Summary
In this chapter, we learned how to scale the Kubernetes resources of our Dapr applications with the Deployment configuration, as well as how to automatically adapt the number of replicas to CPU and memory usage with the HPA.
The concepts we explored in this chapter gave us a more solid approach to testing Dapr applications under specific conditions; is our overall solution, starting with the nodes of the Kubernetes cluster and including the database (state store) and message bus (publish/subscribe), capable of sustaining a specific load?
Even if we had to venture outside the land of C# and .NET to leverage Locust, I think the advantages of learning a popular, developer-oriented load-testing framework justify the effort. Python is also supported in Dapr with an SDK for services and actors, so maybe this could be the next stage of our learning experience with Dapr.
