DbSchema connects to Yellowbrick's MPP warehouse over its PostgreSQL-compatible protocol and reverse-engineers tables, views, distribution keys, and sort keys into a diagram — the data-placement decisions of the cluster, made visible.
Trusted by teams at
Yellowbrick's front end feels like PostgreSQL, but query speed is decided lower down:
how rows distribute across nodes, how sort order aligns with zone maps, whether a join
forces redistribution. Those decisions hide in DDL — one DISTRIBUTE ON
clause at a time.
DbSchema surfaces them: keys and structures on a diagram, so skew and misplaced joins are things you notice rather than discover.
PostgreSQL habits welcome; the appliance-scale details come along.
Reverse engineering captures tables and views with their distribution and sort keys, so the column driving each table's spread across nodes is part of the picture.
Author joins, aggregations, and CTEs; tag interactive work with hints like /* ybworkload:adhoc */; and read EXPLAIN plans to find redistribution steps before they hurt.
Column filters push down to the warehouse, where zone maps skip non-matching segments — so checking a loaded table by date or ID range stays quick.
Export the warehouse model as interactive HTML5 documentation — diagrams, definitions, and comments in a self-contained package. Compliance reviewers and downstream data consumers read it in a browser, with no cluster credentials involved.
Explore the Documentation Tool
Yellowbrick listens on port 5432 — PostgreSQL's port, but the standard
PostgreSQL driver is not interchangeable here. Get the dedicated driver
(yellowbrick-jdbc-*.jar, class com.yellowbrick.jdbc.Driver) from the
Yellowbrick support portal, giving the URL
jdbc:yellowbrick://yellowbrick-host:5432/mydb.
Add ?sslmode=require to enforce SSL — Yellowbrick Cloud requires it and manages
credentials through its console; for on-premises deployments, the client must reach port 5432
on the manager node through the firewall. Details on the
Yellowbrick JDBC driver page.
One free download covers the Community Edition and a 15-day Architect trial — accountless.
Teams working with Yellowbrick often use these engines too. Explore dedicated guides and JDBC setup for each.