I have a Blazor application which is type of WebAssembly hosted model, this is an extra option when creating a Blazor WebAssembly project though Microsoft Visual Studio.
This has an effect that there are three projects one client, one server and another one shared. The client is downloaded into the machine for rendering while server stays inside infrastructure to be able to execute server side services.
If this Blazor app is running on a pod inside a Kubernetes cluster and we identify that we need to scale out, scaling out this pod by increasing the replicas will be enough or do we need to consider also other things?
I imagine that scaling out like this, the more servers, the more clients would be able to be downloaded, "so client and server will be duplicated equally". Of course if any state is saved as static variable inside the app this is something we need to consider, but is this actually the best way to scale out?
Can anyone explain better?