Lost in Data Caught in the Trace

  I for one was/am very excited about Direct Lake as a means of supplying data to Power BI in a meaningful live or near-live fashion. This has long been an issue for reporting and different platforms have dealt with this in different ways. Power BI has typically queried the data at source to avoid copying data into the Power BI model. It's likely if you have tried this you will have experience of where this has slowed down reporting. Direct Lake promises to resolve this by removing the database engine entirely and reading directly from One Lake Parquet Files. For an overview of see here . Within this document, there are a set of limitations. The first of which gave me pause to try to understand what it means and how it limits Direct Lake. Known issues and limitations. Currently, Direct Lake models can only contain tables and views from a single Lakehouse or Data Warehouse. However, tables in the model based on T-SQL-based views cannot be qu...

 Beta is over. Here comes the Live Exam. Based on Pearson Vue the exam is going to go live at the start of April. So there is plenty of time - nicely aligned to a trial subscription length - to crack open the exam content.  The exams seem to be going live the first few days of April (not the 1st  ) . Based on my experience with the Beta version I have compiled the following help. Things to consider when sitting this exam. MS Learn is available. If you have sat the PL-300 on Power BI there is a lot of useful content overlap. The exam will cover technical capability and "what should you do/use in this situation" type questions. Technical things to know. SQL - a good deal of questions were on this - typically select the correct keyword to add to this statement. Know your SQL functions Know your SDC types Pyspark - again, some keyword selection typically how to load to a delta table - overwrite and append Look at this code and this output - evaluate How to Transform data and ...