Analytics Engineering with Microsoft Fabric and Power BI

Chapter 13. Optimize Enterprise-Scale Semantic Models

A Note for Early Release Readers

With Early Release ebooks, you get books in their earliest form—the author’s raw and unedited content as they write—so you can take advantage of these technologies long before the official release of these titles.

This will be the 13th chapter of the final book. Please note that the GitHub repo will be made active later on.

If you’d like to be actively involved in reviewing and commenting on this draft, please reach out to the editor at [e-mail].

A few years ago, a logistics company we’ll call Northwind Cargo asked us to look at one of their Power BI models. The story they told us was familiar. They had a single semantic model that powered the operations dashboard for their entire European network. It had grown gradually over three years: from a handful of routes to 1500, from a few months of shipment history to four years, from 20 users to 350. Nobody had touched the model in months because nobody needed to. It just worked.

Then, one Tuesday in March, the morning refresh failed. The error message was this: Resource Governance: This operation was canceled because there wasn’t enough memory to finish running it.

The team retried it. It failed again. They retried it from a different machine. Same error. They opened the SKU documentation, saw that the F64 capacity gave them 25 GB of memory, looked at their model size in the workspace (around 14 GB at rest), and concluded the platform ...