Share dataset between machines


we have a cluster of n machines, and to run we log into the system

ssh proof

in this way you are redirected to one node of the cluster, let say proof-06. From here I run proof


the problem is that if I subscribe a dataset on proof-06 and I relogin on proof I will be redirected to another machine (proof-01, …) where that dataset is no visibile because every machines have their own sandbox.

Is there a way to share them?


Do the machines share some disk devices?
In such case the solution is to have the dataset directory on such a device (see