With 32 drives you should have at least 3 failure tolerances. Might even want to plan some hot spares in there. The idea is to build a large file system that won't die and take your data with it - nor introduce downtime (with a suitable failure tolerance). And of course, have an offsite backup for actually backing up your data.
I think an observer is a quantum process interacting with another quantum process. Photons are fundamentally different from particles with mass - and they interact differently. Two photons interact like waves at the surface of an ocean - amplitudes and phases changing, for instance. But photons interact with particles with mass by either being absorbed and, later, re-emitted by the other particle or being completely absorbed and altering the particle with mass' state in some way (speed/momentum).
I remember reading somewhere about a geometry which described photons' existences as fundamentally different from particles with mass. Particles with mass have lifelines that change over the course of time. Particles without mass, such as photons, don't experience time and never change themselves. Photons etch out a straight line in this geometry from fixed points to other fixed points. The fixed points represent particles with mass. (something like that)
I find it funny how people are so ready to believe the universe must be computing the same way our models must. The universe simply exists - it's not necessary for it to adhere to our human perceptions - not be a long shot. There's a vast gulf between the model and the modeled. That our theories quickly become uncomputable says more about their unfitness than the universe's functioning. To say otherwise is to negate the fundamental autonomy of the universe. We should be humbled by the universe precisely because our best models to understand it quickly diverge to where our minds and tools may not.