Skip to content

Latest commit

 

History

History
100 lines (49 loc) · 4.44 KB

File metadata and controls

100 lines (49 loc) · 4.44 KB

Lab 7: Dashboarding on Hyperscale (Citus) using Power BI

Using Power BI, you get the ability to centralize your data, simplify it with queries you create, and share it in highly visual reports. Power BI has both direct query and import support for PostgreSQL i.e you can both generate reports that directly run queries/aggregates on the database or by pulling and caching data in Power BI. These features work with both Single Server and Hyperscale (Citus) deployment options for Azure Database for PostgreSQL.

In real-time analytics workloads, you can use Power BI to query both rollup tables and raw tables in Hyperscale (Citus). With larger datasets you could use the direct query functionality for better scalability. Because of sharding and parallelism across multiple nodes dashboard queries would be super snappy. For smaller datasets you could use the import functionality to create reports. Along with performance, Hyperscale (Citus) also lets you have large number of users concurrently running reports in Power BI.

In this exercise, you will connect to your PostgreSQL database cluster in Power BI Desktop and view/create advanced visualizations:

NOTE: While performing steps, If Login prompt appears. Login using odl user credentials found in Environment Details.

1.Open PowerBI application by clicking on PowerBI icon on the left side on the desktop.

2.Select Get Data as shown below:

3.A Get Data block will appear. Select Database > PostgreSQL database and then Connect.

4.Now connect to the database by providing your Server name and Database name (copy both from environment details tab). For Data Connectivity Mode select Direct query. Then select Ok.

5.Navigate to the Environment Details tab, copy and paste your PostgreSQL Database Username and PostgreSQL Database Password in their respective blocks as shown below:

Then select Connect.

6.Now in the Navigator select public.http_request_1min table given under the server and select Load.

7.Then in the right hand side in PowerBI, review the table you selected in the previous step under Fields.

8.Now select Clustered Column Chart as shown below. Then drag site_id under Axis.

9.Now drag error_count, request_count and success_count under Value as shown below:

10.Now add filter by dragging ingest_time under Filters on this page.

11.It will look like as shown below. Then select Filter Type as Basic Filtering from the dropdown.

12.Now select a random ingest time as shown below.

13.The final graph will look similar to the one shown below.

You can explore more on the PowerBI Dashboard to visualize the data. As a reference, we've another PowerBI dashboard draft file saved on desktop which includes a more comprehensive pre-built PowerBi dashboard for the same data. Let's take a look at that:

14.Now Open file Real-Time-Data-Analytics-Draft given on the VM desktop.

15.This will give a pop-up asking for the username and password, select Cancel for it and then Close the Cannot load model note.

16.Select Got it for the popup- Introducing the new filter blade and for Welcome to ArcGIS Maps for PowerBI select Ok.

17.Now go to Transform data and from the dropdown arrow select Data Source Settings.

18.The Data Source Settings tab will open, select the given server hostname and then click on Change Source.

19.Enter your server hostname and database name from the Environment Details tab. Then select Ok and then Close.

20.You will get a warning saying There are pending changes in your queries that haven't been applied., here select Apply Changes.

21.This will direct you to the PowerBI Report.

22.You can explore the dashboard and report configuration here.

23.Click Next on the bottom right of this page.