Filament-wound tanks
Such tanks are routinely designed for hundreds and hundreds of high-pressure cycles, but I suspect that the Shuttle tanks, although they see fewer cycles in normal use, are coasting quite a bit closer to the design limits than, say, a filament-wound scuba tank.
I'm hoping someone is keeping an eye on this. Properly-designed filament-wound tanks do not fail in a catastrophic (explosive) manner, but it's still something you really wouldn't want to have happen inside a Shuttle equipment bay.