Fast Distributed k-Center Clustering with Outliers on Massive Data