This issue has bugged me for the last week and everytime the report gets broken I couldn't find any other solution than re-creating dataset, report and dashboards from scratch.
I'm getting the data from Azure stream analytics and Power BI is the output. I haven't manually created the streaming dataset and let the ASA do its thing with all the input columns as pbi output. After creating the report and a dashboard from that dataset, I enjoyed my working real-time dashboard for some time.
Then I've decided to add a case condition to ASA job query, and instead of "select *", I've written all the field names and a case statement with the same field name as before. I've even copy-pasted from the dataset sample raw data not to make any typos. As soon as I started back the job, all the fields in the report got "Can't display visual" errors.
The error detail says "This visual contains one or more filters with deleted columns, type mismatches, or other breaking modeling changes." Even modifying the expression with the same column name and type, the model breaks. Another error gives "Something's wrong with one or more fields" and does not recognize the same field names on the report table's field list.
Odd thing is, the dashboard is still working fine, updating the real-time data with the latest changes I made with the case statement.
There are other cases where the report's data model breaks and all the visuals get these errors, and I have to start from scratch, such as modifying data types on the dataset, disabling/enabling the historic data. I'm not mentioning those, not to stray further from topic, but I've come to learn how delicate the real-time streaming datasets are.
Thanks for any thoughts.
Image may be NSFW.
Clik here to view.
Image may be NSFW.
Clik here to view.
