Struct DistributedWorker 

pub struct DistributedWorker<W, C, Q, RIS, MS, ES, LS, WISS, CS>
where
    W: Workflow<C, WorkItem: 'static>,
    C: Context + Merge + Default,
    Q: WorkQueue<C, W::WorkItem> + Send + Sync + 'static,
    RIS: RunInfoStore + Send + Sync,
    MS: MetricsStore + Send + Sync,
    ES: ErrorStore + Send + Sync,
    LS: LivenessStore + Send + Sync,
    WISS: WorkItemStateStore<W::WorkItem> + Send + Sync,
    CS: ContextStore<C> + Send + Sync + Clone + 'static,
{ /* private fields */ }

A distributed workflow worker that polls a work queue, processes workflow steps, and updates state in distributed stores.

Use [run_once] to process a single work item, or [run_forever] to continuously poll for work.

Implementations

impl<W, C, Q, RIS, MS, ES, LS, WISS, CS> DistributedWorker<W, C, Q, RIS, MS, ES, LS, WISS, CS>

where W: Workflow<C, WorkItem: 'static> + 'static, C: Context + Merge + Default + 'static, Q: WorkQueue<C, W::WorkItem> + Send + Sync + Clone, RIS: RunInfoStore + Send + Sync + Clone + 'static, MS: MetricsStore + Send + Sync + Clone + 'static, ES: ErrorStore + Send + Sync + Clone + 'static, LS: LivenessStore + Send + Sync + Clone + 'static, WISS: WorkItemStateStore<W::WorkItem> + Send + Sync + Clone + 'static, CS: ContextStore<C> + Send + Sync + Clone + 'static, Self: Clone,

pub fn new( workflow: W, queue: Q, context_store: CS, run_info_store: RIS, metrics_store: MS, error_store: ES, liveness_store: LS, work_item_state_store: WISS, ) -> Self

Create a new distributed worker with all required stores and workflow.

See WorkerBuilder for ergonomic construction with defaults.

pub fn set_retry_policy(&mut self, policy: RetryPolicy)

Set a retry policy for all work items.

pub async fn run_once( &self, worker_id: usize, ) -> Result<Option<(String, W::Output)>, FloxideError>

where C: Debug + Clone + Send + Sync,

Process a single work item from the queue, updating all distributed state.

Returns Ok(Some((run_id, output))) if a work item was processed, Ok(None) if no work was available, or Err on permanent failure.

Instrumentation

This method is instrumented with tracing for async span tracking.

pub async fn run_forever(&self, worker_id: usize) -> Infallible

where C: Debug + Clone + Send + Sync,

Continuously poll for work and process items, sleeping briefly when idle or on error.

This method never returns and is suitable for running in a background task.

Instrumentation

This method is instrumented with tracing for async span tracking.

Note: Returns std::convert::Infallible for compatibility with stable Rust (instead of the experimental ! type).

pub async fn heartbeat(&self, worker_id: usize)

where C: Debug + Clone + Send + Sync,

Heartbeat: update liveness store with current timestamp and update health.

Instrumentation

This method is instrumented with tracing for async span tracking.

Trait Implementations

impl<W, C, Q, RIS, MS, ES, LS, WISS, CS> Clone for DistributedWorker<W, C, Q, RIS, MS, ES, LS, WISS, CS>

where W: Workflow<C, WorkItem: 'static> + Clone, C: Context + Merge + Default + Clone, Q: WorkQueue<C, W::WorkItem> + Send + Sync + 'static + Clone, RIS: RunInfoStore + Send + Sync + Clone, MS: MetricsStore + Send + Sync + Clone, ES: ErrorStore + Send + Sync + Clone, LS: LivenessStore + Send + Sync + Clone, WISS: WorkItemStateStore<W::WorkItem> + Send + Sync + Clone, CS: ContextStore<C> + Send + Sync + Clone + 'static + Clone,

fn clone(&self) -> DistributedWorker<W, C, Q, RIS, MS, ES, LS, WISS, CS>

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.