Skip to main content

Batch Node

The Batch node processes a list in chunks and can run multiple chunks in parallel. It has two outputs:

  • continue: runs once after all batch processing has finished.
  • process: runs for each batch slice.

Required settings​

  • Input Array Path: context path that must resolve to a ContextList.
  • Output Array Path: context path used to collect per-batch outputs.
  • Batch Size: number of items per batch.
  • Parallel Batches: maximum number of active batch slices at once.

Runtime behavior​

  1. The node resolves the input list from Input Array Path.
  2. It schedules up to Parallel Batches batch slices.
  3. Each slice is written back into context at the same input path before invoking the process path.
  4. When all slices complete, outputs are merged and execution continues through continue (if connected).

If process is not connected, the node skips batch execution and immediately attempts the continue path. If continue is not connected, execution ends after batch handling completes.

Validation and failure rules​

The node fails if:

  • BatchSize is less than 1
  • ParallelBatches is less than 1
  • Input Array Path is empty
  • Input Array Path does not resolve
  • resolved value is not a ContextList

Output merge notes​

Per-batch values found at Output Array Path are merged back into a single final context after all slices complete. If Output Array Path is empty, no batch output merge is performed.

The merged context then flows into the continue branch.