Insert Rows
bigquery.insert_rows
ActionDatabases & warehousesconnection required
Calls the BigQuery API POST /datasets/{dataset}/tables/{table}/insertAll.
Stream rows into a BigQuery table (insertAll).
Connection
- Required
- Yes — the node will not save without one
- Auth type
- OAuth 2.0
- Providers
- Google BigQuery
- Secret shape
- {"client_id", "client_secret", "refresh_token"}
- Config it needs
- project_id
- Resolved by
- config.connection_id, else node.connection_id
Configuration
6 fields, 3 required. A dotted path is a nested object and name[] an array of objects — the editor renders the same tree.
| Field | Type | Required | Default | Constraints | Example |
|---|---|---|---|---|---|
project_id GCP project id (overrides connection). | string | No | — | — | — |
dataset Id of the dataset to act on, from bigquery.list_datasets — not its name. | string | Yes | — | — | — |
table Id of the table to act on, from bigquery.list_tables — not its name. | string | Yes | — | — | — |
rows Rows to insert (array of objects, JSON). | array | Yes | — | ≥ 1 characters; array of object | — |
insert_id_field Row key holding a stable id BigQuery de-duplicates on. | string | No | — | — | — |
deduplicate Send a per-step insertId so a retried batch is dropped (default). Turn off for high-throughput inserts: rows with an insertId fall into BigQuery's lower streaming quota tier. | boolean | No | true | — | — |
Example configuration
Every value below is the contract’s own example or default — nothing invented.
{
"deduplicate": true
}Output
One shape.
result
| Field | Type | Example |
|---|---|---|
ok | boolean | — |
inserted | number | — |
table dataset.table the rows went to. | string | — |
deduplicated Rows carried insertIds. BigQuery then drops a batch it has already seen — best effort, and only for about a minute, so a retry that backs off longer than that can still append twice. | boolean | — |
{
"deduplicated": false,
"inserted": 0,
"ok": false,
"table": ""
}See Insert Rows run in a workflow.
In a technical demo we add the Google BigQuery connection, drop the node on the canvas and let Workflow Health pre-flight it before the first real run.