A while back I built a Cloudera 4.x Hadoop Cluster using Cloudera Manager. The cluster runs fine and there are no issues; however when I was troubleshooting a bad drive indicated by smartctl I was curious to see if the drive was actually being used by Hadoop and to my suprise I found every node only using 1 drive out of 4. Each node has 4 2T drives configured asJBOD and only the system drive was configured to be used as hdfs space? I dont recall anything during the Automated install process that specified adding drives to HDFS? Did I miss something or is this supposed to be part of the install? I just assumed that Cloudera Manager used all the drives on each server during set up. Luckily space has not been an issue yet but it will be soon.