Autoscaling your assembly

After your deployment steps are generated and your assembly is deployed, you have to be prepared to handle all kinds of load via your assembly. Fortunately, you chose to deploy via a load balancer, which does most of the work for you. You can configure the scaling up and down of your assembly in advance, so as to handle any anticipated load. It is also important that you set some smart limits on scaling up, so as not to incur any unnecessary costs in case your app decides to go rogue on you and blows the CPU limit due to a bug, rather than genuine load. You can configure your autoscaling options by either going to transition, under the assembly or going to the LB component, under transition.

Here, you can control the scaling of backend compute instances:

By default, it deploys 100%, in this case two instances and maintains a minimum of two instances. However, it does not exceed a maximum of 10 instances. As the load increases, OneOps starts spinning new compute...