[VMtest]

if that's the case, just restart the failing nodes

and of course it's not that simple, they still have to customize the workflow