Forum Discussion
Resolver IDifference: Real-time streaming vs. Direct Query
Hey,
this is a good question, there are three kind of data connectivity modes, but most of the time just two of them are considered. Each mode connects to a Power BI dataset, these modes are
- real-time (push)
- direct query / live connection
- import
The above modes are ordered by the duration between the origin of data and its availability for analysis in Power BI. As one might guess by the name, real-time provides the smallest latency between data origin and analytical availability. The data will be pushed into a Power BI dataset, much more about real-time datasets and some subtle but important differences between the 3 kinds of real-time datasets is described in this article: https://docs.microsoft.com/en-us/power-bi/service-real-time-streaming So, the answer to your question, real-time streaming data is considered push mode.
Whereas data from data sources connected using the import mode, will be imported to a Power BI dataset on a given schedule. Both connectivity modes (real-time / import) have this in common: data will be stored in a Power BI dataset.
Quite often Power BI datasets connected to data sources using the direct query / live connection mode can be considered the oldest datasets as the data has to "loaded" into these data sources, e.g. a data warehouse or a SQL Server Analysis Services data model, but the above ordering helps me a lot to differentiate between these three modes.
Of course, each mode has advantages and disadvantages, import comes with the greatest flexibility and capability regarding data modeling and DAX usage, real-time mode comes with the least flexibility / capability.
From, my point of view, the following is the greatest advantage of a data source that is connected to Power BI dataset using the direct query / live connection mode:
- data is not copied to a Power BI dataset, meaning there is basically no dataset size limit .
Hopefully this adds some additional insights.
Regards,
Tom
wow thanks very much Tom, did not expect such a comprehensive and professional answer.
Just want to add on one thing, not all real time streaming methods store data in the dataset. Find more at https://docs.microsoft.com/en-us/power-bi/service-real-time-streaming
Super UserHey,
this is a good question, there are three kind of data connectivity modes, but most of the time just two of them are considered. Each mode connects to a Power BI dataset, these modes are
- real-time (push)
- direct query / live connection
- import
The above modes are ordered by the duration between the origin of data and its availability for analysis in Power BI. As one might guess by the name, real-time provides the smallest latency between data origin and analytical availability. The data will be pushed into a Power BI dataset, much more about real-time datasets and some subtle but important differences between the 3 kinds of real-time datasets is described in this article: https://docs.microsoft.com/en-us/power-bi/service-real-time-streaming So, the answer to your question, real-time streaming data is considered push mode.
Whereas data from data sources connected using the import mode, will be imported to a Power BI dataset on a given schedule. Both connectivity modes (real-time / import) have this in common: data will be stored in a Power BI dataset.
Quite often Power BI datasets connected to data sources using the direct query / live connection mode can be considered the oldest datasets as the data has to "loaded" into these data sources, e.g. a data warehouse or a SQL Server Analysis Services data model, but the above ordering helps me a lot to differentiate between these three modes.
Of course, each mode has advantages and disadvantages, import comes with the greatest flexibility and capability regarding data modeling and DAX usage, real-time mode comes with the least flexibility / capability.
From, my point of view, the following is the greatest advantage of a data source that is connected to Power BI dataset using the direct query / live connection mode:
- data is not copied to a Power BI dataset, meaning there is basically no dataset size limit .
Hopefully this adds some additional insights.
Regards,
Tom
Resolver Iwow thanks very much Tom, did not expect such a comprehensive and professional answer.
Just want to add on one thing, not all real time streaming methods store data in the dataset. Find more at https://docs.microsoft.com/en-us/power-bi/service-real-time-streaming
- TomMartens
Hey Penn,
sure, you are right PubNub is not a Power BI dataset, but, as I provided the link to the official msft docs where this is explained I resigned to incorporate this into my answer.
Regards,
Tom