We like to think of our solar system as being stable, and there is plenty of evidence that the Earth's orbit has not changed a lot in the several billions of years of life's existence.
It was slightly shocking then, that extrasolar planets discovered to date don't necessarily seem to have behaved this way. In particular, a number of so-called hot Jupiters have been discovered - Jupiter mass planets in very close orbits around their stars. It seems unlikely that they could have formed there, since the environment wouldn't have been cool enough for rocks and metals to condense into planetesimals.
Planetary migration had been considered in theory before the discovery of the hot Jupiters, and some migration is believed to have occurred in the outer planets of our own solar system, so the discoveries were hardly impossible to fit into our theories of planetary formation. The mechanisms are somewhat complicated, and include the interactions with density waves induced in the protoplanetary disk of gas and dust (the so-called Lindblad resonances) as well as potential interactions with planetesimals and other planets and even the star which is orbited. The principle involved, though, is very simple: the migrant exchanges angular momentum with other system components, losing angular momentum to move inward or gaining it to move outward. The amount of angular momentum involved to produce a hot Jupiter is quite large.
If my arithmetic is right, the angular momentum of a body in orbit scales like the square root of orbital radius, so changing from a Jovian orbit to one ten or twenty times smaller means shedding a lot of angular momentum. Given that Jupiter's angular momentum is 60% of the Solar system's total, that's a lot. So presumably a lot of stuff must have gotten booted out of their systems.