Data incorrect with Incremental refresh

[ChayDommeti]

I have a Power BI report which is getting its data from databricks. When i refresh the complete report and publish i get the correct values on power BI matching the data in databricks but if i implement incremental refresh on the data and filter rows for June month, all the data till June shows up correctly but messes up the data after June.

I tried to compare the data from databricks to Power BI report (for July and August) and i have observed that wrong value or an extra value is adding up every day at 12 AM . Any help here is appreciated. 

Below is the date column on which i am applying filter for June month 

and here are the parameters

and below is the incremental refresh implementation

and here is the output comparision 

[Rakeshchandra_9]

I aslo facing the same issue but after the solution which is provided issue got resolved but still i'm not getting how this (=) on both the parameters will be the issue for duplication and that too data on 1st day of the month is getting duplicated for all the remianing days in the month it is fine. Can anyone explain this logically because unable to confirm this change to client.

[lbendlin]

Let's say you have two yearly partitions

2019:   Date between Jan 1 2019 and Jan 1 2020

2020:   Date between Jan 1 2020 and Jan 1 2021

That will mean you will duplicate data for Jan 1 2020

Instead the rule must be 

2019:   Date >= Jan 1 2019 and  < Jan 1 2020

2020:   Date >= Jan 1 2020 and < Jan 1 2021

[Rakeshchandra_9]

Thanks for the quick response.
In my scenari0 im following below rules for the incremental policy. But the values are getting doubled on first of every month. Let's say 20190601 in SQL has sales of 2000$ and when the first dataset refresh the value for the same date 20190601 is coming as 4000$

2019:   Date between Jan 1 2019 and Dec 31 2019

2020:   Date between Jan 1 2020 and Dec 31 2020

Currently the rule has been set as below: 

2019:   Date >= Jan 1 2019 and  <= Dec 31 2019

2020:   Date >= Jan 1 2020 and <= Dec 31 2020

[lbendlin]

Currently the rule has been set as below: 

2019:   Date >= Jan 1 2019 and  <= Dec 31 2019

2020:   Date >= Jan 1 2020 and <= Dec 31 2020

No, it hasn't.  The rules are set by the Power BI Service when it manages the partitions.

All you set are the outer boundaries.

[lbendlin]

You must not include both RangeStart and RangeEnd. One of them has to be excluded.

Either RangeStart < date <= RangeEnd  or RangeStart <= date < RangeEnd.

Duplication will happen if you do not follow that rule.

[ChayDommeti · Accepted solution]

That helped , Also found in Microsoft documention that i shouldnt include equal in both the parameters,  I have changed my existing to [usage_start_time] > RangeStart and [usage_start_time] <= RangeEnd and that fixed the duplication issue