Future generations of mobile networks will require increased backhaul capacity to connect a massive amount of small cells (SCs) to the network. Since having an optical connection to each SC might be infeasible, mmWave links are an interesting alternative due to their large available bandwidth. An advantage of a wireless backhaul is that the topology can be reconfigured to adapt to changing traffic demands, new operator policies, or to rapidly overcome network failures. In this work, we investigate the problem of orchestrating the reconfiguration of mmWave wireless backhaul networks with mechanically steerable antennas assuming green backhaul operation where nodes are turned off when not in use. The orchestration involves scheduling and coordinating the powering on/off of nodes, the rotation of antennas to achieve alignment for link establishment, and setting up and tearing down links to minimize packet loss during the reconfiguration. We model the problem as a Mixed Integer Linear Program (MILP) for optimal orchestration and propose a sub-optimal reduced MILP for larger instances. Numerical results for different topologies using a realistic traffic trace indicate that optimizing reconfiguration orchestration can significantly reduce packet loss in comparison to a straightforward reconfiguration approach, enabling a smooth transition between target mmWave backhaul topologies.