We focus on the routing, spatial mode, and spectrum allocation (RSSA) problem-a basic optimization problem in spectrally and spatially flexible optical networks (SS-FON). RSSA is a very challenging problem since it introduces a new dimension, related to the selection of spatial resources, to the already complex (NP-hard) routing and spectrum allocation (RSA) problem. To allow solving large RSSA problem instances, in particular, in optical backbone networks with tens of nodes and hundreds of demands as well as with optical fibers supporting a numerous number of spatial modes, specialized optimization algorithms are required. In this paper, we propose and study several dedicated optimization procedures and some enhancements in algorithm processing, including parallel processing, which aim at both speeding up and increasing the effectiveness of the RSSA process and at estimating the quality of generated solutions. We combine the proposed procedures into an efficient optimization framework which, as presented numerical results show, is capable of providing high-quality solutions to large instances of the RSSA optimization problem in reasonable computation times. As a case study scenario, we consider an SS-FON with spectral super-channels (SChs) transmitted over bundles of single-mode fibers (SMFB) without spatial mode conversion. Nonetheless, the proposed optimization framework is generic and can be straightforwardly adapted to other SS-FON scenarios.