PowerScale OneFS HDFS Configuration Guide

Configuring virtual HDFS racks

You can create a virtual HDFS rack of nodes on your PowerScale cluster to optimize performance and reduce latency when accessing HDFS data.

OneFS enables you to specify a group of preferred HDFS nodes on your PowerScale cluster and an associated group of Hadoop compute clients as a virtual HDFS rack. Virtual HDFS racks allow you to fine-tune client connectivity by directing Hadoop compute clients to go through quicker, less-busy switches or to faster nodes, depending on your network topology.

When a Hadoop compute client from the specified group connects to the cluster, OneFS returns at least two IP addresses from the group of preferred HDFS nodes. You specify the preferred HDFS nodes by IP address pool. Virtual HDFS racks do not support IP address pools in the IPv6 family.