resumeOnFailure

Resume on failure is a feature in software automation that allows a workflow, task, or test run to continue after an error, rather than aborting the entire process. It is typically achieved by recording progress through checkpoints or resume points and by persisting enough state to recover gracefully. The exact implementation varies by tool, but common goals include reducing rework, speeding recovery, and enabling selective re-execution.

In test automation, resume on failure may let a test suite skip already successful tests and begin from the point of interruption, or re-run only the failing portion while preserving prior results. This often requires tests and actions to be idempotent and to save a durable representation of progress, such as a checked-in report, a database flag, or a checkpoint file.

In data processing and orchestration systems, resume on failure enables a pipeline to restart from the last successful step after a crash or error. This typically involves partitioning work into discrete steps, persisting metadata about completed steps, and handling side effects to avoid duplicate processing.

Considerations for using resume on failure include ensuring data integrity, handling side effects or compensation actions for already completed steps, and choosing appropriate resume points. It may also involve trade-offs between faster recovery and the risk of skipping necessary intermediate validations.

Many tools expose this capability through configuration flags or APIs, often named resume_on_failure or resumeOnFailure. The exact semantics—whether to resume from the next step, from the last successful checkpoint, or to re-run a failed subset—depend on the system in use.