Ah no, we use Slurm. So JLab runs on a head node, and users submit jobs through Slurm. Slurm would let you set a ipcluster which you then drive from a JLab notebook, though, which is a nice set up.