacity of an application in response to client demands includes providing an application configured to run on an application infrastructure comprising a plurality of server nodes and to be accessed by clients via a first network. Next, providing traffic management means directing traffic from the clients to the server nodes of the application infrastructure. Providing monitoring means gathering performance metrics of the application and metrics of the application infrastructure. Providing controlling means configured to change the application infrastructure. Next, monitoring the performance metrics of the application and the metrics of the application infrastructure via the monitoring means thereby obtaining metrics information and then changing the application infrastructure based on the metrics information via the controlling means. Finally directing network traffic targeted to access the application to server nodes of the changed application infrastructure via the traffic management means.