Understand Akka Clustering With LEDs and a 5-Node Mini Cluster

Akka Cluster brings the power, efficiency, resilience, and scalability of Akka actors to massively distributed systems. Developers and architects that understand how to use Akka Cluster effectively in the cloud, on-premise, and even across multiple data centers have found that it is well worth the effort to learn it — just ask anyone that "gets it."

One challenge for Akka champions who want to demo the value to others, however, is that all this magic is happening behind the scenes. It's not easy to immediately visualize how Akka establishes clusters, self-heals nodes, handling network partitions automatically, and so on. Well, now you can.

In this webinar by Lightbend's Eric Loots, Scala and Tooling Practice Lead, and Kikia Carter, Principal Enterprise Architect, we use a simple yet powerful visualization of a 5-node, Raspberry Pi-based cluster to reveal the inner workings of Akka Cluster. In a matter of minutes, you will gain a strong understanding of clustering, even if you don't know anything about Akka. Specifically, Kiki and Eric walk you through visualizations that review:

• Important core Akka Clustering concepts, like cluster node state (Down, Up, Weakly-up, Joining), leader election and leadership role, as well as heartbeating, gossip protocol, and failure detection.
• What happens when we add the chaos of a network partition into the mix-leading to a so-called "Split Brain" scenario-and how can Akka HTTP management help.
• How Akka Split Brain Resolver can automatically recover from a network partition and thus avoid a service being unavailable for a long time.

Watch the Full Presentation and Demo

Other Resources You May Enjoy

• Demo Source Code: monitor this GitHub repo for updates to the source code
• White Paper: "Akka A to ZThe Architect's Guide To Designing, Building, And Running Reactive Systems"
• O'Reilly eBook: Designing Reactive Systems: The Role Of Actors In Distributed Architecture