Insight: Visualizing Redshift View Dependencies with Graphviz

A Visual Twist on a Classic Problem

In a recent post, we explored how to trace recursive view dependencies in Amazon Redshift using SQL and Python. That post was inspired by a real Slack question from our teammate Prasad, who needed to understand how one view depended on another across layers of a legacy data platform. If you missed it, check out the full solution here:

Insight: Recursive View Dependencies in Redshift? Here's the Workaround

Today, we’re taking that solution one step further—by visualizing it.

From Tree to Graph: Why It Matters

When you're working with a large set of views, seeing the dependency tree as a visual diagram can change how you think. Instead of squinting at console logs or parsing text output, you get a clean, directional map that shows the flow of data logic. Arrows make the lineage intuitive. Branching highlights architecture flaws. Cycles (if they exist) pop immediately.

To support this, we've added a Graphviz-compatible generator script to the same GitHub Gist from the original article. The script, generate_view_graph_dot.py, takes a list of view dependencies and outputs a DOT file that can be rendered with any Graphviz tool.

Why the Gist Has More Than One Script

We deliberately kept everything in a single GitHub Gist so readers could see the full context: SQL queries, the Python tree walker, and now the DOT file generator. If you're only interested in the visualizer, just use generate_view_graph_dot.py and ignore the rest—but if you're solving the full problem from scratch, the other files are there to help.


Example Output and How to Run It

Here's what a typical DOT file might look like:

Dot

digraph ViewGraph {
    "v_base" -> "v_mid";
    "v_mid" -> "v_top";
}    

You can visualize this using the online Graphviz viewer at dreampuf.github.io/GraphvizOnline or by installing Graphviz locally and running:

Bash

dot -Tpng view_graph.dot -o view_graph.png   

How to Generate the DOT File

To run the generator, simply pass in your list of dependencies:

Python

from generate_view_graph_dot import generate_dot

deps = [('v_mid', 'v_base'), ('v_top', 'v_mid')]
generate_dot(deps)

This will output a view_graph.dot file you can immediately visualize.


Next Steps

The script can be extended to color root nodes, group schemas, or even flag potential bottlenecks. For now, though, it gives you a fast, readable output with minimal effort—perfect for audits, architecture reviews, and data platform onboarding.

If you'd like to see an advanced version or a CLI tool built around this, let us know. We’re always thinking a few steps ahead.

Need AWS Expertise?

We're happy to help you with your AWS projects!  Feel free to contact us.

Email: info@pacificw.com

Image: Gemini