Starting December 3, join live sessions with database experts and the Microsoft product team to learn just how easy it is to get started
Learn moreGet certified in Microsoft Fabric—for free! For a limited time, get a free DP-600 exam voucher to use by the end of 2024. Register now
When creating a report using live connection to PBI dataset stored in premium capacity workspace, am I right that it works similar to Direct Query and too many simultaneous queries to the dataset (for example from different analysts) may cause high load on PBI node? Or live connection doesn't affect on PBI node performance?
Solved! Go to Solution.
Yes, this creates something similar to direct Query but against the underlying Tabular data set.
Tabular is very good at dealing with concurrent requests (Better than something like SQL) so the additional load should not usually be a problem unless you have an inefficient model and queries or have a lot of concurrent connections. (what a lot means depends on the size of the models and the capacity)
bottom line is that there will be an impact but it should not be significant.
Yes, this creates something similar to direct Query but against the underlying Tabular data set.
Tabular is very good at dealing with concurrent requests (Better than something like SQL) so the additional load should not usually be a problem unless you have an inefficient model and queries or have a lot of concurrent connections. (what a lot means depends on the size of the models and the capacity)
bottom line is that there will be an impact but it should not be significant.
Ok, I guess I just need to try and monitor the node performance. My goal is to optimize datasets in PBI service to avoid storing of similar data, but at the same time to make sure it won't decrease report performance in web.
@Anonymous , If the live connection is on SSAS, then it like a direct query.
But if it on power bi dataset, then it depends on that
you can also check this blog -https://radacad.com/live-connection-when-power-bi-comes-hybrid
Starting December 3, join live sessions with database experts and the Fabric product team to learn just how easy it is to get started.
User | Count |
---|---|
94 | |
90 | |
84 | |
76 | |
49 |
User | Count |
---|---|
143 | |
141 | |
109 | |
69 | |
55 |