Sr3CuIrO6 shows two magnetic phase transitions at ~16K and 5K. Both phases exhibit evidenceof being spin glasses or spin glass-like and from AC susceptibility measurements and no long range order has been observed in detailed neutron diffraction measurements. We suggest that both phases are disordered, but they are distinct, with different origins. The high temperature phase, we postulate, is the result of disorder due to phase separation or hole doping, combined with geometric frustration leading to the formation of interacting magnetic clusters. The lower temperature phase is more likely to arise from geometric frustration. We propose to perform a SANS study which allow us to determine whether the cluster model is appropriate and determine the size of the clusters and the distribution of those clusters.