Clustering fails to shutdown properly

Both nodes fail to shutdown properly. Although the log says 'Shutdown complete' the processes keep running. (That's how I got the thread dumps.)