DAG Join icon

DAG Join

Collect and merge results from all branches into one output.

DAG Join

AI-generated

Summary

Aggregate data from parallel branches of a workflow according to the specified join mode and output format, handling errors and timeouts as configured.

Inputs

  • Expected Branch Count (required) — Number of branches expected to join; must match the branch count defined in the corresponding DagSplit node.
  • Join Mode (required) — Determines when to emit joined data: wait for all branches, first branch, or any branch.
  • Output Format (required) — Controls how the collected branch data is output: merged into one object, as an array of branches, or passthrough of the first branch data.
  • Error Strategy (required) — Defines how errors from branches are handled: continue on error, stop on first failure, or collect all errors.
  • Global Timeout (ms) (required) — Maximum time to wait for branches before emitting data with a timeout flag.

Output shape

a list of items containing the combined results from joined branches according to the selected output format

Outputs one or more execution groups, each keyed by a unique DAG execution ID. If 'stopOnFirst' error strategy is selected and a branch has failed, the node throws an error and halts. If 'collectErrors' is chosen, detailed errors are included in output. Output includes metadata like execution ID, join mode, completion timestamp, and timeout flag if applicable.

Links

Discussion