Create a Lakehouse cluster

This guide explains how to create a Lakehouse cluster in Hybrid Manager (HM). Lakehouse clusters provide scalable analytical compute for fast queries on Apache Iceberg or Delta Lake tables stored in object storage.

For architectural background, see Lakehouse clusters in Hybrid Manager.

Why create a Lakehouse cluster?

By creating a Lakehouse cluster:

• You unlock the ability to run fast Postgres SQL queries on large datasets in object storage.
• You can query Tiered Tables offloaded from PGD or external Iceberg/Delta Lake tables.
• You provide compute for BI tools and ad-hoc analytics on your lakehouse data.

This is a foundational step for building your analytics architecture in Hybrid Manager.

Goals

After completing this How-To, you will be able to:

• Deploy a Lakehouse cluster.
• Connect to it using standard Postgres clients.
• Configure storage and catalogs for querying Iceberg and Delta Lake tables.
• Use the cluster to query existing or newly defined lakehouse tables.

Prerequisites

Before you begin:

• You have access to a Hybrid Manager environment with permissions to create Lakehouse clusters.
• You have object storage available (S3, GCS, MinIO, etc.).
• You have Iceberg or Delta Lake tables available — or plan to define them after cluster creation.
• You have an Iceberg catalog prepared if working with Iceberg tables:
• Configure an Iceberg REST catalog connection

Steps

Step 1: Initiate cluster creation

• In the HM dashboard, click Create New.
• Select Lakehouse Analytics or Analytical Cluster.
• Confirm that you are creating an Analytics Cluster for querying open lakehouse formats.

Step 2: Choose your path

You will be prompted to choose:

• Templates — Use a pre-configured template (if available).
• Custom build — Click Start from Scratch for full configuration control (recommended for most cases).

Step 3: Configure cluster settings

Cluster Settings

• Cluster name — Enter a unique name (e.g. sales_analytics_lakehouse).
• Password — Enter a strong superuser password.
• Tags — Optionally add tags (e.g. environment:production, project:q3-reporting).
• Deployment location — Select cloud provider and region.
• Database type — Choose EDB Postgres Extended Server or Advanced Server.
• Postgres version — Select supported version with Lakehouse extensions.
• Instance size — Select compute size appropriate for your workload.

Additional Settings

• Networking — Configure VPC, subnet, security groups, IP whitelisting.
• Backups — Configure backup settings if applicable.

Step 4: Review and create

• Review the Cluster Summary.
• Confirm all settings.
• Click Create Cluster.
• Monitor progress in the HM dashboard.

What you can do next

Now that your Lakehouse cluster is provisioned, you can:

• Connect to the cluster to run SQL queries: Use standard Postgres clients (psql, DBeaver, pgAdmin) with HM-provided connection details.

• Configure storage and catalogs to access data:

• Configure a PGFS storage location — if querying Delta Lake or filesystem-based Iceberg.

• Configure an Iceberg REST catalog connection — if querying catalog-managed Iceberg tables or using PGD Tiered Tables.

• Define and query Lakehouse tables:

• Query existing Iceberg tables

• Query Delta Lake tables

• Query Tiered Tables offloaded from PGD:

• Query Tiered Tables from PGD and Lakehouse

Related topics

• Lakehouse clusters in Hybrid Manager
• Analytics in Hybrid Manager