Trait BackgroundJobProvider 

pub trait BackgroundJobProvider: Send + Sync {
    // Required methods
    fn definition(&self) -> &JobDefinition;
    fn execute(&self, ctx: JobContext<'_>) -> Result<JobOutcome, FnError>;
}

A background-job provider.

Required Methods

fn definition(&self) -> &JobDefinition

Static definition (schedule, concurrency, timeout, docs).

fn execute(&self, ctx: JobContext<'_>) -> Result<JobOutcome, FnError>

Execute one run of the job.

Threading policy

• Driven from Tokio via tokio::task::spawn_blocking. The host scheduler runs this synchronous method on a blocking worker thread so it never stalls the async runtime.
• Must not block the runtime directly. If the job needs to perform I/O, it must do so on the current (blocking) thread — never call block_on against the host runtime from inside execute.
• Must observe JobContext::cancel cooperatively. Poll CancellationToken::is_cancelled at every safe point (between batches, before long compute, before issuing each query). The scheduler trips the token on shutdown / reload / explicit cancel; an unresponsive job stays alive until the process exits.
• Errors propagate as FnError. Panics are caught at the scheduler boundary and recorded as a failed run; they do not crash the host.

See docs/PLUGIN_THREADING.md for the long-form rationale.

Errors

Returns FnError on execution failure. The host’s scheduler honors the JobDefinition::retry policy.

Dyn Compatibility

This trait is dyn compatible.

In older versions of Rust, dyn compatibility was called "object safety".

Implementors