"total number" would be tough as we do not know the operating frequency range.
But you can simulate them with specified freqeuncy range, and use more randomized dipole sources so they can exite all the modes the cavity supports. You may also add a few different analysis groups in different locaitons, each has a few monitors. This is better than to have one analysis group with many time monitors, I personally believe. Please refer the example: Photonic crystal cavity