Imagine you have a photon with momentum $p$ which hits a mirror $M$ and is reflected back.
Since $M$ is at rest in your frame of reference the total momentum of the system pre-reflection is $p$ and post reflection is $-p$.
This seems quite odd to me and a clear violation of conservation of momentum.
How this can be resolved:
The mirror begins to move/absorbs some momentum after being hit by the photon, but if the photon was reflected with momentum $p$ then we end up having conservation of kinetic energy being violated.
The only conclusion that's reasonable is the photon being reflected back has a different (probably lesser) momentum after hitting the mirror and the mirror then picks up some positive momentum, so that the sum of the momenta remains constant and the sum of their kinetic energies remains constant.
Is this actually what happens? And if so does that mean that almost all mirrors are heating up when they are reflecting photons?