Improve performance of the cursor comparison. #780

We are currently experiencing on high set of data some performance issues related to way we performing checks.
We are looking to rewrite the code so that will benefit from sql joins comparison instead of a hash being calculated.
This will be only done for data that can be compared that way.