CODEWITHSUNDEEP
azureazure-web-app-service
Randy Minder
Randy Minder

Reputation: 48522

Web App Autoscaling - What actually happens when this occurs?

In Azure, you can auto scale your web app. You can create a rule that allows you to determine when to auto scale. When the auto scaling occurs one App Service instance is added (or removed). I'm curious what exactly this means. For example, say I currently have a web app using the S2 service tier. When the scaling occurs, is another S2 instance added? I suspect this is the case. And, if this is the case, what handles the load balancing between the two (or N) instances? Is this handled automatically?

Upvotes: 2

Views: 609

Answers (1)

Rob Reagan
Rob Reagan

Reputation: 7696

You are correct - when you auto-scale, you are adding additional instances of the same type. The scaling is seamless; there is no downtime during the scale out. You are also correct about load balancing. Load balancing is done automatically. Azure will dole out requests between your two (or more) instances. I am 99.99% sure that this is round-robin with the exception of if you have ARRA affinity enabled. Please, someone correct me on round-robin if I am wrong. I can tell you that when watching my Web App instances after scale out, the load is pretty darn even.

Upvotes: 3

Related Questions