DevSum16: Roger Johansson - Akka.NET – Scale up and out!
Ta del av Roger Johanssons session Akka.NET – Scale up and out!
In the era of cloud computing, internet of things and high velocity data, we need more and better tools to deal with scalability. Akka.NET is an unified runtime and programming model for just that. This talk is an introduction to what Akka.NET can do for developers of today in terms of scalability and fault tolerance;
Multi core scale up – By embracing the actor model, systems built using Akka.NET by default become asynchronous and scale to maximize hardware utilization.
Scale out to multiple machines – As actors are ”location transparent”, Akka use the very same model for scale out as for scale up. turning your Akka.NET based solution from a single machine application into a distributed system with dynamic number of nodes is simply a matter of configuration.
Resilient and self-healing systems – The actor model have a unique way to deal with failure called”supervision trees”. Unlike the classical synchronous model where each consumer of a component need to deal with component failures, Akka.NET and the asynchronous model instead direct failures to dedicated fault handlers. Thus preventing application type errors from propagating to consumers of system components and instead resolve errors using those fault handlers.