We've had a lot of these recently on a pair of 9450 arrays with 3.3.1MU3.
They all seemed to resolve instantly but annoying as you don't get resolve messages normally so still have to check.
I suspect it's just occasional packet drops (everything is 'always fine' on our network, at least according to our network team
).
I seem to have fixed it (touch wood) by changing which node has the active network interface, all four of our nodes have the admin interface connected and they are split across different network switch stacks so maybe one of the stacks has a busy uplink.
We also couldn't get into our QW VM, even though we do know the password, however one of our VM team managed to use some admin/hack tools to reset the password after taking the VM down. It only takes 10 mins to deploy a new one though so that was our fallback plan.