See Your CrateDB Shards and Replicas as a Diagram

DbSchema speaks CrateDB's PostgreSQL-compatible protocol, reverse engineers tables — ARRAY and OBJECT columns included — into ER diagrams, and adds SQL editing, data exploration, and documentation for your IoT platform.

Diagram the table structures behind your sensor pipelines
Run analytics SQL with full-text predicates from the desktop
Browse huge event tables without hand-written pagination
Document nested OBJECT columns for the whole data team

available for

ISO-27001 certified

Free Community Edition No signup required 15-day Architect trial included

Free Download

DbSchema reverse engineering a CrateDB cluster into an ER diagram

Trusted by teams at

The Sensors Never Stop Writing
Somebody Still Has to Explain the Tables

Machine data piles up as readings, device metadata, and nested OBJECT payloads spread across shards. The ingestion pipeline knows the layout precisely; the humans mostly remember it.

DbSchema turns the cluster into an ER model — every table, every ARRAY and OBJECT column — so the schema becomes something you look at rather than reconstruct.

Map, Inspect & Explain
CrateDB Without Leaving the Desktop

From wire protocol to diagram, from event table to shareable reference.

Reverse engineer CrateDB tables with ARRAY and OBJECT columns

Diagram What the Pipeline Built

DbSchema introspects tables, columns, and types over CrateDB's PostgreSQL-compatible protocol — capturing the ARRAY and OBJECT structures typical of sensor schemas.

Browse CrateDB event tables in the DbSchema data explorer

Event Tables at Row Level

Filter by timestamp range or device identifier, inspect ARRAY contents inline, and let the explorer paginate tables of any size for you.

Generate HTML documentation for a CrateDB schema

A Reference for the Data Team

Publish HTML documentation of tables and column types, so engineers onboarding to the platform stop guessing what each nested field means.

The Cluster Holds the Data
The Model Holds the Design

DbSchema keeps the CrateDB schema in a model file of its own — editable on a laptop, shareable through Git, and applied only after you have read the SQL it generates.

Design the CrateDB schema without a connection

Work Away from the Cluster

Sketch tables and relationships before a single node is running.

Version the Design

Commit the model beside the ingestion code that fills it.

Read generated CrateDB SQL before running it

Read the SQL First

Every planned change becomes statements you approve before they run.

Promote CrateDB schema changes between environments

Promote with Intent

Carry one reviewed change from staging to the production cluster.

CrateDB Connection Details

Register the CrateDB JDBC driver (io.crate.client.jdbc.CrateDriver, from the CrateDB releases page) in DbSchema's driver manager and use a URL like jdbc:crate://host:5432/ — 5432 is CrateDB's PostgreSQL-compatible port. Newly installed clusters default to the crate superuser with no password; for production, append TLS parameters to the URL. A cluster exposes one SQL endpoint however many nodes it has, so no load-balancer configuration is needed. Setup details are on the CrateDB JDBC driver page.