HowTo: Common Patterns

Throttling Messages

Contributed by: Kaspar Fischer

"A message throttler that ensures that messages are not sent out at too high a rate."

The pattern is described in Throttling Messages in Akka 2.

Balancing Workload Across Nodes

Contributed by: Derek Wyatt

"Often times, people want the functionality of the BalancingDispatcher with the stipulation that the Actors doing the work have distinct Mailboxes on remote nodes. In this post we’ll explore the implementation of such a concept."

The pattern is described Balancing Workload across Nodes with Akka 2.

Work Pulling Pattern to throttle and distribute work, and prevent mailbox overflow

Contributed by: Michael Pollmeier

"This pattern ensures that your mailboxes don’t overflow if creating work is fast than actually doing it – which can lead to out of memory errors when the mailboxes eventually become too full. It also let’s you distribute work around your cluster, scale dynamically scale and is completely non-blocking. This pattern is a specialisation of the above 'Balancing Workload Pattern'."

The pattern is described Work Pulling Pattern to prevent mailbox overflow, throttle and distribute work.

Ordered Termination

Contributed by: Derek Wyatt

"When an Actor stops, its children stop in an undefined order. Child termination is asynchronous and thus non-deterministic.

If an Actor has children that have order dependencies, then you might need to ensure a particular shutdown order of those children so that their postStop() methods get called in the right order."

The pattern is described An Akka 2 Terminator.

Akka AMQP Proxies

Contributed by: Fabrice Drouin

"“AMQP proxies” is a simple way of integrating AMQP with Akka to distribute jobs across a network of computing nodes. You still write “local” code, have very little to configure, and end up with a distributed, elastic, fault-tolerant grid where computing nodes can be written in nearly every programming language."

The pattern is described Akka AMQP Proxies.

Shutdown Patterns in Akka 2

Contributed by: Derek Wyatt

“How do you tell Akka to shut down the ActorSystem when everything’s finished? It turns out that there’s no magical flag for this, no configuration setting, no special callback you can register for, and neither will the illustrious shutdown fairy grace your application with her glorious presence at that perfect moment. She’s just plain mean.

In this post, we’ll discuss why this is the case and provide you with a simple option for shutting down “at the right time”, as well as a not-so-simple-option for doing the exact same thing."

The pattern is described Shutdown Patterns in Akka 2.

Distributed (in-memory) graph processing with Akka

Contributed by: Adelbert Chang

"Graphs have always been an interesting structure to study in both mathematics and computer science (among other fields), and have become even more interesting in the context of online social networks such as Facebook and Twitter, whose underlying network structures are nicely represented by graphs."

The pattern is described Distributed In-Memory Graph Processing with Akka.

Case Study: An Auto-Updating Cache Using Actors

Contributed by: Eric Pederson

"We recently needed to build a caching system in front of a slow backend system with the following requirements:

The data in the backend system is constantly being updated so the caches need to be updated every N minutes. Requests to the backend system need to be throttled. The caching system we built used Akka actors and Scala’s support for functions as first class objects."

The pattern is described Case Study: An Auto-Updating Cache using Actors.

Discovering message flows in actor systems with the Spider Pattern

Contributed by: Raymond Roestenburg

"Building actor systems is fun but debugging them can be difficult, you mostly end up browsing through many log files on several machines to find out what’s going on. I’m sure you have browsed through logs and thought, “Hey, where did that message go?”, “Why did this message cause that effect” or “Why did this actor never get a message?”

This is where the Spider pattern comes in."

The pattern is described Discovering Message Flows in Actor System with the Spider Pattern.

Scheduling Periodic Messages

This pattern describes how to schedule periodic messages to yourself in two different ways.

The first way is to set up periodic message scheduling in the constructor of the actor, and cancel that scheduled sending in postStop or else we might have multiple registered message sends to the same actor.

class ScheduleInConstructor extends Actor {
  import context.dispatcher
  val tick =
    context.system.scheduler.schedule(500 millis, 1000 millis, self, "tick")

  override def postStop() = tick.cancel()

  def receive = {
    case "tick" =>
      // do something useful here
  }
}

The second variant sets up an initial one shot message send in the preStart method of the actor, and the then the actor when it receives this message sets up a new one shot message send. You also have to override postRestart so we don't call preStart and schedule the initial message send again.

class ScheduleInReceive extends Actor {
  import context._

  override def preStart() =
    system.scheduler.scheduleOnce(500 millis, self, "tick")

  // override postRestart so we don't call preStart and schedule a new message
  override def postRestart(reason: Throwable) = {}

  def receive = {
    case "tick" =>
      // send another periodic tick after the specified delay
      system.scheduler.scheduleOnce(1000 millis, self, "tick")
      // do something useful here
  }
}

Template Pattern

Contributed by: N. N.

This is an especially nice pattern, since it does even come with some empty example code:

class ScalaTemplate {
  println("Hello, Template!")
  // uninteresting stuff ...
}

Please keep this pattern at the end of this file.