Executing Data Job duplicates data

Hi, a fairly basic question from a new user I expect!

I’m currently carrying out the Implementation Professional training. When I Execute an extraction it seems to stack a new data set on top of the first. My initial response was to make the Execution a ‘Full Load’ rather than a ‘Delta Load’, but when I ran this, it ran for 2.5 hours without finishing and I had to close it down.

Can anyone provide advice on next steps?

Thanks!
Dan