In our working group, my colleagues and I sometimes let our jobs run on each others computers to better utilize our resources. Is there a way to use snakemake --cluster in this scenario without a "proper" cluster setup? Is there a qsub implementation that simply executes the supplied command on other computers via SSH? (Without requiring the target node to be part of a formal cluster?)
If I understand correctly, snakemake --cluster assumes the same file system to be mounted on all cluster nodes. This is partial true in our case, as our home directories are on NFS.