Full refresh of a df with Incremental

[smpa01]

I have inherited a dataflow which is currently on incremental refresh but was not well maintained at all. As a result, it is missing data from some partitons that are not part of the partitions that are currently passed on to refresh.

I want to initiate a full refresh first before switching back to full refresh of the dataflow to ensure I have all I need.

How do I do it?

@GilbertQ @AlexisOlson @bcdobbs 

[lbendlin]

Very simple - change the query,  Any meta data change will trigger a full refresh, and the next refresh will be incremental again.  

[smpa01]

@lbendlin many thanks for this and it is an awesome trick. But unfortunately, I can't change the existing query of tblMain which has Existing IR cause there is no scope of change. However, I can add a another table as

let

 Source =1

In

Source

Will this trigger the full refresh of tblMain?

 If you can please advise.

I really don't want to turn off the existing IR, do a full refresh and turn on the IR again. But what you are suggesting, looks like I don't have to do that but how do I apply what you are suggesting without creating any havoc.

[lbendlin]

I really don't want to turn off the existing IR, do a full refresh and turn on the IR again

What I proposed is not much different. I don't know if bootstrapping is even possible for dataflows (I know it can be done for datasets).  So net-net if you are making any structural changes your Incremental refresh will be replaced with a full one-time refresh of all partitions before resuming.