Skip to content

Feature request: Retry policies for queued operations #9

Description

@dombrovsky

Feature request

Allow queued operations to opt into retry behavior when they fail transiently.

Rationale

Serialized background work often needs to survive intermittent failures. Without TaskFlow-level support, each consumer must build and maintain its own retry loop, cancellation handling, and error reporting.

Use case

A component owns a periodic synchronization loop. A temporary service or storage failure should not permanently stop that loop, while shutdown must still cancel it promptly.

Desired behavior

  • Eligible failures can cause the operation to be attempted again.
  • Cancellation and TaskFlow disposal stop further attempts promptly.
  • Non-retryable or exhausted failures remain observable by the caller.
  • FIFO ordering and the effect of retries on later queued work are predictable and documented.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions