I have a SQL database that contains datetime columns in some of the tables. The locale of the server is NZ English so the datetimes come through as 15/01/2020, 12:00:00 am
This comes through fine when I connect directly from PBI desktop to the server, and I can work with these dates no problem. Also, if I connect a PBI dataflow to the server via a on prem gateway all is good, the date columns show up just fine in the query editor in the cloud.
However, if I connect PBI desktop to the dataflow, the datetime columns error out with the "We couldn't parse the input provided as a DateTime value" message.
I have figured I can resolve this for columns where it's a date only (i.e. the time is always 12:00:00 am) by converting using Locale to Date > English (New Zealand) in the dataflow.
However, some of the columns do actually have a time that I want to be available in PBI desktop, but when I convert using locale to DateTime or DateTimeZone, they still error out in PBI desktop.
I guess this is some incompatible combination of locale settings between the SQL server, powerbi.com and PBI desktop.
I'd love to be able to resolve this without resorting to either splitting out the time as a string and recombining or just ignoring the time and working only with dates. I have a lot of tables and a lot of columns I want to add to my dataflows so the easiest workaround/solution to this is what I'm going for.
