Getting started

Quickstart

From zero to your first query. You'll install the CLI, sign in, point Polnor at your cloud, and run SQL against an Iceberg table, all on your own infrastructure.

Prerequisite. A Polnor workspace. Don't have one yet? Request a demo, onboarding takes a few days and we provision the namespace with you.

1 · Install the CLI

# macOS / Linux
curl -fsSL https://polnor.net/install.sh | sh
polnor --version

2 · Authenticate

polnor auth login --workspace acme
# opens a browser, or use --email / --password for headless

3 · Connect your cloud

Register the credentials Polnor uses to provision compute and read your storage. Everything stays in your account, see Connect your cloud.

polnor providers add ovh --name prod \
  --application-key "$OVH_APP_KEY" \
  --project-id "$OVH_PROJECT_ID"

polnor storage add --type s3 \
  --endpoint https://s3.gra.perf.cloud.ovh.net \
  --bucket acme-lakehouse --region gra

4 · Start a SQL warehouse

A warehouse provisions a VM in your cloud with DuckDB + Spark sidecars. It auto-stops when idle.

polnor sql warehouse start analytics
# ~3 min to go from "starting" to "running"

5 · Run your first query

polnor sql "SELECT count(*) FROM acme.sales.orders" \
  --warehouse analytics

That read ran on DuckDB, in your cloud, against an Iceberg table on your bucket. Writes route to Spark automatically.

Next steps

Lakehouse & catalog, create databases and tables.
Jobs & pipelines, schedule multi-task DAGs.
SDK & CLI, automate from Python or Go.