As Hadoop clusters grow in size and data volume, it becomes more and more useful to share them between multiple users and to isolate these users. If User 1 is running a ten-hour machine learning job for example, this should not impair a User 2 from running a 2-minute Hive query. In November, I blogged [...]