Incremental Refresh for Dataflows

[jtbint]

Hi all,

I have searched some specific questions I have on this topic, but can't seem to find any existing answers. Hoping someone can enlighten me.

We are using Dataflows to pull from SAP HANA, but the source table does not have a datetime column (only date). From what I have seen from searching online, incremental refresh on datasets allow you to use functions to convert the date column to datetime, and so this is a non-issue. But that does not seem to be the case with dataflows in the service (at least according to the microsoft documentation). Will I be able to use incremental refresh in the service if the source table has a date column rather than a datetime column? I have attempted to get around this by converting the column to datetime within the query itself, but refresh was extremely slow, even slower than refreshing the entire table. I should also mention that currently the query uses Date.AddMonths to filter out any rows that are older than 1 year. Wasn't sure if I needed to get rid of that filter in order for incremental refresh to work correctly.

Any insight would be greatly appreciated, this forum has been very helpful.

[Greg_Deckler]

@jtbint What about using a surrogate integer key like yyyymmdd?