n8n replaces lightweight ETL scripts with visual workflows that extract
data from databases and APIs, transform it through Set and Code nodes,
and load it into data warehouses, Google Sheets, or Airtable. Scheduled
triggers or webhooks drive the pipeline, and built‑in nodes for
PostgreSQL, MySQL, and over 400 connectors make database sync
achievable without writing SQL orchestration code.
[1]

How do you extract data from PostgreSQL and transform it with n8n?

Connect the Postgres node with SSL credentials and write a SELECT
query in the “Execute SQL” operation. The output JSON then passes
through a Set node that renames or reshapes fields—for example
converting timestamps to ISO strings or masking PII columns. Finally,
a second Postgres node inserts the cleaned data into another table.
[1]

For large datasets, use the “Execute in Batches” option on the source
node to fetch records in chunks, avoiding memory issues. This pattern
is also detailed in the
SplitInBatches loop guide
for processing thousands of rows.

How do you sync data between two MySQL databases using n8n?

Set up two MySQL nodes—one for source and one for target. The source
runs a SELECT with a WHERE clause filtered by the last processed
timestamp; the target uses an INSERT … ON DUPLICATE KEY UPDATE (upsert)
for each row. A Set node maps column names between schemas before
the insert.
[2]

Schedule this workflow with a cron trigger for nightly sync. To
guarantee idempotency, always store the last synced ID or timestamp
in a static data node or external key‑value store. For more on
reliable scheduling, see our
n8n trigger types guide.

How do you use Google Sheets as an automated reporting database with n8n?

Use the Google Sheets node with a service account. The “Append” operation
adds new rows to a sheet, while “Update” overwrites specific cells. A
workflow can pull data from an API, transform it with a Set node, and
append a summary row to a tracking sheet every hour.
[3]

For dashboards, combine multiple data sources into one sheet by
running parallel branches that each write to a separate tab. This
approach replaces expensive BI connectors for small‑team reporting,
as demonstrated in our
marketing and data use cases.

How do you automatically sync records between a database and Airtable?

Create a workflow with a database trigger (e.g., new row via polling)
that feeds into the Airtable node. Configure the Airtable node with
a Personal Access Token and select the base/table. Use the “Create”
operation for new records and “Update” for existing ones identified
by a lookup on the Airtable record ID.
[4]

For bidirectional sync between a database and Airtable, pattern the
workflow as described in the
two‑way CRM sync guide
—listening on both ends and using upsert logic to avoid duplicate rows.

How do you map and reshape fields with the Set node in an ETL pipeline?

The Set node maps, renames, and deletes columns using a visual field
editor. Enable “Keep Only Set” to drop all unlisted fields, or toggle
individual fields to add defaults, convert types, or compute expressions
like {{ $json.orderTotal * 1.2 }}. This turns raw API
responses into clean, load‑ready records.
[5]

For advanced reshaping, chain a Code node after the Set node to run
JavaScript transformations on nested structures. The combination
covers both simple and complex ETL mappings. Learn more about
expressions in the
node expressions & techniques guide.

How do you perform an upsert (update or insert) in n8n for databases?

Use the “Upsert” operation available on the Postgres or MySQL node. Set
the “Column to Match On” to a unique key like email or
id. n8n generates a MERGE/ON CONFLICT statement that inserts
new rows or updates matched rows with the specified column values, all
within a single database call.
[1]

This pattern avoids the need for separate SELECT‑then‑INSERT logic.
When the datasource is an API, first transform the payload with a Set
node to match the database column names, then feed it directly into
the upsert node. For error handling during bulk upserts, wrap the
execution in an
error workflow with retry logic.