Choosing to Model in PowerBI vs SSAS Tabular
HTML-код
- Опубликовано: 7 фев 2025
- Get cloud confident today! Download our free cloud migration guide here: success.pragmat...
Learn a few things to consider when deciding whether to build your semantic data layer in Power BI vs SSAS Tabular.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
On-Demand Learning Courses FREE Trial: www.pragmaticw...
Free training every Tuesday at 11am EST pragmaticworks...
Pragmatic Works Blog: blog.pragmaticw...
Let's connect:
Twitter: / pragmaticworks
Facebook: / pragmaticworks
LinkedIn: / prag. .
RUclips: / pragmaticworks
Pragmatic Works
1845 Towncenter Blvd Suite 505 Fleming Island, FL 32003
Phone: (904) 413-1911
Email: hello@pragmaticworks.com
Thanks so much. I am a huge advocate of using SSAS over Power Bi for modeling especially when we want to have a long term solution keeping data volume in mind and reuse by other tens Nd business segments for reporting. Sometimes PowerBI ‘s self service model comes in between this and confuses my clients. Really challenging to explain to clients sometimes. Thanks for the video.
Amazing video, exactly what i was looking for as someone who has not too much experience in either of those products
Great. This really helps clarify things. Seems PowerBI could be the 1 IT guy/data analyst DIY whereas SSAS has the horse power and scope for an entire organization. Thanks for posting.
Chris, we are glad to hear that you liked the video. I am not sure if you know, but we offer many On-Demand Learning courses, such as, Power BI, Azure, SQL Server, Data Science, Business Intelligence and much more. You should check out our FREE trial: www.pragmaticworkstraining.com/trial-registration/?source=odl_youtubeorg - No credit card required and you will get instant access of all our courses.
Perfect!. Short and to the point, thank you so much
AWESOME! I am now deciding this and this is very helpful.
Finished watching
Short and super informative! Thanks alot!
Great video ! Thanks. Anyway I think a lot of things are going to change with latest announcements on Power BI Premium. XMLA endpoint read/write, large datasets with incremental refresh and so on. Shared datasets that allow the separation between modeling and visualization.
Yannick, I agree that the XMLA endpoint will change how we think about these considerations. Also, sharing datasets with "new" workspaces is awesome!
@@bobrubocki5305 Definitely :)
GREAT POST, THANKS FOR THAT.
Good afternoon! Not sure if you will read this comment but here I go. I am dealing with the scenario of adopting Column Level Security to a model which is already prepared for reporting consumption. The strategy is to have a self-service model for BU which is already done and presented in SSAS (composed of 10 or fewer views). A minority of advanced users access these views directly and develop their own models to create the subsequent reports afterward.
At this point, in terms of pros and cons, which security structure would be better? Develop the CLS within SSAS or PBI? Considering the maintainability and scalability to the entire organization.
One approach is to maintain the self-service model (semantic layer) within SSAS, develop the CLS there, and expose this to PBI to let others use this dataset and create their reports. This could be OK as I could secure the data presented to those advanced users as well.
Another approach could be to develop the CLS within PBI with the Tabular Editor extension (which is a thing I haven't made before and I don't know what impacts can have in terms of architecture), but I'm not sure how I will be dealing with the access of those advanced users, maybe certifying and standardizing a dataset in PBI as the source-of-truth for the rest of users?
In summary, what would be the pros and cons of these two approaches regarding Column-Level Security?
Thank you. Was quite informative.
Hi Vasilisa, we are glad you liked the video. I am not sure if you know, but we offer many On-Demand Learning courses, such as, Power BI, Azure, SQL Server, Data Science, Business Intelligence and much more. You should check out our FREE trial: www.pragmaticworkstraining.com/trial-registration/?source=odl_youtubeorg
helpful for me to explain different usecases thanks
You are welcome
Most important one is speed I think. Same data (1GB) ssas vs power bi.. which does open faster?
We are developing reporting solutions on top of our transactional processing system. Many of our customers are wary of the cloud, we'd like to offer a on-premise solution using PowerBI + PowerBI Report Server. However, the report server doesn't support the PBI incremental refresh policy (RangeStart/RangeEnd params + partitioning). Doing some research I found you would have to use SSAS tabular model and implement this process manually. Is this a recommended architecture (e.g. SSAS tabular + PowerBI Report Server & incremental refresh) or are there any gotchas?
Good info, but more details around performance, user count and data size would be great. If pbix model has 1Gb max, how does one practically plan not to exceed it at the onset of a project? It is almost impossible to know the data size growth over time. Also what do 150 concurrent user sessions do to the performance of pbix model?
Thanks. Question: it is possible to import a data model from a PBIx file in SSAS such as a data model from Power Query/Excel? 🤔
Or it would be logical copy-and-paste the model from the PBIx to Power Query and then imported on SSAS? 🤔
It's possible, yes, but not officially supported. With Power BI Desktop open on your PC, an instance of SSAS Tabular runs on your PC. You can find the instance name and port # of the local SSAS Tabular instance. Then use tools like SQL Server Management Studio and Visual Studio to connect to the instance. BTW, the XMLA endpoint (currently available in Premium) will likely change how we develop the Power BI models in the (near) future.
@@bobrubocki5305 Thanks. I will try. I know. The issue is the premium aspect.