Migrate your Power BI Semantic Models to Direct Lake

[@DanielWeikert]

Question would be when this makes sense. You certainly do not want to have all your models in direct lake mode as it requires to much Fabric compute resources. An indeph recommendation guide on that would be helpful

[@anasjabrane894]

Shootout to this genius solution! Making difficult things easy it's extremely difficult but someone has to do it for us all🎉😂

[@sumitnangare2584]

Wow! This is lifesaver ❤

[@sandeepbarge4699]

Is my undetstanding correct that it introduces new tables in Lakehouse? Is it worth migrating existing models to Direct Lake in this manner considering data redundency it creates and overheads it adds for building additional ETL jobs to ingest data in new tables rhat it creates? Woudn't it be worth manually changing initial few steps of each table in report manually to point it to lakehouse tables in direct lake mode?

[@noahhadro8213]

Thanks for the video. Love guy in the cube content. My concern is that when I do this my data flow takes more resources to refresh than my original import semantic model. Using (fabric capacity tracker dashboard from Microsoft). So using more of the capacity means I pay more. In addition, the dataflow refresh takes about 10 to 30 minutes longer pending on the data. Testing I did a Dataflow import of one table vs importing the table with exact same PQ code. Dataflows gen 2 are slower. I’ve followed all the best practices from you guys and Alex powers. Dataflows gen 2 is slower.

[@ProMeiFilmsOfficial]

Question, we use the DEV - STG/UAT - PROD lifecycle we have DEV LH, STG LH, and PRD LH Is it possible to redirect a Semantic model created on STG to PRD?

[@josecardenas2736]

If is have multiple tables imported and a other tables coming from another Semantic Model this will work? I have a shared mentantic model that is used by multiple dashboards but I'm connecting to a external dataset to get some. Tables.

[@Tony-dp1rl]

Can Microsoft stop changing technologies every 3-5 years please. Getting to the point now where it is a hindrance more than an advantage to adopt anything in the Azure data-related product-line.

[@DuAuskenner]

Coolio

[@juanpablorvvv]

They are just trying to sell Fabric. It’s too expensive even for companies.

[@TafadzwaMundida]

Are the alternatives to Fabric cheaper?

[@jasestu]

Come on guys, your a huge company, get better audio.