Databricks and ClickHouse, Better Together

Date: Tuesday, September 29, 2026
Time: 11:00AM PDT / 2:00PM EDT
Duration: 3 hours
Location: Virtual - Zoom Webinar
Cost: FREE
Format: Instructor-led training with hands-on labs

What you'll learn: In this hands-on training, participants will learn how to use Databricks and ClickHouse together to build a high-performance analytics architecture. Rather than replacing Databricks, participants will learn to complement it by setting up a hot&cold architecture, using ClickHouse as the "hot layer" for sub-second dashboards and high-frequency event analysis, while keeping Delta Lake as the "cold layer" for historical archives. Participants will migrate schemas from Databricks to ClickHouse, map Databricks data types to ClickHouse's more granular type system, and design efficient primary keys and sort orders for analytical workloads. The course covers multiple data ingestion methods — including the Spark Connector, S3Queue table engine, and the DataLakeCatalog database engine for querying Unity Catalog data directly — and introduces strategies for handling updates and deletes, building pre-aggregated models with AggregatingMergeTree, and optimizing real-time dashboards using incremental Materialized Views. Through a series of guided labs, learners will build and query a complete hot/cold data pipeline spanning both platforms.

By the end of this training learners will be able to:

• Explain the complementary strengths of Databricks and ClickHouse and when to use each
• Map Databricks concepts (Delta tables, Liquid Clustering, SQL Warehouses) to their ClickHouse equivalents
• Redesign Databricks schemas for ClickHouse using optimal data types, primary keys, and sort orders
• Choose the right data ingestion method (Spark Connector, S3Queue, DataLakeCatalog, ClickPipes) for each migration scenario
• Design a hot/cold architecture that keeps recent, high-query data in ClickHouse and historical data in Delta Lake
• Handle updates and deletes using mutations, ReplacingMergeTree, and lightweight operations
• Build pre-aggregated models with AggregatingMergeTree and incremental Materialized Views for instant-loading dashboards
• Query Delta Lake and Unity Catalog data directly from ClickHouse using the DataLakeCatalog database engine

The material is presented in modules and will include the following:

Module 1: Introduction
Module 2: Understanding ClickHouse
Module 3: Preparing to Migrate
Module 4: Create Data Pipeline
Module 5: Build a Real-Time Dashboard

Prerequisites:

• Basic familiarity with Databricks
• Basic familiarity with ClickHouse
• SQL proficiency