Validate a pipeline

Validation runs a limited preview of a pipeline so you can confirm that Snaps are configured correctly and inspect sample data before you execute the pipeline with full data.

Validation is one of the most common activities when you build a pipeline. When you validate, the Designer performs a limited execution of the pipeline: each Snap generates preview data that you can inspect, without processing the full data set. This lets you catch configuration errors and check intermediate data before you run the pipeline.

In the Designer toolbar, Validate and Execute are separate actions. Validate produces preview data for inspection; Execute runs the pipeline to completion with full data. You do not have to validate before you execute.

Validation button states

The Validate button label reflects the result of the most recent validation:

  • Validate: No validation has run yet for the current changes.
  • Re-validate: The previous validation passed. Click to validate again.
  • Retry: The previous validation failed. A popover shows the validation errors; click Retry to validate again.

When a validation finishes, a popover indicates whether validation passed or failed. If validation fails, the popover opens automatically with the error details.

Reusing cached results versus a full re-validation

By default, a validation reuses cached preview data where possible, which is faster. To force a full re-validation that clears the cached preview data and re-runs every Snap, hold Shift while you click Validate. When a validation fails, you can also choose Start new validation in the popover to clear the cache and run a full validation.

Environment settings for validation

Pipeline validation can be enabled or disabled at the environment (Org) level and at the project level. If validation is disabled, the Validate button is unavailable and Snaps do not generate preview data.

These settings are managed by administrators; standard users cannot change them. Only environment (Org) admins can access Admin Manager.

How to validate a pipeline
  1. Open the pipeline in Designer.
  2. In the toolbar, click the Validate icon.

    The Designer validates the pipeline and updates each Snap's color to reflect its state. The icon shows a green checkmark when validation succeeds and a red retry icon when it fails. If the pipeline has unsaved changes, the Designer saves it first and then validates automatically.

    Note: Validation does nothing if the pipeline contains no Snaps. Validation must also be enabled for your environment; see validate-pipeline.html#task_cg5_pq4_bkc__env-settings.
  3. Review the results.

    Inspect the preview data on each Snap's output view, and review the validation popover. To force a full re-validation that ignores cached data, hold Shift while you click Validate.

Stopping validation
  1. While a validation is in progress, click the Validate icon again.

    While the pipeline is validating, the button shows a stop icon and a Stop Validation tooltip. Clicking it stops the current validation rather than starting a new one.

    Note: Because clicking Validate again stops the in-progress validation, repeatedly clicking the button does not queue or restart validations. After you stop a validation, the preview cache may be incomplete, so the next validation automatically runs in full (without reusing cached data).

    Validation stops, the Execute button is re-enabled, and the pipeline returns to an idle state.

After validation passes, the Snaps show their validated state and preview data, and you can execute the pipeline or continue editing.