Split Brain Scenarios in Akka Clusters

This is a simple application to create split brain scenarios in akka clusters, and how a split brain resolver handles such scenarios. The purpose of this application is to calculate the IQ of a person, given a name.

The following are specific requirements for this application:

• High availability - millions of people may be asking for their IQ at the same time. There should be no downtime, otherwise the app loses popularity.
• Consistent - once an IQ value is assigned to a name, it doesn't change

Application Design

The application is composed of 3 services:

• zookeeper - used for service discovery and cluster setup
• einstein - a service which calculates the IQ, given a person's name. Named after Albert Einstein, as he was very intelligent, hence capable to quickly assign an IQ value to a person's name
• nginx - receives all http requests and forwards them to the einstein service

To fulfill the first requirement, this service was designed to be part of a cluster where multiple einstein services may be running at the same time.

To fulfill the second requirement, the einstein service utilises the actor model to represent names. An actor is assigned a name and is responsible to calculate the IQ and return the same value whenever asked for it. This is used alongside cluster sharding, to make sure that only one actor for the same name exists within the cluster.

Building the application

Perform the following steps:

• mvn clean install to build the application code
• mvn docker:build to build the image and tag it with the latest tag

The akka-cluster-custom-downing_2.11 dependency is hosted by bintray (https://bintray.com/). For successful compilation, either add the dependency manually to your local repository by checking out the repository https://github.com/TanUkkii007/akka-cluster-custom-downing and building with sbt or configure maven to use bintray. To perform the latter, add the following to your settings.xml file:

   <profiles>
     <profile>
       <repositories>
         <repository>
           <snapshots>
              <enabled>false</enabled>
           </snapshots>
           <id>bintray</id>
           <name>bintray</name>
           <url>https://dl.bintray.com/tanukkii007/maven</url>
         </repository>
       </repositories>
       <pluginRepositories>
         <pluginRepository>
           <snapshots>
             <enabled>false</enabled>
           </snapshots>
           <id>bintray</id>
           <name>bintray-plugins</name>
           <url>https://dl.bintray.com/tanukkii007/maven</url>
          </pluginRepository>
        </pluginRepositories>
        <id>bintray</id>
     </profile>
   </profiles>

   <activeProfiles>
     <activeProfile>bintray</activeProfile>
   </activeProfiles>

Running the application

To run this application, make sure that you have docker and docker-compose installed.

Starting the application

Navigate to the docker directory and run the following command:

docker-compose -f application.yml up

To run multiple instances of the einstein service:

docker-compose -f application.yml up --scale einstein=3 einstein zookeeper nginx

Stopping the application

Navigate to the docker directory and run the following command:

docker-compose -f application.yml down

Asking for an IQ

To get an IQ value for a name, perform a GET request on the following URL:

http://localhost:8080/iq/{name}