Why Fabric Capacity Throttles During Off-Peak Hours Despite Low Daily Usage
Microsoft Fabric smooths background operations such as scheduled notebook jobs and dataset refreshes over a 24-hour window, while interactive operations are smoothed over one minute. As a result, a heavy overnight workload can trigger throttling warnings even when total daily compute use looks low. The article notes that engineers often misread these symptoms as a sign they need a larger capacity SKU. To diagnose the issue, it recommends checking the Background and Interactive tabs in the Fabric Capacity Metrics app, and suggests staggering background jobs with pipeline Wait activities. It also advises optimizing Direct Lake semantic models and Lakehouse tables, for example with Liquid Clustering, to reduce unnecessary compute consumption.
This is an AI-generated summary. ShortSingh links to the original source for the complete article.
Discussion (0)
Log in to join the discussion and vote.
Log in