This article shows an interesting workflow for logging output of data pipeline activities into a KQL database (which is very commonly used for logging). In this case, it's the Semantic Model Refresh activity, but could be extended to use for many activity types, pretty cool!