Power BI Performance Tuning Aggregation Part 1 Overview

[@SomeGuyInBrisVegas]

Fascinating to see old patterns reborn in new software. This pattern has been around for over 20 years in my experience. Truly the basics of computer science do not change. 🙂 Great video.

[@NikSargent]

loved this video and part 2, which is so helpful.. . but I am stuck on the following 2 things: 1) what is a good design pattern to build aggregation tables for facts that are timebased, so I need hourly, daily, weekly, month views - i am stuck on how to join these tables with the right relationship (create a combined field such as date-hour to join facts to the date-hour aggregation?) and also secondly, which is more of a blocker: how to set up aggregations in fabric directlake, using the online report editor (not desktop). Is this automatic aggregation something that is limited to desktop only?

[@tanababa]

Thank you Reza for the great explanation. I am a little confused with the following scenario: In the model you presented, Say there is a measure [Total rows] which counts the number of records from the fatcTable. New records are created in the table (in the DB) every second. However, using a card to show the [Total rows], PBI will use the imported aggregated table. That number will not show me the correct amount of records (because the report is refreshed once a day) Naturally, I would like to enjoy a faster-optimized model ...but still display the correct results What am I missing? Thank you, Tamir

[@Milhouse77BS]

Ooh, now that I have Fabric capacity, aggregations could be useful. Thanks.

[@RADACAD]

You're welcome :)

[@hosseinaghamohamadi3887]

خیلی مخلصیم آقا رضا