Actor

Overview

The fundamental building blocks of GoAkt are actors.

• Have a unique name across the entire system even in cluster mode. One a name is given to an actor, it is immutable and it is globally unique.

• They are independent, isolated unit of computation with their own state.

• They can be long-lived actors or be passivated after some period of time that is configured during their creation. Use this feature with care when dealing with persistent actors (actors that require their state to be persisted).

• They are automatically thread-safe without having to use locks or any other shared-memory synchronization mechanisms.

• They can be stateful and stateless depending upon the system to build.

• Every actor in GoAkt:

  • has a process id PID. Via the process id any allowable action can be executed by the actor.

  • has a lifecycle via the following methods: PreStart, PostStop.

    • PreStart hook is used to initialise actor state. It is like the actor constructor.

    • PostStop hook is used to clean up resources used by the Actor. It is like the actor destructor. It means it can live and die like any other process.

  • handles and responds to messages via the method Receive. While handling messages it can:

  • create other (child) actors via their process id PID SpawnChild method

  • send messages to other actors locally or remotely via their process id PID Ask, RemoteAsk(request/response fashion) and Tell, RemoteTell(fire-and-forget fashion) methods

  • stop (child) actors via their process id PID

  • watch/unwatch (child) actors via their process id PID Watch and UnWatch methods

  • supervise the failure behavior of (child) actors.

  • remotely lookup for an actor on another node via their process id PID RemoteLookup. This allows it to send messages remotely via RemoteAsk or RemoteTell methods

  • stash/unstash messages. See Stashing

  • can adopt various form using the Behavior feature. See Behaviors

  • can be restarted (respawned)

  • can be gracefully stopped (killed). Every message in the mailbox prior to stoppage will be processed within a configurable time period.

PID

• Acts as a unique, immutable identifier for each actor within the system. See PID for more information.
• Serves as the primary handle for sending messages to the actor, ensuring location transparency.

• Encapsulates the actor’s address and operational semantics (e.g., supervision and lifecycle management).

• Enables message delivery by directing both asynchronous (“tell”) and synchronous (“ask”) interactions.

• Facilitates supervision, where parent actors can monitor, restart, or stop child actors using their PIDs.

• Supports remote messaging by incorporating network location data, making inter-node communication seamless.

• Combines actor reference, mailbox, metrics, and synchronization primitives.

• Provides a rich API with methods like Tell, Ask, RemoteTell, and RemoteAsk to cover different messaging patterns.

• Integrates with the actor’s mailbox and supervision framework, enabling effective lifecycle management and monitoring.

Get Started

To define an actor one needs to implement the Actorinterface:

type Actor interface {
    // PreStart is called once before the actor starts processing messages.
    //
    // Use this method to initialize dependencies such as database clients,
    // caches, or external service connections and persistent state recovery. If PreStart returns an error,
    // the actor will not be started, and the failure will be handled by its supervisor.
    PreStart(ctx *Context) error
    
    // Receive handles all messages sent to the actor's mailbox.
    //
    // This is the heart of the actor's behavior. Messages can include user-defined
    // commands/events as well as internal system messages such as PostStart or lifecycle signals.
    //
    // Actors can reply to messages using async messaging patterns or configure replies inline
    // where supported. Avoid heavy synchronous workflows as they may degrade throughput in high-load scenarios.
    //
    // Tip: Use pattern matching or typed message handlers to organize complex message workflows.
    Receive(ctx *ReceiveContext)
    
    // PostStop is called when the actor is about to shut down.
    //
    // This lifecycle hook is invoked after the actor has finished processing all messages
    // in its mailbox and is guaranteed to run before the actor is fully terminated.
    //
    // Use this method to perform final cleanup actions such as:
    //   - Releasing resources (e.g., database connections, goroutines, open files)
    //   - Flushing logs or metrics
    //   - Notifying other systems of termination (e.g., via events or pub/sub)
    //
    // This method is especially important passivated actors, as it is also
    // called during passivation when an idle actor is stopped to free up resources.
    //
    // Note: If PostStop returns an error, the error is logged but does not prevent the actor
    // from being stopped. Keep PostStop logic fast and resilient to avoid delaying system shutdowns.
    PostStop(ctx *Context) error
}