> > yes, I guess that prevents the issue most times (but it is not 100%) > > mhm i can't imagine a situation where this can happen. The important part is > that the expiretime is BIGGER than the kvm timeout. Ah. What exactly is the kvm timeout?