Batch Operations

Batch nodes process multiple items efficiently in a single step. Instead of using a For Each loop (which processes one item at a time), batch nodes handle arrays in bulk with configurable concurrency and error handling.

Batch vs. For Each

Configuration

Operations

Batch HTTP Request

Send the same HTTP request for each item in the array:

Batch Create Record

Create multiple database rows at once:

Batch Update Record

Update multiple rows based on a matching key:

Batch Delete Record

Delete rows matching each item:

Output variables

{{batch.results}}          → array of results (one per input item)
{{batch.succeeded}}        → count of successful items
{{batch.failed}}           → count of failed items
{{batch.errors}}           → array of error details for failed items
{{batch.total}}            → total items processed

Error handling modes

Stop all

Stops processing immediately when any item fails. Items already processed stay processed; there's no rollback.

Skip and continue

Skip the failed item and continue with the rest. Failed items are reported in {{batch.errors}} at the end.

Retry then skip

Retry failed items up to N times with exponential backoff. If all retries fail, skip and continue.

Common patterns

Bulk API sync

Find All (contacts updated today) → Batch HTTP Request →
  POST to CRM API for each contact (concurrency: 10) →
  If-Else (any failures?) → Send Email (report failed syncs)

Bulk enrichment

CSV Import → Filter (rows without company data) →
  Batch HTTP Request (enrichment API, concurrency: 5) →
  Batch Update Record (write enrichment data back)

Data cleanup

Manual Trigger → Find All (inactive users, older than 1 year) →
  Batch Delete Record (concurrency: 20) →
  Log (deleted {{batch.succeeded}} inactive users)

Performance tips