Tinkering with Akka Cluster

This is a toy application developed to play a bit with Akka Cluster. It is based on the sample repo provided by Akka, and on the official documentation. Do you want to play with it? Feel free to contribute! 😄

The project is "docker-first", so we need to create the docker images to run using sbt Native Packager. Run the command sbt docker:publishLocal, it will create the image elleflorio/akka-cluster-playground.

Once you have the image, type docker-compose up. Docker compose will run a seed node along with node1 and node2.

I integrated Akka HTTP, so you can interact with the cluster using the following REST API. The port for the nodes are the following:

• seed: 8000
• node1: 8001
• node2: 8002

Check if the nodes are up and running calling a /health endpoint:

GET http://localhost:{{port}}/health

Response:

OK

Check the members inside the cluster using the /status/members endpoint:

GET http://localhost:{{port}}/status/members

Response:

[
    "akka.tcp://cluster-playground@node1:1600",
    "akka.tcp://cluster-playground@node2:1600",
    "akka.tcp://cluster-playground@seed:2552"
]

Ask for the computation of Fibonacci number n with the /process/fibonacci/n endpoint:

GET http://localhost:{{port}}/process/fibonacci/10

Response:

{
    "nodeId": "seed",
    "result": 55
}

The nodeId represents the node where the processor that executed the computation is running.

The cluster uses Cluster Aware Routers for load balancing, with a round-robin strategy. You can verify this asking for the Fibonacci number computation to a random node. Everytime the nodeId in the response will be different, following a round-robin strategy.

That's it! Enjoy! 😁