Reputation: 1
In the current architecture, Power BI datasets are connected to oracle database with gateways and datasets on the service gets refreshes as per schedule.
However, due to inconsistent of oracle due to huge code and huge data in tables, cline is looking to change the entire architecture into Kafka and Cassandra database which will be real-time data. So, client is looking for real time data on power BI so, how can we implement streaming datasets which should connect to Cassandra database and push the real-time data to dashboards..
Thanks!
Upvotes: 0
Views: 489
Reputation: 4967
If you have huge data, then Power BI streaming datasets will not be suitable for this type of workload. Please review the limitations of Power BI Streaming Dataset here. You haven't mentioned the use case of what streaming data requirements you need, but streaming dataset sizes are limited to 200,000 rows for those that are the 1 hour history retention, or 5,000,000 rows for datasets with a retention policy.
If the data volumes are low enough you would have to send data from Cassandra, to Azure Event Hubs, to Streaming Analytics to Power BI. If your data volumes are large, then you will need to push the data to a database that can handle streaming, for example Azure Databricks, then connect to Power BI on that data or use Databricks to push it to a database, so Power BI can use direct query mode and get the latest data when queried.
Upvotes: 0