Connecting to the engine - e6data Docs

Connect BI tools, SQL clients, drivers, and your own code to an e6data query cluster. e6data exposes one authenticated endpoint that speaks three protocols, so you can connect with the e6data drivers, with plain HTTP, or with any PostgreSQL client you already have - no new driver required for the Postgres path.

Pick a path

You want to connect…	Use	Driver to install
Tableau, pgAdmin, psql, any Postgres tool	PostgreSQL protocol	None (stock Postgres)
DBeaver, DbVisualizer, a Java app, Zeppelin	JDBC	e6data JDBC driver
Python, a notebook	Python connector or REST	`pip install` (connector) or none (HTTP)
A REST integration in any language	REST API for SQL queries	None

See One-endpoint / three-protocols model for how this fits together.

Core concepts

Cluster name - the routing key. Every connection names the target cluster (a header on the gRPC/HTTP paths; encoded in the database name on the Postgres path). You never address individual nodes.
Catalog and schema - what you query.
Access token - a personal access token authenticates you (the password on the Postgres path, or exchanged for a session on HTTP).
TLS is always on, and the first query to a suspended cluster auto-resumes it (allow ~30–90 s).

Find your exact values - host, cluster names, and a ready-to-copy connection string per tool - in the Console under Cluster → Connectivity, and create a token under Profile → Settings.

Connection-string cheat sheet

e6data JDBC : jdbc:e6data://<HOST>:443/secure=true&cluster-name=<CLUSTER>&catalog=<CATALOG>&database=<SCHEMA>
PG JDBC     : jdbc:postgresql://<PG_HOST>:5432/<CLUSTER>/<CATALOG>?user=<EMAIL>&sslmode=require
psql        : psql "host=<PG_HOST> port=5432 dbname=<CLUSTER>/<CATALOG> user=<EMAIL> sslmode=require"
SQLAlchemy  : e6data://<EMAIL>:<TOKEN>@<HOST>:443/<CLUSTER>?secure=true
HTTP REST   : POST https://<HOST>/api/v1/authenticate → POST https://<HOST>/api/v2/sql/statements

See also

Last modified on September 29, 2026