Scaling out Azure function with Linux container deployment on Dedicated hosting plan is not working as expected

Question

I have Azure function with Linux container deployment (hosting Python function) on Dedicated hosting plan - Premium v3 P2V3, that comes with 16GB of RAM. I am using Service Bus trigger. When I use Manual scale out method and set instance count to 3, I would expect to get 3 instances of my function, each having it's own 16GB of RAM. Instead, it seems that I got 3 instances sharing the 16 GB of RAM and other hosting plan resources. It is like that number of "hosting" VM instances is not increased, only the number of Azure function instances, sharing the same hosting plan resources. Can someone confirm if this is actually how things work with Dedicated hosting plan?

Answer 1

Hello @Mladenovic, Dusan (EXT)

In a Dedicated hosting plan, when you manually scale out your function app, you are increasing the number of instances of the Functions host, not the number of VMs. Each instance of the Functions host shares the same resources, including the 16GB of RAM, with the other instances running on the same VM.

So, in your case, when you manually scaled out to 3 instances, you got 3 instances of the Functions host running on the same VM, each sharing the 16GB of RAM and other hosting plan resources.

If you need more resources for your function app, you can consider scaling up to a higher pricing tier with more resources or scaling out to more VMs.

I hope this clarifies your query.