There's no need to fade up the second, lower gain track as the gain drops from the first track. Assuming you're only ducking a track behind, for example, someone talking.
Track 2 (I'm skipping track 1 - talking) has the music soundtrack, running at some level. When the talking starts, that track should drop under the speaking.
Track 3 has the same content as Track 2, and it's at some lower level (gain). Edit Track 2 to end exactly where the speaking begins. Fade out back before the cut (that is, start the fade before speaking begins). You'll effective drop the gain to Track 3's level just as if you moved a console's slider.
This works because Track 2 and Track 3 are in sync and have the same material. If 2 and 3 have different material, a crossfade is needed. Oh, and boosting the final gain is only the dropping process in reverse.