Akka (toolkit)

History

Distinguishing features

• Asynchronous and non-blocking communication, distribution, and concurrency: Akka applications are event-based, asynchronous, and non-blocking: no mutable data are shared, and no synchronization primitives are used; Akka implements the actor model with support for streaming, Publish-Subscribe, HTTP, gRPC, and multiple other protocols.
• Location transparency: The way Akka-based services/agents interact is the same whether they are on the same host or separate hosts, communicating directly or through routing facilities. This means that the topology of the application is not fixed but dynamic and may be altered at deployment time through a configuration mechanism, allowing a program to be scaled up, out, and clustered, without code modifications.
• Self-healing through declarative failure management: Actors are arranged hierarchically in so-called ‘supervisor hierarchies’. Failures are treated as immutable facts, reified events, and sent asynchronously to the component’s supervisor, who can manage the failure in a safe and healthy context outside of the failed component. In contrast to Erlang, Akka enforces parental supervision, which means that each actor is created and supervised by its parent actor.
• Durable replicated in-memory persistence: Akka services/agents are durable with their in-memory state acting as the source of truth and each state-changing event is logged to disk in the order they arrive leveraging so-called event-sourcing. Replaying the event log allows them to gracefully recover from failure, sourcing replicas, and provides a built-in audit log of everything that has happened in the system.
• Multi-region/multi-cloud clustering: Akka services/agents are clustered automatically “from within”. Each service is its own fully replicated and sharded cluster of nodes that can span multiple data centers, regions, clouds, or span from the cloud to the edge.

• Akka Libraries is an open-ended toolkit for building distributed systems. It has a modular structure, with a core module providing actors. Other modules are available to add features such as network distribution of actors, cluster support, Command and Event Sourcing, data distribution and management, integration with various third-party systems through the streaming integration module, and even support for other concurrency models such as asynchronous stream-processing and Futures.
• Akka SDK is a high-level and opinionated framework built on top of the Akka Libraries. It encodes the most common and useful patterns and best practices learned from the use of Akka Libraries through a set of discrete composable components, allowing developers to build highly responsive, scalable, resilient, distributed agentic and services-based applications.

Relation to other libraries

• The Play framework for developing web applications offers integration with Akka
• Up until version 1.6, Apache Spark used Akka for communication between nodes
• The Socko Web Server library supports the implementation of REST APIs for Akka applications
• The eventsourced library provides event-driven architecture support for Akka actors
• The Gatling stress test tool for load-testing web servers is built upon Akka
• The Scalatra web framework offers integration with Akka.
• The Vaadin web app development framework can integrate with Akka
• The Apache Flink RPC system is built using Akka but isolated since v1.14.
• The Lagom framework for building reactive microservices is implemented on top of Akka.

Publications about Akka

• Akka Essentials
• Akka Code Examples
• Akka Concurrency
• Akka in Action, Second Edition
• Akka in Action
• Effective Akka
• Composable Futures with Akka 2.0, Featuring Java, Scala and Akka Code Examples

• P. Haller's "Actors in Scala"
• N. Raychaudhuri's "Scala in Action"
• D. Wampler's "Functional Programming for Java Developers"
• A. Alexander's "Scala Cookbook"
• V. Subramaniam's "Programming Concurrency on the JVM"
• M. Bernhardt's "Reactive Web Applications"