Delight

Visualizes executor core usage broken down by categories and aligned with Spark job timelines, making it easy to correlate performance issues like slow shuffles, as shown in the before/after comparison images.

Reports peak memory usage per executor segmented into JVM, Python, and other processes, helping tune container memory sizes to avoid waste, demonstrated in the memory graph screenshot.

Entirely free of charge and works on any Spark platform—open-source, commercial, cloud, or on-premise—without vendor lock-in, as stated in the compatibility section.

Provides access to the Spark UI without setting up and maintaining a separate Spark History Server, simplifying deployment for users.