When testing out the REST API, I created a push dataset with a table called Contacts. This table was given two columns: "id" as an Int64 and "name" as a string. If I insert two rows with the same name value, I end up with one row containing an id value which is the sum of the value from both records. This appears to implicitly make the "name" column a primary key, with the implicit behavior of adding any numeric fields together. To make matters worse, I can't find any way to view the rows after they are inserted. Not from the API, the service, or the desktop app. How am I supposed to know if the ids are even being added together or if this is just how every visualization chooses to display the data? This black-box style of storage and configuration is quite unappealing compared to any SQL database experience.
I've spent a lot of time reading documentation for Power BI, including the information for push datasets (which is what I set the "defaultMode" to). I did not find anything mentioning this sort of behavior. I'm really hoping there is some sort of reference explaining how to control the primary key (if there is one?) and what happens for duplicate key inserts. For my use case, I need the primary key to be a number and, if possible, replace values if a row is inserted with duplicate key, not add them together.
